Network Working Group                                            J. Dunn
Request for Comments: 3116                                     C. Martin
Category: Informational                                        ANC, Inc.
                                                              June 2001


                   Methodology for ATM Benchmarking

Status of this Memo

  This memo provides information for the Internet community.  It does
  not specify an Internet standard of any kind.  Distribution of this
  memo is unlimited.

Copyright Notice

  Copyright (C) The Internet Society (2001).  All Rights Reserved.

Abstract

  This document discusses and defines a number of tests that may be
  used to describe the performance characteristics of ATM (Asynchronous
  Transfer Mode) based switching devices.  In addition to defining the
  tests this document also describes specific formats for reporting the
  results of the tests.

  This memo is a product of the Benchmarking Methodology Working Group
  (BMWG) of the Internet Engineering Task Force (IETF).

Table of Contents

  1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  4
  2. Background . . . . . . . . . . . . . . . . . . . . . . . . . .  5
  2.1. Test Device Requirements . . . . . . . . . . . . . . . . . .  5
  2.2. Systems Under Test (SUTs). . . . . . . . . . . . . . . . . .  5
  2.3. Test Result Evaluation . . . . . . . . . . . . . . . . . . .  5
  2.4. Requirements . . . . . . . . . . . . . . . . . . . . . . . .  5
  2.5. Test Configurations for SONET. . . . . . . . . . . . . . . .  6
  2.6. SUT Configuration. . . . . . . . . . . . . . . . . . . . . .  7
  2.7. Frame Formats. . . . . . . . . . . . . . . . . . . . . . . .  8
  2.8. Frame Sizes. . . . . . . . . . . . . . . . . . . . . . . . .  8
  2.9. Verifying Received IP PDU's. . . . . . . . . . . . . . . . .  9
  2.10. Modifiers . . . . . . . . . . . . . . . . . . . . . . . . .  9
  2.10.1. Management IP PDU's . . . . . . . . . . . . . . . . . . .  9
  2.10.2. Routing Update IP PDU's . . . . . . . . . . . . . . . . . 10
  2.11. Filters . . . . . . . . . . . . . . . . . . . . . . . . . . 10
  2.11.1. Filter Addresses. . . . . . . . . . . . . . . . . . . . . 11
  2.12. Protocol Addresses. . . . . . . . . . . . . . . . . . . . . 12



Dunn & Martin                Informational                      [Page 1]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  2.13. Route Set Up. . . . . . . . . . . . . . . . . . . . . . . . 12
  2.14. Bidirectional Traffic . . . . . . . . . . . . . . . . . . . 12
  2.15. Single Stream Path. . . . . . . . . . . . . . . . . . . . . 12
  2.16. Multi-port. . . . . . . . . . . . . . . . . . . . . . . . . 13
  2.17. Multiple Protocols. . . . . . . . . . . . . . . . . . . . . 14
  2.18. Multiple IP PDU Sizes . . . . . . . . . . . . . . . . . . . 14
  2.19. Testing Beyond a Single SUT . . . . . . . . . . . . . . . . 14
  2.20. Maximum IP PDU Rate . . . . . . . . . . . . . . . . . . . . 15
  2.21. Busty Traffic . . . . . . . . . . . . . . . . . . . . . . . 15
  2.22. Trial Description . . . . . . . . . . . . . . . . . . . . . 16
  2.23. Trial Duration. . . . . . . . . . . . . . . . . . . . . . . 16
  2.24. Address Resolution. . . . . . . . . . . . . . . . . . . . . 16
  2.25. Synchronized Payload Bit Pattern. . . . . . . . . . . . . . 16
  2.26. Burst Traffic Descriptors . . . . . . . . . . . . . . . . . 17
  3. Performance Metrics. . . . . . . . . . . . . . . . . . . . . . 17
  3.1. Physical Layer-SONET . . . . . . . . . . . . . . . . . . . . 17
  3.1.1. Pointer Movements. . . . . . . . . . . . . . . . . . . . . 17
  3.1.1.1. Pointer Movement Propagation . . . . . . . . . . . . . . 17
  3.1.1.2. Cell Loss due to Pointer Movement. . . . . . . . . . . . 19
  3.1.1.3. IP Packet Loss due to Pointer Movement . . . . . . . . . 20
  3.1.2. Transport Overhead (TOH) Error Count . . . . . . . . . . . 21
  3.1.2.1. TOH Error Propagation. . . . . . . . . . . . . . . . . . 21
  3.1.2.2. Cell Loss due to TOH Error . . . . . . . . . . . . . . . 22
  3.1.2.3. IP Packet Loss due to TOH Error. . . . . . . . . . . . . 23
  3.1.3. Path Overhead (POH) Error Count. . . . . . . . . . . . . . 24
  3.1.3.1. POH Error Propagation. . . . . . . . . . . . . . . . . . 24
  3.1.3.2. Cell Loss due to POH Error . . . . . . . . . . . . . . . 25
  3.1.3.3. IP Packet Loss due to POH Error. . . . . . . . . . . . . 26
  3.2. ATM Layer. . . . . . . . . . . . . . . . . . . . . . . . . . 27
  3.2.1. Two-Point Cell Delay Variation (CDV) . . . . . . . . . . . 27
  3.2.1.1. Test Setup . . . . . . . . . . . . . . . . . . . . . . . 27
  3.2.1.2. Two-point CDV/Steady Load/One VCC. . . . . . . . . . . . 27
  3.2.1.3. Two-point CDV/Steady Load/Twelve VCCs. . . . . . . . . . 28
  3.2.1.4. Two-point CDV/Steady Load/Maximum VCCs . . . . . . . . . 30
  3.2.1.5. Two-point CDV/Bursty VBR Load/One VCC. . . . . . . . . . 31
  3.2.1.6. Two-point CDV/Bursty VBR Load/Twelve VCCs. . . . . . . . 32
  3.2.1.7. Two-point CDV/Bursty VBR Load/Maximum VCCs . . . . . . . 34
  3.2.1.8. Two-point CDV/Mixed Load/Three VCC's . . . . . . . . . . 35
  3.2.1.9. Two-point CDV/Mixed Load/Twelve VCCs . . . . . . . . . . 36
  3.2.1.10. Two-point CDV/Mixed Load/Maximum VCCs . . . . . . . . . 38
  3.2.2. Cell Error Ratio (CER) . . . . . . . . . . . . . . . . . . 39
  3.2.2.1. Test Setup . . . . . . . . . . . . . . . . . . . . . . . 39
  3.2.2.2. CER/Steady Load/One VCC. . . . . . . . . . . . . . . . . 40
  3.2.2.3. CER/Steady Load/Twelve VCCs. . . . . . . . . . . . . . . 41
  3.2.2.4. CER/Steady Load/Maximum VCCs . . . . . . . . . . . . . . 42
  3.2.2.5. CER/Bursty VBR Load/One VCC. . . . . . . . . . . . . . . 43
  3.2.2.6. CER/Bursty VBR Load/Twelve VCCs. . . . . . . . . . . . . 44
  3.2.2.7. CER/Bursty VBR Load/Maximum VCCs . . . . . . . . . . . . 46



Dunn & Martin                Informational                      [Page 2]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  3.2.3. Cell Loss Ratio (CLR). . . . . . . . . . . . . . . . . . . 47
  3.2.3.1. CLR/Steady Load/One VCC. . . . . . . . . . . . . . . . . 47
  3.2.3.2. CLR/Steady Load/Twelve VCCs. . . . . . . . . . . . . . . 48
  3.2.3.3. CLR/Steady Load/Maximum VCCs . . . . . . . . . . . . . . 49
  3.2.3.4. CLR/Bursty VBR Load/One VCC. . . . . . . . . . . . . . . 51
  3.2.3.5. CLR/Bursty VBR Load/Twelve VCCs. . . . . . . . . . . . . 52
  3.2.3.6. CLR/Bursty VBR Load/Maximum VCCs . . . . . . . . . . . . 53
  3.2.4. Cell Misinsertion Rate (CMR) . . . . . . . . . . . . . . . 54
  3.2.4.1. CMR/Steady Load/One VCC. . . . . . . . . . . . . . . . . 54
  3.2.4.2. CMR/Steady Load/Twelve VCCs. . . . . . . . . . . . . . . 55
  3.2.4.3. CMR/Steady Load/Maximum VCCs . . . . . . . . . . . . . . 57
  3.2.4.4. CMR/Bursty VBR Load/One VCC. . . . . . . . . . . . . . . 58
  3.2.4.5. CMR/Bursty VBR Load/Twelve VCCs. . . . . . . . . . . . . 59
  3.2.4.6. CMR/Bursty VBR Load/Maximum VCCs . . . . . . . . . . . . 60
  3.2.5. CRC Error Ratio (CRC-ER) . . . . . . . . . . . . . . . . . 62
  3.2.5.1. CRC-ER/Steady Load/One VCC . . . . . . . . . . . . . . . 62
  3.2.5.2. CRC-ER/Steady Load/Twelve VCCs . . . . . . . . . . . . . 63
  3.2.5.3. CRC-ER/Steady Load/Maximum VCCs. . . . . . . . . . . . . 64
  3.2.5.4. CRC-ER/Bursty VBR Load/One VCC . . . . . . . . . . . . . 65
  3.2.5.5. CRC-ER/Bursty VBR Load/Twelve VCCs . . . . . . . . . . . 66
  3.2.5.6. CRC-ER/Bursty VBR Load/Maximum VCCs. . . . . . . . . . . 68
  3.2.5.7. CRC-ER/Bursty UBR Load/One VCC . . . . . . . . . . . . . 69
  3.2.5.8. CRC-ER/Bursty UBR Load/Twelve VCCs . . . . . . . . . . . 70
  3.2.5.9. CRC-ER/Bursty UBR Load/Maximum VCCs. . . . . . . . . . . 71
  3.2.5.10. CRC-ER/Bursty Mixed Load/Three VCC. . . . . . . . . . . 73
  3.2.5.11. CRC-ER/Bursty Mixed Load/Twelve VCCs. . . . . . . . . . 74
  3.2.5.12. CRC-ER/Bursty Mixed Load/Maximum VCCs . . . . . . . . . 75
  3.2.6. Cell Transfer Delay (CTD). . . . . . . . . . . . . . . . . 76
  3.2.6.1. Test Setup . . . . . . . . . . . . . . . . . . . . . . . 76
  3.2.6.2. CTD/Steady Load/One VCC. . . . . . . . . . . . . . . . . 77
  3.2.6.3. CTD/Steady Load/Twelve VCCs. . . . . . . . . . . . . . . 78
  3.2.6.4. CTD/Steady Load/Maximum VCCs . . . . . . . . . . . . . . 79
  3.2.6.5. CTD/Bursty VBR Load/One VCC. . . . . . . . . . . . . . . 81
  3.2.6.6. CTD/Bursty VBR Load/Twelve VCCs. . . . . . . . . . . . . 82
  3.2.6.7. CTD/Bursty VBR Load/Maximum VCCs . . . . . . . . . . . . 83
  3.2.6.8. CTD/Bursty UBR Load/One VCC. . . . . . . . . . . . . . . 85
  3.2.6.9. CTD/Bursty UBR Load/Twelve VCCs. . . . . . . . . . . . . 86
  3.2.6.10. CTD/Bursty UBR Load/Maximum VCCs. . . . . . . . . . . . 87
  3.2.6.11. CTD/Mixed Load/Three VCC's. . . . . . . . . . . . . . . 88
  3.2.6.12. CTD/Mixed Load/Twelve VCCs. . . . . . . . . . . . . . . 90
  3.2.6.13. CTD/Mixed Load/Maximum VCCs . . . . . . . . . . . . . . 91
  3.3. ATM Adaptation Layer (AAL) Type 5 (AAL5) . . . . . . . . . . 93
  3.3.1. IP Packet Loss due to AAL5 Re-assembly Errors. . . . . . . 93
  3.3.2. AAL5 Re-assembly Time. . . . . . . . . . . . . . . . . . . 94
  3.3.3. AAL5 CRC Error Ratio . . . . . . . . . . . . . . . . . . . 95
  3.3.3.1. Test Setup . . . . . . . . . . . . . . . . . . . . . . . 95
  3.3.3.2. AAL5-CRC-ER/Steady Load/One VCC. . . . . . . . . . . . . 95
  3.3.3.3. AAL5-CRC-ER/Steady Load/Twelve VCCs. . . . . . . . . . . 96



Dunn & Martin                Informational                      [Page 3]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  3.3.3.4. AAL5-CRC-ER/Steady Load/Maximum VCCs . . . . . . . . . . 97
  3.3.3.5. AAL5-CRC-ER/Bursty VBR Load/One VCC. . . . . . . . . . . 99
  3.3.3.6. AAL5-CRC-ER/Bursty VBR Load/Twelve VCCs. . . . . . . . .100
  3.3.3.7. AAL5-CRC-ER/Bursty VBR Load/Maximum VCCs . . . . . . . .101
  3.3.3.8. AAL5-CRC-ER/Mixed Load/Three VCC's . . . . . . . . . . .102
  3.3.3.9. AAL5-CRC-ER/Mixed Load/Twelve VCCs . . . . . . . . . . .104
  3.3.3.10. AAL5-CRC-ER/Mixed Load/Maximum VCCs . . . . . . . . . .105
  3.4. ATM Service: Signaling . . . . . . . . . . . . . . . . . . .106
  3.4.1. CAC Denial Time and Connection Establishment Time. . . . .106
  3.4.2. Connection Teardown Time . . . . . . . . . . . . . . . . .107
  3.4.3. Crankback Time . . . . . . . . . . . . . . . . . . . . . .108
  3.4.4. Route Update Response Time . . . . . . . . . . . . . . . .109
  3.5. ATM Service: ILMI. . . . . . . . . . . . . . . . . . . . . .110
  3.5.1. MIB Alignment Time . . . . . . . . . . . . . . . . . . . .110
  3.5.2. Address Registration Time. . . . . . . . . . . . . . . . .111
  4. Security Considerations  . . . . . . . . . . . . . . . . . . .112
  5. Notices. . . . . . . . . . . . . . . . . . . . . . . . . . . .112
  6. References . . . . . . . . . . . . . . . . . . . . . . . . . .113
  7. Authors' Addresses . . . . . . . . . . . . . . . . . . . . . .113
  APPENDIX A  . . . . . . . . . . . . . . . . . . . . . . . . . . .114
  APPENDIX B  . . . . . . . . . . . . . . . . . . . . . . . . . . .114
  APPENDIX C  . . . . . . . . . . . . . . . . . . . . . . . . . . .116
  Full Copyright Statement  . . . . . . . . . . . . . . . . . . . .127

1. Introduction

  This document defines a specific set of tests that vendors can use to
  measure and report the performance characteristics of ATM network
  devices.  The results of these tests will provide the user comparable
  data from different vendors with which to evaluate these devices.
  The methods defined in this memo are based on RFC 2544 "Benchmarking
  Methodology for Network Interconnect Devices".

  The document "Terminology for ATM Benchmarking" (RFC 2761), defines
  many of the terms that are used in this document.  The terminology
  document should be consulted before attempting to make use of this
  document.

  The BMWG produces two major classes of documents: Benchmarking
  Terminology documents and Benchmarking Methodology documents.  The
  Terminology documents present the benchmarks and other related terms.
  The Methodology documents define the procedures required to collect
  the benchmarks cited in the corresponding Terminology documents.








Dunn & Martin                Informational                      [Page 4]

RFC 3116            Methodology for ATM Benchmarking           June 2001


2. Background

2.1. Test Device Requirements

  This document is based on the requirement that a test device is
  available.  The test device can either be off the shelf or can be
  easily built with current technologies.  The test device must have a
  transmitting and receiving port for the interface type under test.
  The test device must be configured to transmit test PDUs and to
  analyze received PDUs.  The test device should be able to transmit
  and analyze received data at the same time.

2.2. Systems Under Test (SUTs)

  There are a number of tests described in this document that do not
  apply to each SUT.  Vendors should perform all of the tests that can
  be supported by a specific product type.  It will take some time to
  perform all of the recommended tests under all of the recommended
  conditions.

2.3. Test Result Evaluation

  Performing all of the tests in this document will result in a great
  deal of data.  The applicability of this data to the evaluation of a
  particular SUT will depend on its expected use and the configuration
  of the network in which it will be used.  For example, the time
  required by a switch to provide ILMI services will not be a pertinent
  measurement in a network that does not use the ILMI protocol, such as
  an ATM WAN.  Evaluating data relevant to a particular network
  installation may require considerable experience, which may not be
  readily available.  Finally, test selection and evaluation of test
  results must be done with an understanding of generally accepted
  testing practices regarding repeatability, variance and the
  statistical significance of a small numbers of trials.

2.4. Requirements

  In this document, the words that are used to define the significance
  of each particular requirement are capitalized.  These words are:

  *  "MUST" This word, or the words "REQUIRED" and "SHALL" mean that
     the item is an absolute requirement of the specification.

  *  "SHOULD" This word or the adjective "RECOMMENDED" means that there
     may exist valid reasons in particular circumstances to ignore this
     item, but the full implications should be understood and the case
     carefully weighed before choosing a different course.




Dunn & Martin                Informational                      [Page 5]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  *  "MAY" This word or the adjective "OPTIONAL" means that this item
     is truly optional.  One vendor may choose to include the item
     because a particular marketplace requires it or because it
     enhances the product, for example; another vendor may omit the
     same item.

  An implementation is not compliant if it fails to satisfy one or more
  of the MUST requirements for the protocols it implements.  An
  implementation that satisfies all the MUST and all the SHOULD
  requirements for its protocols is said to be "unconditionally
  compliant"; one that satisfies all the MUST requirements but not all
  the SHOULD requirements for its protocols is said to be
  "conditionally compliant".

2.5. Test Configurations for SONET

  The test device can be connected to the SUT in a variety of
  configurations depending on the test point.  The following
  configurations will be used for the tests described in this document.

  1) Uni-directional connection: The test devices transmit port
     (labeled Tx) is connected to the SUT receive port (labeled Rx).
     The SUTs transmit port is connected to the test device receive
     port (see Figure 1).  In this configuration, the test device can
     verify that all transmitted packets are acknowledged correctly.
     Note that this configuration does not verify internal system
     functions, but verifies one port on the SUT.

           +-------------+               +-------------+
           |           Tx|-------------->|Rx           |
           |    Test   Rx|<--------------|Tx   SUT     |
           |   Device    |               |             |
           +-------------+               +-------------+

                           Figure 1

  2) Bi-directional connection: The test devices first transmit port is
     connected to the SUTs first receive port.  The SUTs first transmit
     port is connected to the test devices first receive port.  The
     test devices second transmit port is connected to the SUTs second
     receive port.  The SUTs second transmit port is connected to the
     test devices second receive port (see Figure 2).  In this
     configuration, the test device can determine if all of the
     transmitted packets were received and forwarded correctly.  Note
     that this configuration does verify internal system functions,
     since it verifies two ports on the SUT.





Dunn & Martin                Informational                      [Page 6]

RFC 3116            Methodology for ATM Benchmarking           June 2001


           +-------------+               +-------------+
           |     Test  Tx|-------------->|Rx           |
           |    Device Rx|<--------------|Tx   SUT     |
           |    Tx   Rx  |               |   Tx   Rx   |
           +-------------+               +-------------+
                 |   ^                        |    ^
                 |   |                        |    |
                 |   +------------------------+    |
                 |                                 |
                 |---------------------------------|

                            Figure 2

  3) Uni-directional passthrough connection: The test devices first
     transmit port is connected to the SUT1 receive port.  The SUT1
     transmit port is connected to the test devices first receive port.
     The test devices second transmit port is connected to the SUT2
     receive port.  The SUT2 transmit port is connected to the test
     devices second receive port (see Figure 3).  In this
     configuration, the test device can determine if all of the packets
     transmitted by SUT1 were correctly acknowledged by SUT2.  Note
     that this configuration does not verify internal system functions,
     but verifies one port on each SUT.

  +-------------+           +-------------+           +-------------+
  |           Tx|---------->|Rx         Tx|---------->|Rx           |
  |     SUT1  Rx|<----------|Tx   Test  Rx|<----------|Tx   SUT2    |
  |             |           |    Device   |           |             |
  +-------------+           +-------------+           +-------------+

                             Figure 3

2.6. SUT Configuration

  The SUT MUST be configured as described in the SUT users guide.
  Specifically, it is expected that all of the supported protocols will
  be configured and enabled.  It is expected that all of the tests will
  be run without changing the configuration or setup of the SUT in any
  way other than that required to do the specific test.  For example,
  it is not acceptable to disable all but one transport protocol when
  testing the throughput of that protocol.  If PNNI or BISUP is used to
  initiate switched virtual connections (SVCs), the SUT configuration
  SHOULD include the normally recommended routing update intervals and
  keep alive frequency.  The specific version of the software and the
  exact SUT configuration, including what functions are disabled and
  used during the tests MUST be included as part of the report of the
  results.




Dunn & Martin                Informational                      [Page 7]

RFC 3116            Methodology for ATM Benchmarking           June 2001


2.7. Frame formats

  The formats of the test IP PDUs to use for TCP/IP and UPC/IP over ATM
  are shown in Appendix C: Test Frame Formats.  Note that these IP PDUs
  are in accordance with RFC 2225.  These exact IP PDU formats SHOULD
  be used in the tests described in this document for this
  protocol/media combination.  These IP PDUs will be used as a template
  for testing other protocol/media combinations.  The specific formats
  that are used to define the test IP PDUs for a particular test series
  MUST be included in the report of the results.

2.8. Frame sizes

  All of the described tests SHOULD be performed using a number of IP
  PDU sizes.  Specifically, the sizes SHOULD include the maximum and
  minimum legitimate sizes for the protocol under test on the media
  under test and enough sizes in between to be able to get a full
  characterization of the SUT performance.  Except where noted, at
  least five IP PDU sizes SHOULD be tested for each test condition.

  Theoretically the minimum size UDP Echo request IP PDU would consist
  of an IP header (minimum length 20 octets), a UDP header (8 octets),
  AAL5 trailer (8 octets) and an LLC/SNAP code point header (8 octets);
  therefore, the minimum size PDU will fit into one ATM cell.  The
  theoretical maximum IP PDU size is determined by the size of the
  length field in the IP header.  In almost all cases the actual
  maximum and minimum sizes are determined by the limitations of the
  media.  In the case of ATM, the maximum IP PDU size SHOULD be the ATM
  MTU size, which is 9180 octets.

  In theory it would be ideal to distribute the IP PDU sizes in a way
  that would evenly distribute the theoretical IP PDU rates.  These
  recommendations incorporate this theory but specify IP PDU sizes,
  which are easy to understand and remember.  In addition, many of the
  same IP PDU sizes are specified on each of the media types to allow
  for easy performance comparisons.

  Note: The inclusion of an unrealistically small IP PDU size on some
  of the media types (i.e., with little or no space for data) is to
  help characterize the per-IP PDU processing overhead of the SUT.

  The IP PDU sizes that will be used are:

  44, 64, 128, 256, 1024, 1518, 2048, 4472, 9180

  The minimum size IP PDU for UDP on ATM is 44 octets, the minimum size
  of 44 is recommended to allow direct comparison to token ring
  performance.  The IP PDU size of 4472 is recommended instead of the



Dunn & Martin                Informational                      [Page 8]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  theoretical FDDI maximum size of 4500 octets in order to permit the
  same type of comparison.  An IP (i.e., not UDP) IP PDU may be used in
  addition if a higher data rate is desired, in which case the minimum
  IP PDU size is 28 octets.

2.9. Verifying received IP PDUs

  The test equipment SHOULD discard any IP PDUs received during a test
  run that are not actual forwarded test IP PDUs.  For example, keep-
  alive and routing update IP PDUs SHOULD NOT be included in the count
  of received IP PDUs.  In any case, the test equipment SHOULD verify
  the length of the received IP PDUs and check that they match the
  expected length.

  Preferably, the test equipment SHOULD include sequence numbers in the
  transmitted IP PDUs and check for these numbers on the received IP
  PDUs.  If this is done, the reported results SHOULD include, in
  addition to the number of IP PDUs dropped, the number of IP PDUs that
  were received out of order, the number of duplicate IP PDUs received
  and the number of gaps in the received IP PDU numbering sequence.
  This functionality is required for some of the described tests.

2.10. Modifiers

  It is useful to characterize the SUTs performance under a number of
  conditions.  Some of these conditions are noted below.  The reported
  results SHOULD include as many of these conditions as the test
  equipment is able to generate.  The suite of tests SHOULD be run
  first without any modifying conditions, then repeated under each of
  the modifying conditions separately.  To preserve the ability to
  compare the results of these tests, any IP PDUs that are required to
  generate the modifying conditions (excluding management queries) will
  be included in the same data stream as that of the normal test IP
  PDUs and in place of one of the test IP PDUs.  They MUST not be
  supplied to the SUT on a separate network port.

2.10.1. Management IP PDUs

  Most ATM data networks now make use of ILMI, signaling and OAM.  In
  many environments, there can be a number of management stations
  sending queries to the same SUT at the same time.

  Management queries MUST be made in accordance with the applicable
  specification, e.g., ILMI sysUpTime getNext requests will be made in
  accordance with ILMI 4.0.  The response to the query MUST be verified
  by the test equipment.  Note that, for each management protocol in





Dunn & Martin                Informational                      [Page 9]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  use, this requires that the test equipment implement the associated
  protocol state machine.  One example of the specific query IP PDU
  (ICMP) that should be used is shown in Appendix C.

2.10.2. Routing update IP PDUs

  The processing of PNNI updates could have a significant impact on the
  ability of a switch to forward cells and complete calls.  If PNNI is
  configured on the SUT, one routing update MUST be transmitted before
  the first test IP PDU is transmitted during the trial.  The test
  SHOULD verify that the SUT has properly processed the routing update.

  PNNI routing update IP PDUs SHOULD be sent at the rate specified in
  Appendix B.  Appendix C defines one routing update PDU for the TCP/IP
  over ATM example.  The routing updates are designed to change the
  routing on a number of networks that are not involved in the
  forwarding of the test data.  The first IP PDU sets the routing table
  state to "A", the second one changes the state to "B".  The IP PDUs
  MUST be alternated during the trial.  The test SHOULD verify that the
  SUT has properly processed the routing update.

2.11. Filters

  Filters are added to switches to selectively inhibit the forwarding
  of cells that would normally be forwarded.  This is usually done to
  implement security controls on the data that is accepted between one
  area and another.  Different products have different capabilities to
  implement filters.  Filters are applicable only if the SUT supports
  the filtering feature.

  The SUT SHOULD be first configured to add one filter condition and
  the tests performed.  This filter SHOULD permit the forwarding of the
  test data stream.  This filter SHOULD be of the form as described in
  the SUT Users Guide.

  The SUT SHOULD be then reconfigured to implement a total of 25
  filters.  The first 24 of these filters SHOULD be based on 24
  separate ATM NSAP Network Prefix addresses.

  The 24 ATM NSAP Network Prefix addresses SHOULD not be any that are
  represented in the test data stream.  The last filter SHOULD permit
  the forwarding of the test data stream.  By "first" and "last" we
  mean to ensure that in the second case, 25 conditions must be checked
  before the data IP over ATM PDUs will match the conditions that
  permit the forwarding of the IP PDU.  Of course, if the SUT reorders
  the filters or does not use a linear scan of the filter rules the
  effect of the sequence in which the filters are input is properly
  lost.



Dunn & Martin                Informational                     [Page 10]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  The exact filters configuration command lines used SHOULD be included
  with the report of the results.

2.11.1. Filter Addresses

  Two sets of filter addresses are required, one for the single filter
  case and one for the 25 filter case.

  The single filter case should permit traffic from ATM address [Switch
  Network Prefix] 00 00 00 00 00 01 00 to ATM address [Switch Network
  Prefix] 00 00 00 00 00 02 00 and deny all other traffic.  Note that
  the 13 octet Switch Network Prefix MUST be configured before this
  test can be run.

  The 25 filter case should follow the following sequence.

        deny [Switch Network Prefix] 00 00 00 00 00 01 00
             to [Switch Network Prefix] 00 00 00 00 00 03 00
        deny [Switch Network Prefix] 00 00 00 00 00 01 00
             to [Switch Network Prefix] 00 00 00 00 00 04 00
        deny [Switch Network Prefix] 00 00 00 00 00 01 00
             to [Switch Network Prefix] 00 00 00 00 00 05 00
        ...
        deny [Switch Network Prefix] 00 00 00 00 00 01 00
             to [Switch Network Prefix] 00 00 00 00 00 0C 00
        deny [Switch Network Prefix] 00 00 00 00 00 01 00
             to [Switch Network Prefix] 00 00 00 00 00 0D 00
        allow [Switch Network Prefix] 00 00 00 00 00 01 00
             to [Switch Network Prefix] 00 00 00 00 00 02 00
        deny [Switch Network Prefix] 00 00 00 00 00 01 00
             to [Switch Network Prefix] 00 00 00 00 00 0E 00
        deny [Switch Network Prefix] 00 00 00 00 00 01 00
             to [Switch Network Prefix] 00 00 00 00 00 0F 00
         ...
        deny [Switch Network Prefix] 00 00 00 00 00 01 00
             to [Switch Network Prefix] 00 00 00 00 00 18 00
        deny all else

  All previous filter conditions should be cleared from the switch
  before this sequence is entered.  The sequence is selected to test to
  see if the switch sorts the filter conditions or accepts them in the
  order that they were entered.  Both of these procedures will result
  in a greater impact on performance than will some form of hash
  coding.







Dunn & Martin                Informational                     [Page 11]

RFC 3116            Methodology for ATM Benchmarking           June 2001


2.12. Protocol addresses

  It is easier to implement these tests using a single logical stream
  of data, with one source ATM address and one destination ATM address,
  and for some conditions like the filters described above, a practical
  requirement.  Networks in the real world are not limited to single
  streams of data.  The test suite SHOULD be first run with a single
  ATM source and destination address pair.  The tests SHOULD then be
  repeated with using a random destination address.  In the case of
  testing single switches, the addresses SHOULD be random and uniformly
  distributed over a range of 256 seven octet user parts.  In the case
  of testing multiple interconnected switches, the addresses SHOULD be
  random and uniformly distributed over the 256 network prefixes, each
  of which should support 256 seven octet user parts.  The specific
  address ranges to use for ATM are shown in Appendix A.  IP to ATM
  address mapping MUST be accomplished as described in RFC 2225.

2.13. Route Set Up

  It is not reasonable that all of the routing information necessary to
  forward the test stream, especially in the multiple address case,
  will be manually set up.  If PNNI and/or ILMI are running, at the
  start of each trial a routing update MUST be sent to the SUT.  This
  routing update MUST include all of the ATM addresses that will be
  required for the trial.  This routing update will have to be repeated
  at the interval required by PNNI or ILMI.  An example of the format
  and repetition interval of the update IP PDUs is given in Appendix B
  (interval and size) and Appendix C (format).

2.14. Bidirectional traffic

  Bidirectional performance tests SHOULD be run with the same data rate
  being offered from each direction.  The sum of the data rates should
  not exceed the theoretical limit for the media.

2.15. Single stream path

  The full suite of tests SHOULD be run with the appropriate modifiers
  for a single receive and transmit port on the SUT.  If the internal
  design of the SUT has multiple distinct pathways, for example,
  multiple interface cards each with multiple network ports, then all
  possible permutations of pathways SHOULD be tested separately.  If
  multiple interconnected switches are tested, the test MUST specify
  routes, which allow only one path between source and destination ATM
  addresses.






Dunn & Martin                Informational                     [Page 12]

RFC 3116            Methodology for ATM Benchmarking           June 2001


2.16. Multi-port

  Many switch products provide several network ports on the same
  interface module.  Each port on an interface module SHOULD be
  stimulated in an identical manner.  Specifically, half of the ports
  on each module SHOULD be receive ports and half SHOULD be transmit
  ports.  For example if a SUT has two interface module each of which
  has four ports, two ports on each interface module be receive ports
  and two will be transmit ports.  Each receive port MUST be offered
  the same data rate.  The addresses in the input data streams SHOULD
  be set so that an IP PDU will be directed to each of the transmit
  ports in sequence.  That is, all transmit ports will receive an
  identical distribution of IP PDUs from a particular receive port.

  Consider the following 6 port SUT:

              --------------
     ---------| Rx A   Tx X|--------
     ---------| Rx B   Tx Y|--------
     ---------| Rx C   Tx Z|--------
              --------------

  The addressing of the data streams for each of the inputs SHOULD be:

     stream sent to Rx A:
       IP PDU to Tx X, IP PDU to Tx Y, IP PDU to Tx Z
     stream sent to Rx B:
       IP PDU to Tx X, IP PDU to Tx Y, IP PDU to Tx Z
     stream sent to Rx C
       IP PDU to Tx X, IP PDU to Tx Y, IP PDU to Tx Z

  Note: Each stream contains the same sequence of IP destination
  addresses; therefore, each transmit port will receive 3 IP PDUs
  simultaneously.  This procedure ensures that the SUT will have to
  process multiple IP PDUs addressed to the same transmit port
  simultaneously.

  The same configuration MAY be used to perform a bi-directional
  multi-stream test.  In this case all of the ports are considered both
  receive and transmit ports.  Each data stream MUST consist of IP PDUs
  whose addresses correspond to the ATM addresses all of the other
  ports.









Dunn & Martin                Informational                     [Page 13]

RFC 3116            Methodology for ATM Benchmarking           June 2001


2.17. Multiple protocols

  This document does not address the issue of testing the effects of a
  mixed protocol environment other than to suggest that if such tests
  are wanted then PDUs SHOULD be distributed between all of the test
  protocols.  The distribution MAY approximate the conditions on the
  network in which the SUT would be used.

2.18. Multiple IP PDU sizes

  This document does not address the issue of testing the effects of a
  mixed IP PDU size environment other than to suggest that, if such
  tests are required, then IP PDU size SHOULD be evenly distributed
  among all of the PDU sizes listed in this document.  The distribution
  MAY approximate the conditions on the network in which the SUT would
  be used.

2.19. Testing beyond a single SUT

  In the performance testing of a single SUT, the paradigm can be
  described as applying some input to a SUT and monitoring the output.
  The results of which can be used to form a basis of characterization
  of that device under those test conditions.

  This model is useful when the test input and output are homogeneous
  (e.g., 64-byte IP, AAL5 PDUs into the SUT; 64 byte IP, AAL5 PDUs
  out).

  By extending the single SUT test model, reasonable benchmarks
  regarding multiple SUTs or heterogeneous environments may be
  collected.  In this extension, the single SUT is replaced by a system
  of interconnected network SUTs.  This test methodology would support
  the benchmarking of a variety of device/media/service/protocol
  combinations.  For example, a configuration for a LAN-to-WAN-to-LAN
  test might be:

     (1) ATM UNI -> SUT 1 -> BISUP -> SUT 2 -> ATM UNI

  Or an extended LAN configuration might be:

     (2) ATM UNI -> SUT 1 -> PNNI Network -> SUT 2 -> ATM UNI

  In both examples 1 and 2, end-to-end benchmarks of each system could
  be empirically ascertained.  Other behavior may be characterized
  through the use of intermediate devices.  In example 2, the
  configuration may be used to give an indication of the effect of PNNI
  routing on IP throughput.




Dunn & Martin                Informational                     [Page 14]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  Because multiple SUTs are treated as a single system, there are
  limitations to this methodology.  For instance, this methodology may
  yield an aggregate benchmark for a tested system.  That benchmark
  alone, however, may not necessarily reflect asymmetries in behavior
  between the SUTs, latencies introduced by other apparatus (e.g.,
  CSUs/DSUs, switches), etc.

  Further, care must be used when comparing benchmarks of different
  systems by ensuring that the SUTs' features and configuration of the
  tested systems have the appropriate common denominators to allow
  comparison.

2.20. Maximum IP PDU rate

  The maximum IP PDU rates that should be used when testing LAN
  connections SHOULD be the listed theoretical maximum rate for the IP
  PDU size on the media.

  The maximum IP PDU rate that should be used when testing WAN
  connections SHOULD be greater than the listed theoretical maximum
  rate for the IP PDU size on that speed connection.  The higher rate
  for WAN tests is to compensate for the fact that some vendors employ
  various forms of header compression.

  A list of maximum IP PDU rates for LAN connections is included in
  Appendix B.

2.21. Bursty traffic

  It is convenient to measure the SUT performance under steady state
  load; however, this is an unrealistic way to gauge the functioning of
  a SUT.  Actual network traffic normally consists of bursts of IP
  PDUs.

  Some of the tests described below SHOULD be performed with both
  constant bit rate, bursty Unspecified Bit Rate (UBR) Best Effort
  [AF-TM4.1] and Variable Bit Rate Non-real Time (VBR-nrt) Best Effort
  [AF-TM4.1].  The IP PDUs within a burst are transmitted with the
  minimum legitimate inter-IP PDU gap.

  The objective of the test is to determine the minimum interval
  between bursts that the SUT can process with no IP PDU loss.  Tests
  SHOULD be run with burst sizes of 10% of Maximum Burst Size (MBS),
  20% of MBS, 50% of MBS and 100% MBS.  Note that the number of IP PDUs
  in each burst will depend on the PDU size.  For UBR, the MBS refers
  to the associated VBR traffic parameters.





Dunn & Martin                Informational                     [Page 15]

RFC 3116            Methodology for ATM Benchmarking           June 2001


2.22. Trial description

  A particular test consists of multiple trials.  Each trial returns
  one piece of information, for example the loss rate at a particular
  input IP PDU rate.  Each trial consists of five of phases:

  a) If the SUT is a switch supporting PNNI, send the routing update to
     the SUT receive port and wait two seconds to be sure that the
     routing has settled.

  b) Send an ATM ARP PDU to determine the ATM address corresponding to
     the destination IP address.  The formats of the ATM ARP PDU that
     should be used are shown in the Test Frame Formats document and
     MUST be in accordance with RFC 2225.

  c) Stimulate SUT with traffic load.

  d) Wait for two seconds for any residual IP PDUs to be received.

  e) Wait for at least five seconds for the SUT to restabilize.

2.23. Trial duration

  The objective of the tests defined in this document is to accurately
  characterize the behavior of a particular piece of network equipment
  under varying traffic loads.  The choice of test duration must be a
  compromise between this objective and keeping the duration of the
  benchmarking test suite within reasonable bounds.  The SUT SHOULD be
  stimulated for at least 60 seconds.  If this time period results in a
  high variance in the test results, the SUT SHOULD be stimulated for
  at least 300 seconds.

2.24. Address resolution

  The SUT MUST be able to respond to address resolution requests sent
  by another SUT, an ATM ARP server or the test equipment in accordance
  with RFC 2225.

2.25. Synchronized Payload Bit Pattern.

  Some measurements assume that both the transmitter and receiver
  payload information is synchronized.  Synchronization MUST be
  achieved by supplying a known bit pattern to both the transmitter and
  receiver.  This bit pattern MUST be one of the following: PRBS-15,
  PRBS-23, 0xFF00, or 0xAA55.






Dunn & Martin                Informational                     [Page 16]

RFC 3116            Methodology for ATM Benchmarking           June 2001


2.26. Burst Traffic Descriptors.

  Some measurements require busty traffic patterns.  These patterns
  MUST conform to one of the following traffic descriptors:

1) PCR=100% allotted line rate, SCR=50% allotted line rate, and MBS=8192

2) PCR=100% allotted line rate, SCR=50% allotted line rate, and MBS=4096

3) PCR=90% allotted line rate, SCR=50% allotted line rate, and MBS=8192

4) PCR=90% allotted line rate, SCR=50% allotted line rate, and MBS=4096

5) PCR=90% allotted line rate, SCR=45% allotted line rate, and MBS=8192

6) PCR=90% allotted line rate, SCR=45% allotted line rate, and MBS=4096

7) PCR=80% allotted line rate, SCR=40% allotted line rate, and MBS=65536

8) PCR=80% allotted line rate, SCR=40% allotted line rate, and MBS=32768

  The allotted line rate refers to the total available line rate
  divided by the number of VCCs in use.

3. Performance Metrics

3.1. Physical Layer-SONET

3.1.1. Pointer Movements

3.1.1.1. Pointer Movement Propagation.

  Objective: To determine that the SUT does not propagate pointer
  movements as defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the uni-directional
      configuration.

  2)  Send a specific number of IP PDUs at a specific rate through the
      SUT.  Since this test is not a throughput test, the rate should
      not be greater than 90% of line rate.  The cell payload SHOULD
      contain valid IP PDUs.  The IP PDUs MUST be encapsulated in AAL5.







Dunn & Martin                Informational                     [Page 17]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  3)  Count the IP PDUs that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test, else lower the test device
      traffic rate until the counts are the same.

  4)  Inject one forward payload pointer movement.  Verify that the SUT
      does not change the pointer.

  5)  Inject one forward payload pointer movement every 1 second.
      Verify that the SUT does not change the pointer.

  6)  Discontinue the payload pointer movement.

  7)  Inject five forward payload pointer movements every 1 second.
      Verify that the SUT does not change the pointer.

  8)  Discontinue the payload pointer movement.

  9)  Inject one backward payload pointer movement.  Verify that the
      SUT does not change the pointer.

  10) Inject one backward payload pointer movement every 1 second.
      Verify that the SUT does not change the pointer.

  11) Discontinue the payload pointer movement.

  12) Inject five backward payload pointer movements every 1 second.
      Verify that the SUT does not change the pointer.

  13) Discontinue the payload pointer movement.

  Reporting Format:

     The results of the pointer movement propagation test SHOULD be
     reported in a form of a table.  The rows SHOULD be labeled single
     pointer movement, one pointer movement per second, and five
     pointer movements per second.  The columns SHOULD be labeled
     pointer movement and loss of pointer.  The elements of the table
     SHOULD be either True or False, indicating whether the particular
     condition was observed for each test.

     The table MUST also indicate the IP PDU size in octets and traffic
     rate in IP PDUs per second as generated by the test device.








Dunn & Martin                Informational                     [Page 18]

RFC 3116            Methodology for ATM Benchmarking           June 2001


3.1.1.2. Cell Loss due to Pointer Movement.

  Objective: To determine if the SUT will drop cells due to pointer
  movements as defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the uni-directional
      configuration.

  2)  Send a specific number of cells at a specific rate through the
      SUT.  Since this test is not a throughput test, the rate should
      not be greater than 90% of line rate.  The cell payload SHOULD
      contain valid IP PDUs.  The IP PDUs MUST be encapsulated in AAL5.

  3)  Count the cells that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  4)  Inject one forward payload pointer movement.  Verify that the SUT
      does not drop any cells.

  5)  Inject one forward payload pointer movement every 1 second.
      Verify that the SUT does not drop any cells.

  6)  Discontinue the payload pointer movement.

  7)  Inject five forward payload pointer movements every 1 second.
      Verify that the SUT does not drop any cells.

  8)  Discontinue the payload pointer movement.

  9)  Inject one backward payload pointer movement.  Verify that the
      SUT does not drop any cells.

  10) Inject one backward payload pointer movement every 1 second.
      Verify that the SUT does not drop any cells.

  11) Discontinue the payload pointer movement.

  12) Inject five backward payload pointer movements every 1 second.
      Verify that the SUT does not drop any cells.

  13) Discontinue the payload pointer movement.






Dunn & Martin                Informational                     [Page 19]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  Reporting Format:

     The results of the cell loss due to pointer movement test SHOULD
     be reported in a form of a table.  The rows SHOULD be labeled
     single pointer movement, one pointer movement per second, and five
     pointer movements per second.  The columns SHOULD be labeled cell
     loss and number of cells lost.  The elements of column 1 SHOULD be
     either True or False, indicating whether the particular condition
     was observed for each test.  The elements of column 2 SHOULD be
     non-negative integers.

     The table MUST also indicate the traffic rate in IP PDUs per
     second as generated by the test device.

3.1.1.3. IP Packet Loss due to Pointer Movement.

  Objective: To determine if the SUT will drop IP packets due to
  pointer movements as defined in RFC 2761 "Terminology for ATM
  Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the uni-directional
      configuration.

  2)  Send a specific number of IP packets at a specific rate through
      the SUT.  Since this test is not a throughput test, the rate
      should not be greater than 90% of line rate.  The IP PDUs MUST be
      encapsulated in AAL5.

  3)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  4)  Inject one forward payload pointer movement.  Verify that the SUT
      does not drop any packets.

  5)  Inject one forward payload pointer movement every 1 second.
      Verify that the SUT does not drop any packets.

  6)  Discontinue the payload pointer movement.

  7)  Inject five forward payload pointer movements every 1 second.
      Verify that the SUT does not drop any packets.

  8)  Discontinue the payload pointer movement.




Dunn & Martin                Informational                     [Page 20]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  9)  Inject one backward payload pointer movement.  Verify that the
      SUT does not drop any packets.

  10) Inject one backward payload pointer movement every 1 second.
      Verify that the SUT does not drop any packets.

  11) Discontinue the payload pointer movement.

  12) Inject five backward payload pointer movements every 1 second.
      Verify that the SUT does not drop any packets.

  13) Discontinue the payload pointer movement.

  Reporting Format:

     The results of the IP packet loss due to pointer movement test
     SHOULD be reported in a form of a table.  The rows SHOULD be
     labeled single pointer movement, one pointer movement per second,
     and five pointer movements per second.  The columns SHOULD be
     labeled packet loss and number of packets lost.  The elements of
     column 1 SHOULD be either True or False, indicating whether the
     particular condition was observed for each test.  The elements of
     column 2 SHOULD be non-negative integers.

     The table MUST also indicate the packet size in octets and traffic
     rate in packets per second as generated by the test device.

3.1.2. Transport Overhead (TOH) Error Count

3.1.2.1. TOH Error Propagation.

  Objective: To determine that the SUT does not propagate TOH errors as
  defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the uni-directional
      configuration.

  2)  Send a specific number of IP PDUs at a specific rate through the
      SUT.  Since this test is not a throughput test, the rate should
      not be greater than 90% of line rate.  The cell payload SHOULD
      contain valid IP PDUs.  The IP PDUs MUST be encapsulated in AAL5.

  3)  Count the IP PDUs that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test, else lower the test device
      traffic rate until the counts are the same.



Dunn & Martin                Informational                     [Page 21]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  4)  Inject one error in the first bit of the A1 and A2 Frameword.
      Verify that the SUT does not propagate the error.

  5)  Inject one error in the first bit of the A1 and A2 Frameword
      every 1 second.  Verify that the SUT does not propagate the
      error.

  6)  Discontinue the Frameword error.

  7)  Inject one error in the first bit of the A1 and A2 Frameword for
      4 consecutive IP PDUs in every 6 IP PDUs.  Verify that the SUT
      indicates Loss of Frame.

  8)  Discontinue the Frameword error.

  Reporting Format:

     The results of the TOH error propagation test SHOULD be reported
     in a form of a table.  The rows SHOULD be labeled single error,
     one error per second, and four consecutive errors every 6 IP PDUs.
     The columns SHOULD be labeled error propagated and loss of IP PDU.
     The elements of the table SHOULD be either True or False,
     indicating whether the particular condition was observed for each
     test.

     The table MUST also indicate the IP PDU size in octets and traffic
     rate in IP PDUs per second as generated by the test device.

3.1.2.2. c TOH Error.

  Objective: To determine if the SUT will drop cells due TOH Errors as
  defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the uni-directional
      configuration.

  2)  Send a specific number of cells at a specific rate through the
      SUT.  Since this test is not a throughput test, the rate should
      not be greater than 90% of line rate.  The cell payload SHOULD
      contain valid IP PDUs.  The IP PDUs MUST be encapsulated in AAL5.

  3)  Count the cells that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.




Dunn & Martin                Informational                     [Page 22]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  4)  Inject one error in the first bit of the A1 and A2 Frameword.
      Verify that the SUT does not drop any cells.

  5)  Inject one error in the first bit of the A1 and A2 Frameword
      every 1 second.  Verify that the SUT does not drop any cells.

  6)  Discontinue the Frameword error.

  7)  Inject one error in the first bit of the A1 and A2 Frameword for
      4 consecutive IP PDUs in every 6 IP PDUs.  Verify that the SUT
      does drop cells.

  8)  Discontinue the Frameword error.

  Reporting Format:

     The results of the Cell Loss due to TOH errors test SHOULD be
     reported in a form of a table.  The rows SHOULD be labeled single
     error, one error per second, and four consecutive errors every 6
     IP PDUs.  The columns SHOULD be labeled cell loss and number of
     cells lost.  The elements of column 1 SHOULD be either True or
     False, indicating whether the particular condition was observed
     for each test.  The elements of column 2 SHOULD be non-negative
     integers.

     The table MUST also indicate the traffic rate in IP PDUs per
     second as generated by the test device.

3.1.2.3. IP Packet Loss due to TOH Error.

  Objective: To determine if the SUT will drop IP packets due to TOH
  errors as defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the uni-directional
      configuration.

  2)  Send a specific number of IP packets at a specific rate through
      the SUT.  Since this test is not a throughput test, the rate
      should not be greater than 90% of line rate.  The IP PDUs MUST be
      encapsulated in AAL5.

  3)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.




Dunn & Martin                Informational                     [Page 23]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  4)  Inject one error in the first bit of the A1 and A2 Frameword.
      Verify that the SUT does not drop any packets.

  5)  Inject one error in the first bit of the A1 and A2 Frameword
      every 1 second.  Verify that the SUT does not drop any packets.

  6)  Discontinue the Frameword error.

  7)  Inject one error in the first bit of the A1 and A2 Frameword for
      4 consecutive IP PDUs in every 6 IP PDUs.  Verify that the SUT
      does drop packets.

  8)  Discontinue the Frameword error.

  Reporting Format:

     The results of the IP packet loss due to TOH errors test SHOULD be
     reported in a form of a table.  The rows SHOULD be labeled single
     error, one error per second, and four consecutive errors every 6
     IP PDUs.  The columns SHOULD be labeled packet loss and number of
     packets lost.  The elements of column 1 SHOULD be either True or
     False, indicating whether the particular condition was observed
     for each test.  The elements of column 2 SHOULD be non-negative
     integers.

     The table MUST also indicate the packet size in octets and traffic
     rate in packets per second as generated by the test device.

3.1.3. Path Overhead (POH) Error Count

3.1.3.1. POH Error Propagation.

  Objective: To determine that the SUT does not propagate POH errors as
  defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the uni-directional
      configuration.

  2)  Send a specific number of IP PDUs at a specific rate through the
      SUT.  Since this test is not a throughput test, the rate should
      not be greater than 90% of line rate.  The cell payload SHOULD
      contain valid IP PDUs.  The IP PDUs MUST be encapsulated in AAL5.







Dunn & Martin                Informational                     [Page 24]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  3)  Count the IP PDUs that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test, else lower the test device
      traffic rate until the counts are the same.

  4)  Inject one error in the B3 (Path BIP8) byte.  Verify that the SUT
      does not propagate the error.

  5)  Inject one error in the B3 byte every 1 second.  Verify that the
      SUT does not propagate the error.

  6)  Discontinue the POH error.

  Reporting Format:

      The results of the POH error propagation test SHOULD be reported
      in a form of a table.  The rows SHOULD be labeled single error
      and one error per second.  The columns SHOULD be labeled error
      propagated and loss of IP PDU.  The elements of the table SHOULD
      be either True or False, indicating whether the particular
      condition was observed for each test.

      The table MUST also indicate the IP PDU size in octets and
      traffic rate in IP PDUs per second as generated by the test
      device.

3.1.3.2. Cell Loss due to POH Error.

  Objective: To determine if the SUT will drop cells due POH Errors as
  defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the uni-directional
      configuration.

  2)  Send a specific number of cells at a specific rate through the
      SUT.  Since this test is not a throughput test, the rate should
      not be greater than 90% of line rate.  The cell payload SHOULD
      contain valid IP PDUs.  The IP PDUs MUST be encapsulated in AAL5.

  3)  Count the cells that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  4)  Inject one error in the B3 (Path BIP8) byte.  Verify that the SUT
      does not drop any cells.



Dunn & Martin                Informational                     [Page 25]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  5)  Inject one error in the B3 byte every 1 second.  Verify that the
      SUT does not drop any cells.

  6)  Discontinue the POH error.

  Reporting Format:

     The results of the Cell Loss due to POH errors test SHOULD be
     reported in a form of a table.  The rows SHOULD be labeled single
     error and one error per second.  The columns SHOULD be labeled
     cell loss and number of cells lost.  The elements of column 1
     SHOULD be either True or False, indicating whether the particular
     condition was observed for each test.  The elements of column 2
     SHOULD be non-negative integers.

     The table MUST also indicate the traffic rate in IP PDUs per
     second as generated by the test device.

3.1.3.3. IP Packet Loss due to POH Error.

  Objective: To determine if the SUT will drop IP packets due to POH
  errors as defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the uni-directional
      configuration.

  2)  Send a specific number of IP packets at a specific rate through
      the SUT.  Since this test is not a throughput test, the rate
      should not be greater than 90% of line rate.  The IP PDUs MUST be
      encapsulated in AAL5.

  3)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  4)  Inject one error in the B3 (Path BIP8) byte.  Verify that the SUT
      does not drop any packets.

  5)  Inject one error in the B3 byte every 1 second.  Verify that the
      SUT does not drop any packets.

  6)  Discontinue the POH error.






Dunn & Martin                Informational                     [Page 26]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  Reporting Format:

     The results of the IP packet loss due to POH errors test SHOULD be
     reported in a form of a table.  The rows SHOULD be labeled single
     error and one error per second.  The columns SHOULD be labeled
     packet loss and number of packets lost.  The elements of column 1
     SHOULD be either True or False, indicating whether the particular
     condition was observed for each test.  The elements of column 2
     SHOULD be non-negative integers.

     The table MUST also indicate the packet size in octets and traffic
     rate in packets per second as generated by the test device.

3.2. ATM Layer

3.2.1. Two-Point Cell Delay Variation (CDV)

3.2.1.1. Test Setup

  The cell delay measurements assume that both the transmitter and
  receiver timestamp information is synchronized.  Synchronization
  SHOULD be achieved by supplying a common clock signal (minimum of 100
  Mhz or 10 ns resolution) to both the transmitter and receiver.  The
  maximum timestamp values MUST be recorded to ensure synchronization
  in the case of counter rollover.  The cell delay measurements SHOULD
  utilize the O.191 cell (ITUT-O.191) encapsulated in a valid IP
  packet.  If the O.191 cell is not available, a test cell encapsulated
  in a valid IP packet MAY be used.  The test cell MUST contain a
  transmit timestamp which can be correlated with a receive timestamp.
  A description of the test cell MUST be included in the test results.
  The description MUST include the timestamp length (in bits), counter
  rollover value, and the timestamp accuracy (in ns).

3.2.1.2. Two-point CDV/Steady Load/One VCC

  Objective: To determine the SUT variation in cell transfer delay with
  one VCC as defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with one VCC.  The VCC SHOULD
      contain one VPI/VCI.  The VCC MUST be configured as either a CBR,
      VBR, or UBR connection.  The VPI/VCI MUST not be one of the
      reserved ATM signaling channels (e.g., [0,5], [0,16]).




Dunn & Martin                Informational                     [Page 27]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  3)  Send a specific number of IP packets containing timestamps at a
      specific constant rate through the SUT via the defined test VCC.
      Since this test is not a throughput test, the rate should not be
      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device.

  Reporting Format:

     The results of the Two-point CDV/Steady Load/One VCC test SHOULD
     be reported in a form of text, graph, and histogram.

     The text results SHOULD display the numerical values of the CDV.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, maximum
     and minimum CDV during the test in us, and peak-to-peak CDV in us.

     The graph results SHOULD display the cell delay values.  The x-
     coordinate SHOULD be the test run time in either seconds, minutes
     or days depending on the total length of the test.  The x-
     coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the cell delay in us.  The integration time per point MUST be
     indicated.

     The histogram results SHOULD display the peak-to-peak cell delay.
     The x-coordinate SHOULD be the cell delay in us with at least 256
     bins.  The y-coordinate SHOULD be the number of cells observed in
     each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     bearer class of the created VCC MUST also be indicated.

3.2.1.3. Two-point CDV/Steady Load/Twelve VCCs

  Objective: To determine the SUT variation in cell transfer delay with
  twelve VCCs as defined in RFC 2761 "Terminology for ATM
  Benchmarking".




Dunn & Martin                Informational                     [Page 28]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCCs, using 1 VPI
      and 12 VCIs.  The VCC's MUST be configured as either a CBR, VBR,
      or UBR connection.  The VPI/VCIs MUST not be one of the reserved
      ATM signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing timestamps at a
      specific constant rate through the SUT via the defined test VCCs.
      All of the VPI/VCI pairs will generate traffic at the same
      traffic rate.  Since this test is not a throughput test, the rate
      should not be greater than 90% of line rate.  The IP PDUs MUST be
      encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device for all VCCs.

  Reporting Format:

     The results of the Two-point CDV/Steady Load/Twelve VCCs test
     SHOULD be reported in a form of text, graph, and histograms.

     The text results SHOULD display the numerical values of the CDV.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI values, total number of cells transmitted and received on
     each VCC during the test in positive integers, maximum and minimum
     CDV on each VCC during the test in us, and peak-to-peak CDV on
     each VCC in us.

     The graph results SHOULD display the cell delay values.  The x-
     coordinate SHOULD be the test run time in either seconds, minutes
     or days depending on the total length of the test.  The x-
     coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the cell delay for each VCC in ms.  There SHOULD be 12 curves
     on the graph, one curves indicated and labeled for each VCC.  The
     integration time per point MUST be indicated.







Dunn & Martin                Informational                     [Page 29]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The histograms SHOULD display the peak-to-peak cell delay.  There
     will be one histogram for each VCC.  The x-coordinate SHOULD be
     the cell delay in us with at least 256 bins.  The y-coordinate
     SHOULD be the number of cells observed in each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     bearer class of the created VCC MUST also be indicated.

3.2.1.4. Two-point CDV/Steady Load/Maximum VCCs

  Objective: To determine the SUT variation in cell transfer delay with
  the maximum number VCCs supported on the SUT as defined in RFC 2761
  "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with the maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per
      VPI.  The VCC's MUST be configured as either a CBR, VBR, or UBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing timestamps at a
      specific constant rate through the SUT via the defined test VCCs.
      All of the VPI/VCI pairs will generate traffic at the same
      traffic rate.  Since this test is not a throughput test, the rate
      should not be greater than 90% of line rate.  The IP PDUs MUST be
      encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device for all VCCs.

  Reporting Format:

     The results of the Two-point CDV/Steady Load/Maximum VCCs test
     SHOULD be reported in a form of text, graphs, and histograms.




Dunn & Martin                Informational                     [Page 30]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The text results SHOULD display the numerical values of the CDV.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI values, total number of cells transmitted and received on
     each VCC during the test in positive integers, maximum and minimum
     CDV on each VCC during the test in us, and peak-to-peak CDV on
     each VCC in us.

     The graph results SHOULD display the cell delay values.  There
     will be (Max number of VCCs/10) graphs, with 10 VCCs indicated on
     each graph.  The x-coordinate SHOULD be the test run time in
     either seconds, minutes or days depending on the total length of
     the test.  The x-coordinate time SHOULD be configurable.  The y-
     coordinate SHOULD be the cell delay for each VCC in us.  There
     SHOULD be no more than 10 curves on each graph, one curve
     indicated and labeled for each VCC.  The integration time per
     point MUST be indicated.

     The histograms SHOULD display the peak-to-peak cell delay.  There
     will be one histogram for each VCC.  The x-coordinate SHOULD be
     the cell delay in us with at least 256 bins.  The y-coordinate
     SHOULD be the number of cells observed in each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     bearer class of the created VCC MUST also be indicated.

3.2.1.5. Two-point CDV/Bursty VBR Load/One VCC

  Objective: To determine the SUT variation in cell transfer delay with
  one VCC as defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with one VCC.  The VCC SHOULD
      contain one VPI/VCI.  The VCC MUST be configured as either a CBR
      or VBR connection.  The VPI/VCI MUST not be one of the reserved
      ATM signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing timestamps at a
      specific VBR through the SUT via the defined test VCC.  Since
      this test is not a throughput test, the rate should not be
      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.




Dunn & Martin                Informational                     [Page 31]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device.

  Reporting Format:

     The results of the Two-point CDV/Bursty VBR Load/One VCC test
     SHOULD be reported in a form of text, graph, and histogram.

     The text results SHOULD display the numerical values of the CDV.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, maximum
     and minimum CDV during the test in us, and peak-to-peak CDV in us.

     The graph results SHOULD display the cell delay values.  The x-
     coordinate SHOULD be the test run time in either seconds, minutes
     or days depending on the total length of the test.  The x-
     coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the cell delay in us.  The integration time per point MUST be
     indicated.

     The histogram results SHOULD display the peak-to-peak cell delay.
     The x-coordinate SHOULD be the cell delay in us with at least 256
     bins.  The y-coordinate SHOULD be the number of cells observed in
     each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.1.6. Two-point CDV/Bursty VBR Load/Twelve VCCs

  Objective: To determine the SUT variation in cell transfer delay with
  twelve VCCs as defined in RFC 2761 "Terminology for ATM
  Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.




Dunn & Martin                Informational                     [Page 32]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  2)  Configure the SUT and test device with twelve VCCs, using 1 VPI
      and 12 VCIs.  The VCC's MUST be configured as either a CBR or VBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing timestamps at a
      specific VBR through the SUT via the defined test VCCs.  All of
      the VPI/VCI pairs will generate traffic at the same traffic rate.
      Since this test is not a throughput test, the rate should not be
      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device for all VCCs.

  Reporting Format:

     The results of the Two-point CDV/Bursty VBR Load/Twelve VCCs test
     SHOULD be reported in a form of text, graph, and histograms.

     The text results SHOULD display the numerical values of the CDV.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI values, total number of cells transmitted and received on
     each VCC during the test in positive integers, maximum and minimum
     CDV on each VCC during the test in us, and peak-to-peak CDV on
     each VCC in us.

     The graph results SHOULD display the cell delay values.  The x-
     coordinate SHOULD be the test run time in either seconds, minutes
     or days depending on the total length of the test.  The x-
     coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the cell delay for each VCC in ms.  There SHOULD be 12 curves
     on the graph, one curves indicated and labeled for each VCC.  The
     integration time per point MUST be indicated.

     The histograms SHOULD display the peak-to-peak cell delay.  There
     will be one histogram for each VCC.  The x-coordinate SHOULD be
     the cell delay in us with at least 256 bins.  The y-coordinate
     SHOULD be the number of cells observed in each bin.







Dunn & Martin                Informational                     [Page 33]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.1.7. Two-point CDV/Bursty VBR Load/Maximum VCCs

  Objective: To determine the SUT variation in cell transfer delay with
  the maximum number VCCs supported on the SUT as defined in RFC 2761
  "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with the maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per
      VPI.  The VCC's MUST be configured as either a CBR or VBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing timestamps at a
      specific VBR through the SUT via the defined test VCCs.  All of
      the VPI/VCI pairs will generate traffic at the same traffic rate.
      Since this test is not a throughput test, the rate should not be
      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device for all VCCs.

  Reporting Format:

     The results of the Two-point CDV/Bursty VBR Load/Maximum VCCs test
     SHOULD be reported in a form of text, graphs, and histograms.

     The text results SHOULD display the numerical values of the CDV.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI values, total number of cells transmitted and received on




Dunn & Martin                Informational                     [Page 34]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     each VCC during the test in positive integers, maximum and minimum
     CDV on each VCC during the test in us, and peak-to-peak CDV on
     each VCC in us.

     The graph results SHOULD display the cell delay values.  There
     will be (Max number of VCCs/10) graphs, with 10 VCCs indicated on
     each graph.  The x-coordinate SHOULD be the test run time in
     either seconds, minutes or days depending on the total length of
     the test.  The x-coordinate time SHOULD be configurable.  The y-
     coordinate SHOULD be the cell delay for each VCC in us.  There
     SHOULD be no more than 10 curves on each graph, one curve
     indicated and labeled for each VCC.  The integration time per
     point MUST be indicated.

     The histograms SHOULD display the peak-to-peak cell delay.  There
     will be one histogram for each VCC.  The x-coordinate SHOULD be
     the cell delay in us with at least 256 bins.  The y-coordinate
     SHOULD be the number of cells observed in each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.1.8. Two-point CDV/Mixed Load/Three VCC's

  Objective: To determine the SUT variation in cell transfer delay with
  three VCC's as defined in RFC 2761 "Terminology for ATM
  Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with three VCC's.  Each VCC
      MUST be defined as a different Bearer class: one CBR, one UBR and
      one VBR.  Each VCC SHOULD contain one VPI/VCI.  The VPI/VCI MUST
      not be one of the reserved ATM signaling channels (e.g., [0,5],
      [0,16]).

  3)  Send a specific number of IP packets containing timestamps
      through the SUT via the defined test VCCs.  Each generated VCC
      stream MUST match the corresponding VCC Bearer class.  All of the
      VPI/VCI pairs will generate traffic at the same traffic rate.





Dunn & Martin                Informational                     [Page 35]

RFC 3116            Methodology for ATM Benchmarking           June 2001


      Since this test is not a throughput test, the rate should not be
      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device for all VCC's.

  Reporting Format:

     The results of the Two-point CDV/Mixed Load/Three VCC test SHOULD
     be reported in a form of text, graph, and histogram.

     The text results SHOULD display the numerical values of the CDV.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, maximum
     and minimum CDV during the test in us, and peak-to-peak CDV in us.

     The graph results SHOULD display the cell delay values.  The x-
     coordinate SHOULD be the test run time in either seconds, minutes
     or days depending on the total length of the test.  The x-
     coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the cell delay in us.  The integration time per point MUST be
     indicated.

     The histogram results SHOULD display the peak-to-peak cell delay.
     The x-coordinate SHOULD be the cell delay in us with at least 256
     bins.  The y-coordinate SHOULD be the number of cells observed in
     each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.1.9. Two-point CDV/Mixed Load/Twelve VCCs

  Objective: To determine the SUT variation in cell transfer delay with
  twelve VCCs as defined in RFC 2761 "Terminology for ATM
  Benchmarking".





Dunn & Martin                Informational                     [Page 36]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCC's.  Each VCC
      MUST be defined as one of the Bearer classes for a total of four
      CBR, four UBR and four VBR VCC's.  Each VCC SHOULD contain one
      VPI/VCI.  The VPI/VCI MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing timestamps
      through the SUT via the defined test VCCs.  Each generated VCC
      stream MUST match the corresponding VCC Bearer class.  All of the
      VPI/VCI pairs will generate traffic at the same traffic rate.
      Since this test is not a throughput test, the rate should not be
      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device for all VCCs.

  Reporting Format:

     The results of the Two-point CDV/Mixed Load/Twelve VCCs test
     SHOULD be reported in a form of text, graph, and histograms.

     The text results SHOULD display the numerical values of the CDV.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI values, total number of cells transmitted and received on
     each VCC during the test in positive integers, maximum and minimum
     CDV on each VCC during the test in us, and peak-to-peak CDV on
     each VCC in us.

     The graph results SHOULD display the cell delay values.  The x-
     coordinate SHOULD be the test run time in either seconds, minutes
     or days depending on the total length of the test.  The x-
     coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the cell delay for each VCC in ms.  There SHOULD be 12 curves
     on the graph, one curves indicated and labeled for each VCC.  The
     integration time per point MUST be indicated.





Dunn & Martin                Informational                     [Page 37]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The histograms SHOULD display the peak-to-peak cell delay.  There
     will be one histogram for each VCC.  The x-coordinate SHOULD be
     the cell delay in us with at least 256 bins.  The y-coordinate
     SHOULD be the number of cells observed in each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.1.10. Two-point CDV/Mixed Load/Maximum VCCs

  Objective: To determine the SUT variation in cell transfer delay with
  the maximum number VCCs supported on the SUT as defined in RFC 2761
  "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per
      VPI.  Each VCC MUST be defined as one of the Bearer classes for a
      total of (max VCC/3) CBR, (max VCC/3) UBR and (max VCC/3) VBR
      VCC's.  If the maximum number of VCC's is not divisible by 3, the
      total for each bearer class MUST be within 3 VCC's of each other.
      The VPI/VCI MUST not be one of the reserved ATM signaling
      channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing timestamps
      through the SUT via the defined test VCCs.  Each generated VCC
      stream MUST match the corresponding VCC Bearer class.  All of the
      VPI/VCI pairs will generate traffic at the same traffic rate.
      Since this test is not a throughput test, the rate should not be
      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device for all VCCs.




Dunn & Martin                Informational                     [Page 38]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  Reporting Format:

     The results of the Two-point CDV/Mixed Load/Maximum VCCs test
     SHOULD be reported in a form of text, graphs, and histograms.

     The text results SHOULD display the numerical values of the CDV.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI values, total number of cells transmitted and received on
     each VCC during the test in positive integers, maximum and minimum
     CDV on each VCC during the test in us, and peak-to-peak CDV on
     each VCC in us.

     The graph results SHOULD display the cell delay values.  There
     will be (Max number of VCCs/10) graphs, with 10 VCCs indicated on
     each graph.  The x-coordinate SHOULD be the test run time in
     either seconds, minutes or days depending on the total length of
     the test.  The x-coordinate time SHOULD be configurable.  The y-
     coordinate SHOULD be the cell delay for each VCC in us.  There
     SHOULD be no more than 10 curves on each graph, one curve
     indicated and labeled for each VCC.  The integration time per
     point MUST be indicated.

     The histograms SHOULD display the peak-to-peak cell delay.  There
     will be one histogram for each VCC.  The x-coordinate SHOULD be
     the cell delay in us with at least 256 bins.  The y-coordinate
     SHOULD be the number of cells observed in each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.2. Cell Error Ratio (CER)

3.2.2.1. Test Setup

  The cell error ratio measurements assume that both the transmitter
  and receiver payload information is synchronized.  Synchronization
  MUST be achieved by supplying a known bit pattern to both the
  transmitter and receiver.  If this bit pattern is longer than the
  packet size, the receiver MUST synchronize with the transmitter
  before tests can be run.








Dunn & Martin                Informational                     [Page 39]

RFC 3116            Methodology for ATM Benchmarking           June 2001


3.2.2.2. CER/Steady Load/One VCC

  Objective: To determine the SUT ratio of errored cells on one VCC in
  a transmission in relation to the total cells sent as defined in RFC
  2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with one VCC.  The VCC SHOULD
      contain one VPI/VCI.  The VCC MUST be configured as either a CBR,
      VBR, or UBR connection.  The VPI/VCI MUST not be one of the
      reserved ATM signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a constant rate through the SUT via the
      defined test VCC.  Since this test is not a throughput test, the
      rate should not be greater than 90% of line rate.  The IP PDUs
      MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the number of bit errors at the receiver end of the test
      device.

  Reporting Format:

     The results of the CER/Steady Load/One VCC test SHOULD be reported
     in a form of text and graph.

     The text results SHOULD display the numerical values of the CER.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, and the
     CER for the entire test.

     The graph results SHOULD display the cell error ratio values.  The
     x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the CER.  The integration time per point MUST be indicated.





Dunn & Martin                Informational                     [Page 40]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST be indicated.  The generated bit pattern MUST
     also be indicated.

3.2.2.3. CER/Steady Load/Twelve VCCs

  Objective: To determine the SUT ratio of errored cells on twelve VCCs
  in a transmission in relation to the total cells sent as defined in
  RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCCs, using 1 VPI
      and 12 VCIs.  The VCC's MUST be configured as either a CBR, VBR,
      or UBR connection.  The VPI/VCIs MUST not be one of the reserved
      ATM signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a constant rate through the SUT via the
      defined test VCCs.  All of the VPI/VCI pairs will generate
      traffic at the same traffic rate.  Since this test is not a
      throughput test, the rate should not be greater than 90% of line
      rate.  The IP PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of bit errors at the receiver end of the test
      device for all VCCs.

  Reporting Format:

     The results of the CER/Steady Load/Twelve VCCs test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CER.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, and the
     CER for the entire test.



Dunn & Martin                Informational                     [Page 41]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The graph results SHOULD display the cell error ratio values.  The
     x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the CER for each VCC.  There should be 12 curves on the graph,
     on curve indicated and labeled for each VCC.  The integration time
     per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST be indicated.  The generated bit pattern MUST
     also be indicated.

3.2.2.4. CER/Steady Load/Maximum VCCs

  Objective: To determine the SUT ratio of errored cells with the
  maximum number VCCs supported on the SUT in a transmission in
  relation to the total cells sent as defined in RFC 2761 "Terminology
  for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with the maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per
      VPI.  The VCC's MUST be configured as either a CBR, VBR, or UBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a constant rate through the SUT via the
      defined test VCCs.  All of the VPI/VCI pairs will generate
      traffic at the same traffic rate.  Since this test is not a
      throughput test, the rate should not be greater than 90% of line
      rate.  The IP PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of bit errors at the receiver end of the test
      device for all VCCs.



Dunn & Martin                Informational                     [Page 42]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  Reporting Format:

     The results of the CER/Steady Load/Maximum VCCs test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CER.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, and the
     CER for the entire test.

     The graph results SHOULD display the cell error ratio values.
     There will be (Max number of VCCs/10) graphs, with 10 VCCs
     indicated on each graph.  The x-coordinate SHOULD be the test run
     time in either seconds, minutes or days depending on the total
     length of the test.  The x-coordinate time SHOULD be configurable.
     The y-coordinate SHOULD be the CER for each VCC.  There SHOULD be
     no more than 10 curves on each graph, one curve indicated and
     labeled for each VCC.  The integration time per point MUST be
     indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST be indicated.  The generated bit pattern MUST
     also be indicated.

3.2.2.5. CER/Bursty VBR Load/One VCC

  Objective: To determine the SUT ratio of errored cells on one VCC in
  a transmission in relation to the total cells sent as defined in RFC
  2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with one VCC.  The VCC SHOULD
      contain one VPI/VCI.  The VCC MUST be configured as either a CBR
      or VBR connection.  The VPI/VCI MUST not be one of the reserved
      ATM signaling channels (e.g., [0,5], [0,16]).  The PCR, SCR, and
      MBS must be configured using one of the specified traffic
      descriptors.






Dunn & Martin                Informational                     [Page 43]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific VBR rate through the SUT via
      the defined test VCC.  Since this test is not a throughput test,
      the rate should not be greater than 90% of line rate.  The IP
      PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the number of bit errors at the receiver end of the test
      device.

  Reporting Format:

     The results of the CER/Bursty VBR Load/One VCC test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CER.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, and the
     CER for the entire test.

     The graph results SHOULD display the cell error ratio values.  The
     x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the CER.  The integration time per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST be indicated.  The generated bit pattern MUST
     also be indicated.

3.2.2.6. CER/Bursty VBR Load/Twelve VCCs

  Objective: To determine the SUT ratio of errored cells on twelve VCCs
  in a transmission in relation to the total cells sent as defined in
  RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.



Dunn & Martin                Informational                     [Page 44]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  2)  Configure the SUT and test device with twelve VCCs, using 1 VPI
      and 12 VCIs.  The VCC's MUST be configured as either a CBR or VBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).  The PCR, SCR, and MBS
      must be configured using one of the specified traffic
      descriptors.

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific VBR rate through the SUT via
      the defined test VCCs.  All of the VPI/VCI pairs will generate
      traffic at the same traffic rate.  Since this test is not a
      throughput test, the rate should not be greater than 90% of line
      rate.  The PCR, SCR, and MBS must be indicated.  The IP PDUs MUST
      be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of bit errors at the receiver end of the test
      device for all VCCs.

  Reporting Format:

     The results of the CER/Bursty VBR Load/Twelve VCCs test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CER.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, and the
     CER for the entire test.

     The graph results SHOULD display the cell error ratio values.  The
     x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the CER for each VCC.  There should be 12 curves on the graph,
     on curve indicated and labeled for each VCC.  The integration time
     per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST be indicated.  The generated bit pattern MUST
     also be indicated.



Dunn & Martin                Informational                     [Page 45]

RFC 3116            Methodology for ATM Benchmarking           June 2001


3.2.2.7. CER/Bursty VBR Load/Maximum VCCs

  Objective: To determine the SUT ratio of errored cells with the
  maximum number VCCs supported on the SUT in a transmission in
  relation to the total cells sent as defined in RFC 2761 "Terminology
  for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with the maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per
      VPI.  The VCC's MUST be configured as either a CBR or VBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).  The PCR, SCR, and MBS
      must be configured using one of the specified traffic
      descriptors.

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific VBR rate through the SUT via
      the defined test VCCs.  All of the VPI/VCI pairs will generate
      traffic at the same traffic rate.  Since this test is not a
      throughput test, the rate should not be greater than 90% of line
      rate.  The IP PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of bit errors at the receiver end of the test
      device for all VCCs.

  Reporting Format:

     The results of the CER/Bursty VBR Load/Maximum VCCs test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CER.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, and the
     CER for the entire test.





Dunn & Martin                Informational                     [Page 46]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The graph results SHOULD display the cell error ratio values.
     There will be (Max number of VCCs/10) graphs, with 10 VCCs
     indicated on each graph.  The x-coordinate SHOULD be the test run
     time in either seconds, minutes or days depending on the total
     length of the test.  The x-coordinate time SHOULD be configurable.
     The y-coordinate SHOULD be the CER for each VCC.  There SHOULD be
     no more than 10 curves on each graph, one curve indicated and
     labeled for each VCC.  The integration time per point MUST be
     indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST be indicated.  The generated bit pattern MUST
     also be indicated.

3.2.3. Cell Loss Ratio (CLR)

3.2.3.1. CLR/Steady Load/One VCC

  Objective: To determine the SUT ratio of lost cells on one VCC in a
  transmission in relation to the total cells sent as defined in RFC
  2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with one VCC.  The VCC SHOULD
      contain one VPI/VCI.  The VCC MUST be configured as either a CBR,
      VBR, or UBR connection.  The VPI/VCI MUST not be one of the
      reserved ATM signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets at a specific constant rate
      through the SUT via the defined test VCC.  Since this test is not
      a throughput test, the rate should not be greater than 90% of
      line rate.  The IP PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the number of cells transmitted and received on the test
      device.




Dunn & Martin                Informational                     [Page 47]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  Reporting Format:

     The results of the CLR/Steady Load/One VCC test SHOULD be reported
     in a form of text and graph.

     The text results SHOULD display the numerical values of the CLR.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, and the
     CLR for the entire test.

     The graph results SHOULD display the Cell Loss ratio values.  The
     x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the CLR.  The integration time per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.3.2. CLR/Steady Load/Twelve VCCs

  Objective: To determine the SUT ratio of lost cells on twelve VCCs in
  a transmission in relation to the total cells sent as defined in RFC
  2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCCs, using 1 VPI
      and 12 VCIs.  The VCC's MUST be configured as either a CBR, VBR,
      or UBR connection.  The VPI/VCIs MUST not be one of the reserved
      ATM signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets at a specific constant rate
      through the SUT via the defined test VCCs.  All of the VPI/VCI
      pairs will generate traffic at the same traffic rate.  Since this
      test is not a throughput test, the rate should not be greater
      than 90% of line rate.  The IP PDUs MUST be encapsulated in AAL5.







Dunn & Martin                Informational                     [Page 48]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of cells transmitted and received per VCC on
      the test device.

  Reporting Format:

     The results of the CLR/Steady Load/Twelve VCCs test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CLR.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, and the
     CLR for the entire test.

     The graph results SHOULD display the Cell Loss ratio values.  The
     x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the CLR for each VCC.  There should be 12 curves on the graph,
     on curve indicated and labeled for each VCC.  The integration time
     per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.3.3. CLR/Steady Load/Maximum VCCs

  Objective: To determine the SUT ratio of lost cells with the maximum
  number VCCs supported on the SUT in a transmission in relation to the
  total cells sent as defined in RFC 2761 "Terminology for ATM
  Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with the maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per



Dunn & Martin                Informational                     [Page 49]

RFC 3116            Methodology for ATM Benchmarking           June 2001


      VPI.  The VCC's MUST be configured as either a CBR, VBR, or UBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets at a specific constant rate
      through the SUT via the defined test VCCs.  All of the VPI/VCI
      pairs will generate traffic at the same traffic rate.  Since this
      test is not a throughput test, the rate should not be greater
      than 90% of line rate.  The IP PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of cells transmitted and received per VCC on
      the test device.

  Reporting Format:

     The results of the CLR/Steady Load/Maximum VCCs test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CLR.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, and the
     CLR for the entire test.

     The graph results SHOULD display the Cell Loss ratio values.
     There will be (Max number of VCCs/10) graphs, with 10 VCCs
     indicated on each graph.  The x-coordinate SHOULD be the test run
     time in either seconds, minutes or days depending on the total
     length of the test.  The x-coordinate time SHOULD be configurable.
     The y-coordinate SHOULD be the CLR for each VCC.  There SHOULD be
     no more than 10 curves on each graph, one curve indicated and
     labeled for each VCC.  The integration time per point MUST be
     indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.







Dunn & Martin                Informational                     [Page 50]

RFC 3116            Methodology for ATM Benchmarking           June 2001


3.2.3.4. CLR/Bursty VBR Load/One VCC

  Objective: To determine the SUT ratio of lost cells on one VCC in a
  transmission in relation to the total cells sent as defined in RFC
  2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with one VCC.  The VCC SHOULD
      contain one VPI/VCI.  The VCC MUST be configured as either a CBR
      or VBR connection.  The VPI/VCI MUST not be one of the reserved
      ATM signaling channels (e.g., [0,5], [0,16]).  The PCR, SCR, and
      MBS must be configured using one of the specified traffic
      descriptors.

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific rate through the SUT via the
      defined test VCC.  Since this test is not a throughput test, the
      rate should not be greater than 90% of line rate.  The IP PDUs
      MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the number of cells transmitted and received on the test
      device.

  Reporting Format:

     The results of the CLR/Bursty VBR Load/One VCC test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CLR.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, and the
     CLR for the entire test.

     The graph results SHOULD display the Cell Loss ratio values.  The
     x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the CLR.  The integration time per point MUST be indicated.



Dunn & Martin                Informational                     [Page 51]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.3.5. CLR/Bursty VBR Load/Twelve VCCs

  Objective: To determine the SUT ratio of lost cells on twelve VCCs in
  a transmission in relation to the total cells sent as defined in RFC
  2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCCs, using 1 VPI
      and 12 VCIs.  The VCC MUST be configured as either a CBR or VBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).  The PCR, SCR, and MBS
      must be configured using one of the specified traffic
      descriptors.

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific rate through the SUT via the
      defined test VCCs.  All of the VPI/VCI pairs will generate
      traffic at the same traffic rate.  Since this test is not a
      throughput test, the rate should not be greater than 90% of line
      rate.  The PCR, SCR, and MBS must be indicated.  The IP PDUs MUST
      be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of cells transmitted and received per VCC on
      the test device.

  Reporting Format:

     The results of the CLR/Bursty VBR Load/Twelve VCCs test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CLR.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on



Dunn & Martin                Informational                     [Page 52]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     the given VPI/VCI during the test in positive integers, and the
     CLR for the entire test.

     The graph results SHOULD display the Cell Loss ratio values.  The
     x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the CLR for each VCC.  There should be 12 curves on the graph,
     on curve indicated and labeled for each VCC.  The integration time
     per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.3.6. CLR/Bursty VBR Load/Maximum VCCs

  Objective: To determine the SUT ratio of lost cells with the maximum
  number VCCs supported on the SUT in a transmission in relation to the
  total cells sent as defined in RFC 2761 "Terminology for ATM
  Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with the maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per
      VPI.  The VCC MUST be configured as either a CBR or VBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).  The PCR, SCR, and MBS
      must be configured using one of the specified traffic
      descriptors.

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific rate through the SUT via the
      defined test VCCs.  All of the VPI/VCI pairs will generate
      traffic at the same traffic rate.  Since this test is not a
      throughput test, the rate should not be greater than 90% of line
      rate.  The IP PDUs MUST be encapsulated in AAL5.







Dunn & Martin                Informational                     [Page 53]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of cells transmitted and received per VCC on
      the test device.

  Reporting Format:

     The results of the CLR/Bursty VBR Load/Maximum VCCs test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CLR.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, and the
     CLR for the entire test.

     The graph results SHOULD display the Cell Loss ratio values.
     There will be (Max number of VCCs/10) graphs, with 10 VCCs
     indicated on each graph.  The x-coordinate SHOULD be the test run
     time in either seconds, minutes or days depending on the total
     length of the test.  The x-coordinate time SHOULD be configurable.
     The y-coordinate SHOULD be the CLR for each VCC.  There SHOULD be
     no more than 10 curves on each graph, one curve indicated and
     labeled for each VCC.  The integration time per point MUST be
     indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.4. Cell Misinsertion Rate (CMR)

3.2.4.1. CMR/Steady Load/One VCC

  Objective: To determine the SUT ratio of cell misinsertion on one VCC
  in a transmission in relation to the total cells sent as defined in
  RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.




Dunn & Martin                Informational                     [Page 54]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  2)  Configure the SUT and test device with one VCC.  The VCC MUST be
      configured as either a CBR, VBR, or UBR connection.  The VCC
      SHOULD contain one VPI/VCI.  The VPI/VCI MUST not be one of the
      reserved ATM signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets at a specific constant rate
      through the SUT via the defined test VCC.  Since this test is not
      a throughput test, the rate should not be greater than 90% of
      line rate.  The IP PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the number of cell misinsertion errors at the receiver end
      of the test device.

  Reporting Format:

     The results of the CMR/Steady Load/One VCC test SHOULD be reported
     in a form of text and graph.

     The text results SHOULD display the numerical values of the CMR.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, and the
     CMR for the entire test.

     The graph results SHOULD display the Cell misinsertion rate
     values.  The x-coordinate SHOULD be the test run time in either
     seconds, minutes or days depending on the total length of the
     test.  The x-coordinate time SHOULD be configurable.  The y-
     coordinate SHOULD be the CMR.  The integration time per point MUST
     be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.4.2. CMR/Steady Load/Twelve VCCs

  Objective: To determine the SUT rate of misinserted cells on twelve
  VCCs in a transmission in relation to the total cells sent as defined
  in RFC 2761 "Terminology for ATM Benchmarking".




Dunn & Martin                Informational                     [Page 55]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCCs, using 1 VPI
      and 12 VCIs.  The VCC's MUST be configured as either a CBR, VBR,
      or UBR connection.  The VPI/VCIs MUST not be one of the reserved
      ATM signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets at a specific constant rate
      through the SUT via the defined test VCCs.  All of the VPI/VCI
      pairs will generate traffic at the same traffic rate.  Since this
      test is not a throughput test, the rate should not be greater
      than 90% of line rate.  The IP PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of cell misinsertion errors at the receiver end
      of the test device per VCC.

  Reporting Format:

     The results of the CMR/Steady Load/Twelve VCCs test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CMR.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, and the
     CMR for the entire test.

     The graph results SHOULD display the Cell misinsertion rate
     values.  The x-coordinate SHOULD be the test run time in either
     seconds, minutes or days depending on the total length of the
     test.  The x-coordinate time SHOULD be configurable.  The y-
     coordinate SHOULD be the CMR for each VCC.  There should be 12
     curves on the graph, on curve indicated and labeled for each VCC.
     The integration time per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.



Dunn & Martin                Informational                     [Page 56]

RFC 3116            Methodology for ATM Benchmarking           June 2001


3.2.4.3. CMR/Steady Load/Maximum VCCs

  Objective: To determine the SUT rate of misinserted cells with the
  maximum number VCCs supported on the SUT in a transmission in
  relation to the total cells sent as defined in RFC 2761 "Terminology
  for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with the maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per
      VPI.  The VCC's MUST be configured as either a CBR, VBR, or UBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets at a specific constant rate
      through the SUT via the defined test VCCs.  All of the VPI/VCI
      pairs will generate traffic at the same traffic rate.  Since this
      test is not a throughput test, the rate should not be greater
      than 90% of line rate.  The IP PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of cell misinsertion errors at the receiver end
      of the test device per VCC.

  Reporting Format:

     The results of the CMR/Steady Load/Maximum VCCs test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CMR.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, and the
     CMR for the entire test.

     The graph results SHOULD display the Cell misinsertion rate
     values.  There will be (Max number of VCCs/10) graphs, with 10
     VCCs indicated on each graph.  The x-coordinate SHOULD be the test
     run time in either seconds, minutes or days depending on the total



Dunn & Martin                Informational                     [Page 57]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     length of the test.  The x-coordinate time SHOULD be configurable.
     The y-coordinate SHOULD be the CMR for each VCC.  There SHOULD be
     no more than 10 curves on each graph, one curve indicated and
     labeled for each VCC.  The integration time per point MUST be
     indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.4.4. CMR/Bursty VBR Load/One VCC

  Objective: To determine the SUT rate of misinserted cells on one VCC
  in a transmission in relation to the total cells sent as defined in
  RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with one VCC.  The VCC SHOULD
      contain one VPI/VCI.  The VCC MUST be configured as either a CBR
      or VBR connection.  The VPI/VCI MUST not be one of the reserved
      ATM signaling channels (e.g., [0,5], [0,16]).  The PCR, SCR, and
      MBS must be configured using one of the specified traffic
      descriptors.

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific rate through the SUT via the
      defined test VCC.  Since this test is not a throughput test, the
      rate should not be greater than 90% of line rate.  The IP PDUs
      MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the number of cell misinsertion errors at the receiver end
      of the test device.

  Reporting Format:

     The results of the CMR/Bursty VBR Load/One VCC test SHOULD be
     reported in a form of text and graph.



Dunn & Martin                Informational                     [Page 58]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The text results SHOULD display the numerical values of the CMR.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, and the
     CMR for the entire test.

     The graph results SHOULD display the Cell misinsertion rate
     values.  The x-coordinate SHOULD be the test run time in either
     seconds, minutes or days depending on the total length of the
     test.  The x-coordinate time SHOULD be configurable.  The y-
     coordinate SHOULD be the CMR.  The integration time per point MUST
     be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.4.5. CMR/Bursty VBR Load/Twelve VCCs

  Objective: To determine the SUT rate of misinserted cells on twelve
  VCCs in a transmission in relation to the total cells sent as defined
  in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCCs, using 1 VPI
      and 12 VCIs.  The VCC's MUST be configured as either a CBR or VBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).  The PCR, SCR, and MBS
      must be configured using one of the specified traffic
      descriptors.

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific rate through the SUT via the
      defined test VCCs.  All of the VPI/VCI pairs will generate
      traffic at the same traffic rate.  Since this test is not a
      throughput test, the rate should not be greater than 90% of line
      rate.  The PCR, SCR, and MBS must be indicated.  The IP PDUs MUST
      be encapsulated in AAL5.







Dunn & Martin                Informational                     [Page 59]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of cell misinsertion errors at the receiver end
      of the test device per VCC.

  Reporting Format:

     The results of the CMR/Bursty VBR Load/Twelve VCCs test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CMR.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, and the
     CMR for the entire test.

     The graph results SHOULD display the Cell misinsertion rate
     values.  The x-coordinate SHOULD be the test run time in either
     seconds, minutes or days depending on the total length of the
     test.  The x-coordinate time SHOULD be configurable.  The y-
     coordinate SHOULD be the CMR for each VCC.  There should be 12
     curves on the graph, on curve indicated and labeled for each VCC.
     The integration time per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.4.6. CMR/Bursty VBR Load/Maximum VCCs

  Objective: To determine the SUT rate of misinserted cells with the
  maximum number VCCs supported on the SUT in a transmission in
  relation to the total cells sent as defined in RFC 2761 "Terminology
  for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with the maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per



Dunn & Martin                Informational                     [Page 60]

RFC 3116            Methodology for ATM Benchmarking           June 2001


      VPI.  The VCC's MUST be configured as either a CBR or VBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).  The PCR, SCR, and MBS
      must be configured using one of the specified traffic
      descriptors.

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific rate through the SUT via the
      defined test VCCs.  All of the VPI/VCI pairs will generate
      traffic at the same traffic rate.  Since this test is not a
      throughput test, the rate should not be greater than 90% of line
      rate.  The IP PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of cell misinsertion errors at the receiver end
      of the test device per VCC.

  Reporting Format:

     The results of the CMR/Bursty VBR Load/Maximum VCCs test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CMR.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, and the
     CMR for the entire test.

     The graph results SHOULD display the Cell misinsertion rate
     values.  There will be (Max number of VCCs/10) graphs, with 10
     VCCs indicated on each graph.  The x-coordinate SHOULD be the test
     run time in either seconds, minutes or days depending on the total
     length of the test.  The x-coordinate time SHOULD be configurable.
     The y-coordinate SHOULD be the CMR for each VCC.  There SHOULD be
     no more than 10 curves on each graph, one curve indicated and
     labeled for each VCC.  The integration time per point MUST be
     indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.




Dunn & Martin                Informational                     [Page 61]

RFC 3116            Methodology for ATM Benchmarking           June 2001


3.2.5. CRC Error Ratio (CRC-ER)

3.2.5.1. CRC-ER/Steady Load/One VCC

  Objective: To determine the SUT ratio of CRC errors on one VCC in a
  transmission in relation to the total cells sent as defined in RFC
  2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with one VCC.  The VCC SHOULD
      contain one VPI/VCI.  The VCC MUST be configured as either a CBR,
      VBR, or UBR connection.  The VPI/VCI MUST not be one of the
      reserved ATM signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets at a specific constant rate
      through the SUT via the defined test VCC.  Since this test is not
      a throughput test, the rate should not be greater than 90% of
      line rate.  The IP PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the number of CRC errored cells received on the test
      device.

  Reporting Format:

     The results of the CRC-ER/Steady Load/One VCC test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CRC-
     ER.  The values given SHOULD include: time period of test in s,
     test VPI/VCI value, total number of cells transmitted and received
     on the given VPI/VCI during the test in positive integers, and the
     CRC-ER for the entire test.

     The graph results SHOULD display the CRC Error ratio values.  The
     x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the CRC-ER.  The integration time per point MUST be indicated.




Dunn & Martin                Informational                     [Page 62]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.5.2. CRC-ER/Steady Load/Twelve VCCs

  Objective: To determine the SUT ratio of lost cells on twelve VCCs in
  a transmission in relation to the total cells sent as defined in RFC
  2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCCs, using 1 VPI
      and 12 VCIs.  The VCC's MUST be configured as either a CBR, VBR,
      or UBR connection.  The VPI/VCIs MUST not be one of the reserved
      ATM signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets at a specific constant rate
      through the SUT via the defined test VCCs.  All of the VPI/VCI
      pairs will generate traffic at the same traffic rate.  Since this
      test is not a throughput test, the rate should not be greater
      than 90% of line rate.  The IP PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of CRC errored cells received per VCC on the
      test device.

  Reporting Format:

     The results of the CRC-ER/Steady Load/Twelve VCCs test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CRC-
     ER.  The values given SHOULD include: time period of test in s,
     test VPI/VCI value, total number of cells transmitted and received
     on the given VPI/VCI during the test in positive integers, and the
     CRC-ER for the entire test.





Dunn & Martin                Informational                     [Page 63]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The graph results SHOULD display the CRC Error ratio values.  The
     x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the CRC-ER for each VCC.  There should be 12 curves on the
     graph, on curve indicated and labeled for each VCC.  The
     integration time per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.5.3. CRC-ER/Steady Load/Maximum VCCs

  Objective: To determine the SUT ratio of lost cells with the maximum
  number VCCs supported on the SUT in a transmission in relation to the
  total cells sent as defined in RFC 2761 "Terminology for ATM
  Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with the maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per
      VPI.  The VCC's MUST be configured as either a CBR, VBR, or UBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets at a specific constant rate
      through the SUT via the defined test VCCs.  All of the VPI/VCI
      pairs will generate traffic at the same traffic rate.  Since this
      test is not a throughput test, the rate should not be greater
      than 90% of line rate.  The IP PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of CRC errored cells received per VCC on the
      test device.





Dunn & Martin                Informational                     [Page 64]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  Reporting Format:

     The results of the CRC-ER/Steady Load/Maximum VCCs test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CRC-
     ER.  The values given SHOULD include: time period of test in s,
     test VPI/VCI value, total number of cells transmitted and received
     on the given VPI/VCI during the test in positive integers, and the
     CRC-ER for the entire test.

     The graph results SHOULD display the CRC Error ratio values.
     There will be (Max number of VCCs/10) graphs, with 10 VCCs
     indicated on each graph.  The x-coordinate SHOULD be the test run
     time in either seconds, minutes or days depending on the total
     length of the test.  The x-coordinate time SHOULD be configurable.
     The y-coordinate SHOULD be the CRC-ER for each VCC.  There SHOULD
     be no more than 10 curves on each graph, one curve indicated and
     labeled for each VCC.  The integration time per point MUST be
     indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.5.4. CRC-ER/Bursty VBR Load/One VCC

  Objective: To determine the SUT ratio of lost cells on one VCC in a
  transmission in relation to the total cells sent as defined in RFC
  2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with one VCC.  The VCC SHOULD
      contain one VPI/VCI.  The VCC MUST be configured as either a CBR
      or VBR connection.  The VPI/VCI MUST not be one of the reserved
      ATM signaling channels (e.g., [0,5], [0,16]).  The PCR, SCR, and
      MBS must be configured using one of the specified traffic
      descriptors.

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific rate through the SUT via the
      defined test VCC.  Since this test is not a throughput test, the



Dunn & Martin                Informational                     [Page 65]

RFC 3116            Methodology for ATM Benchmarking           June 2001


      rate should not be greater than 90% of line rate.  The IP PDUs
      MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the number of CRC errored cells received per VCC on the
      test device.

  Reporting Format:

     The results of the CRC-ER/Bursty VBR Load/One VCC test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CRC-
     ER.  The values given SHOULD include: time period of test in s,
     test VPI/VCI value, total number of cells transmitted and received
     on the given VPI/VCI during the test in positive integers, and the
     CRC-ER for the entire test.

     The graph results SHOULD display the CRC Error ratio values.  The
     x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the CRC-ER.  The integration time per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.5.5. CRC-ER/Bursty VBR Load/Twelve VCCs

  Objective: To determine the SUT ratio of lost cells on twelve VCCs in
  a transmission in relation to the total cells sent as defined in RFC
  2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCCs, using 1 VPI
      and 12 VCIs.  The VCC MUST be configured as either a CBR or VBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM



Dunn & Martin                Informational                     [Page 66]

RFC 3116            Methodology for ATM Benchmarking           June 2001


      signaling channels (e.g., [0,5], [0,16]).  The PCR, SCR, and MBS
      must be configured using one of the specified traffic
      descriptors.

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific rate through the SUT via the
      defined test VCCs.  All of the VPI/VCI pairs will generate
      traffic at the same traffic rate.  Since this test is not a
      throughput test, the rate should not be greater than 90% of line
      rate.  The PCR, SCR, and MBS must be indicated.  The IP PDUs MUST
      be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of CRC errored cells received per VCC on the
      test device for all VCCs.

  Reporting Format:

     The results of the CRC-ER/Bursty VBR Load/Twelve VCCs test SHOULD
     be reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CRC-
     ER.  The values given SHOULD include: time period of test in s,
     test VPI/VCI value, total number of cells transmitted and received
     on the given VPI/VCI during the test in positive integers, and the
     CRC-ER for the entire test.

     The graph results SHOULD display the CRC Error ratio values.  The
     x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the CRC-ER for each VCC.  There should be 12 curves on the
     graph, on curve indicated and labeled for each VCC.  The
     integration time per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.







Dunn & Martin                Informational                     [Page 67]

RFC 3116            Methodology for ATM Benchmarking           June 2001


3.2.5.6. CRC-ER/Bursty VBR Load/Maximum VCCs

  Objective: To determine the SUT ratio of lost cells with the maximum
  number VCCs supported on the SUT in a transmission in relation to the
  total cells sent as defined in RFC 2761 "Terminology for ATM
  Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with the maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per
      VPI.  The VCC MUST be configured as either a CBR or VBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).  The PCR, SCR, and MBS
      must be configured using one of the specified traffic
      descriptors.

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific rate through the SUT via the
      defined test VCCs.  All of the VPI/VCI pairs will generate
      traffic at the same traffic rate.  Since this test is not a
      throughput test, the rate should not be greater than 90% of line
      rate.  The IP PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of CRC errored cells received per VCC on the
      test device for all VCCs.

  Reporting Format:

     The results of the CRC-ER/Bursty VBR Load/Maximum VCCs test SHOULD
     be reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CRC-
     ER.  The values given SHOULD include: time period of test in s,
     test VPI/VCI value, total number of cells transmitted and received
     on the given VPI/VCI during the test in positive integers, and the
     CRC-ER for the entire test.

     The graph results SHOULD display the CRC Error ratio values.



Dunn & Martin                Informational                     [Page 68]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     There will be (Max number of VCCs/10) graphs, with 10 VCCs
     indicated on each graph.  The x-coordinate SHOULD be the test run
     time in either seconds, minutes or days depending on the total
     length of the test.  The x-coordinate time SHOULD be configurable.
     The y-coordinate SHOULD be the CRC-ER for each VCC.  There SHOULD
     be no more than 10 curves on each graph, one curve indicated and
     labeled for each VCC.  The integration time per point MUST be
     indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.5.7. CRC-ER/Bursty UBR Load/One VCC

  Objective: To determine the SUT ratio of lost cells on one VCC in a
  transmission in relation to the total cells sent as defined in RFC
  2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with one VCC.  The VCC SHOULD
      contain one VPI/VCI.  The VCC MUST be configured as a UBR
      connection.  The VPI/VCI MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).  The PCR, SCR, and MBS
      must be configured using one of the specified traffic
      descriptors.

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific rate through the SUT via the
      defined test VCC.  Since this test is not a throughput test, the
      rate should not be greater than 90% of line rate.  The IP PDUs
      MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the number of CRC errored cells received per VCC on the
      test device.





Dunn & Martin                Informational                     [Page 69]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  Reporting Format:

     The results of the CRC-ER/Bursty UBR Load/One VCC test SHOULD be
     reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CRC-
     ER.  The values given SHOULD include: time period of test in s,
     test VPI/VCI value, total number of cells transmitted and received
     on the given VPI/VCI during the test in positive integers, and the
     CRC-ER for the entire test.

     The graph results SHOULD display the CRC Error ratio values.  The
     x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the CRC-ER.  The integration time per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.5.8. CRC-ER/Bursty UBR Load/Twelve VCCs

  Objective: To determine the SUT ratio of lost cells on twelve VCCs in
  a transmission in relation to the total cells sent as defined in RFC
  2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCCs, using 1 VPI
      and 12 VCIs.  The VCC MUST be configured as a UBR connection.
      The VPI/VCIs MUST not be one of the reserved ATM signaling
      channels (e.g., [0,5], [0,16]).  The PCR, SCR, and MBS must be
      configured using one of the specified traffic descriptors.

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific rate through the SUT via the
      defined test VCCs.  All of the VPI/VCI pairs will generate
      traffic at the same traffic rate.  Since this test is not a
      throughput test, the rate should not be greater than 90% of line
      rate.  The PCR, SCR, and MBS must be indicated.  The IP PDUs MUST
      be encapsulated in AAL5.




Dunn & Martin                Informational                     [Page 70]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of CRC errored cells received per VCC on the
      test device for all VCCs.

  Reporting Format:

     The results of the CRC-ER/Bursty UBR Load/Twelve VCCs test SHOULD
     be reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CRC-
     ER.  The values given SHOULD include: time period of test in s,
     test VPI/VCI value, total number of cells transmitted and received
     on the given VPI/VCI during the test in positive integers, and the
     CRC-ER for the entire test.

     The graph results SHOULD display the CRC Error ratio values.  The
     x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the CRC-ER for each VCC.  There should be 12 curves on the
     graph, on curve indicated and labeled for each VCC.  The
     integration time per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.5.9. CRC-ER/Bursty UBR Load/Maximum VCCs

  Objective: To determine the SUT ratio of lost cells with the maximum
  number VCCs supported on the SUT in a transmission in relation to the
  total cells sent as defined in RFC 2761 "Terminology for ATM
  Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with the maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per



Dunn & Martin                Informational                     [Page 71]

RFC 3116            Methodology for ATM Benchmarking           June 2001


      VPI.  The VCC MUST be configured as a UBR connection.  The
      VPI/VCIs MUST not be one of the reserved ATM signaling channels
      (e.g., [0,5], [0,16]).  The PCR, SCR, and MBS must be configured
      using one of the specified traffic descriptors.

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific rate through the SUT via the
      defined test VCCs.  All of the VPI/VCI pairs will generate
      traffic at the same traffic rate.  Since this test is not a
      throughput test, the rate should not be greater than 90% of line
      rate.  The IP PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of CRC errored cells received per VCC on the
      test device for all VCCs.

  Reporting Format:

     The results of the CRC-ER/Bursty UBR Load/Maximum VCCs test SHOULD
     be reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CRC-
     ER.  The values given SHOULD include: time period of test in s,
     test VPI/VCI value, total number of cells transmitted and received
     on the given VPI/VCI during the test in positive integers, and the
     CRC-ER for the entire test.

     The graph results SHOULD display the CRC Error ratio values.
     There will be (Max number of VCCs/10) graphs, with 10 VCCs
     indicated on each graph.  The x-coordinate SHOULD be the test run
     time in either seconds, minutes or days depending on the total
     length of the test.  The x-coordinate time SHOULD be configurable.
     The y-coordinate SHOULD be the CRC-ER for each VCC.  There SHOULD
     be no more than 10 curves on each graph, one curve indicated and
     labeled for each VCC.  The integration time per point MUST be
     indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.





Dunn & Martin                Informational                     [Page 72]

RFC 3116            Methodology for ATM Benchmarking           June 2001


3.2.5.10. CRC-ER/Bursty Mixed Load/Three VCC

  Objective: To determine the SUT ratio of lost cells on three VCC's in
  relation to the total cells sent as defined in RFC 2761 "Terminology
  for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with three VCC's.  Each VCC
      MUST be defined as a different Bearer class; one CBR, one UBR and
      one VBR.  Each VCC SHOULD contain one VPI/VCI.  The VPI/VCI MUST
      not be one of the reserved ATM signaling channels (e.g., [0,5],
      [0,16]).  The PCR, SCR, and MBS must be configured using one of
      the specified traffic descriptors.

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns through the SUT via the defined test VCCs.
      Each generated VCC stream MUST match the corresponding VCC Bearer
      class.  All of the VPI/VCI pairs will generate traffic at the
      same traffic rate.  Since this test is not a throughput test, the
      rate should not be greater than 90% of line rate.  The IP PDUs
      MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the number of CRC errored cells received per VCC on the
      test device.

  Reporting Format:

     The results of the CRC-ER/Bursty Mixed Load/Three VCC test SHOULD
     be reported in in a form of text and graph.

     The text results SHOULD display the numerical values of the CRC-
     ER.  The values given SHOULD include: time period of test in s,
     test VPI/VCI value, total number of cells transmitted and received
     on the given VPI/VCI during the test in positive integers, and the
     CRC-ER for the entire test.

     The graph results SHOULD display the CRC Error ratio values.  The
     x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The



Dunn & Martin                Informational                     [Page 73]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the CRC-ER for each VCC.  There should be 12 curves on the
     graph, on curve indicated and labeled for each VCC.  The
     integration time per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.5.11. CRC-ER/Bursty Mixed Load/Twelve VCCs

  Objective: To determine the SUT ratio of lost cells on twelve VCCs in
  a transmission in relation to the total cells sent as defined in RFC
  2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCC's.  Each VCC
      MUST be defined as one of the Bearer classes for a total of four
      CBR, four UBR and four VBR VCC's.  Each VCC SHOULD contain one
      VPI/VCI.  The VPI/VCI MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns through the SUT via the defined test VCCs.
      Each generated VCC stream MUST match the corresponding VCC Bearer
      class.  All of the VPI/VCI pairs will generate traffic at the
      same traffic rate.  Since this test is not a throughput test, the
      rate should not be greater than 90% of line rate.  The IP PDUs
      MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of CRC errored cells received per VCC on the
      test device for all VCCs.

  Reporting Format:

     The results of the CRC-ER/Bursty Mixed Load/Twelve VCCs test
     SHOULD be reported in a form of text and graph.



Dunn & Martin                Informational                     [Page 74]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The text results SHOULD display the numerical values of the CRC-
     ER.  The values given SHOULD include: time period of test in s,
     test VPI/VCI value, total number of cells transmitted and received
     on the given VPI/VCI during the test in positive integers, and the
     CRC-ER for the entire test.

     The graph results SHOULD display the CRC Error ratio values.  The
     x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on  the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the CRC-ER for each VCC.  There should be 12 curves on the
     graph, on curve indicated and labeled for each VCC.  The
     integration time per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.5.12. CRC-ER/Bursty Mixed Load/Maximum VCCs

  Objective: To determine the SUT ratio of lost cells with the maximum
  number VCCs supported on the SUT in a transmission in relation to the
  total cells sent as defined in RFC 2761 "Terminology for ATM
  Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per
      VPI.  Each VCC MUST be defined as one of the Bearer classes for a
      total of (max VCC/3) CBR, (max VCC/3) UBR and (max VCC/3) VBR
      VCC's.  The VPI/VCI MUST not be one of the reserved ATM signaling
      channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns through the SUT via the defined test VCCs.
      Each generated VCC stream MUST match the corresponding VCC Bearer
      class.  All of the VPI/VCI pairs will generate traffic at the
      same traffic rate.  Since this test is not a throughput test, the
      rate should not be greater than 90% of line rate.  The IP PDUs
      MUST be encapsulated in AAL5.




Dunn & Martin                Informational                     [Page 75]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of CRC errored cells received per VCC on the
      test device for all VCCs.

  Reporting Format:

     The results of the CRC-ER/Bursty Mixed Load/Maximum VCCs test
     SHOULD be reported in a form of text and graph.

     The text results SHOULD display the numerical values of the CRC-
     ER.  The values given SHOULD include: time period of test in s,
     test VPI/VCI value, total number of cells transmitted and received
     on the given VPI/VCI during the test in positive integers, and the
     CRC-ER for the entire test.

     The graph results SHOULD display the CRC Error ratio values.
     There will be (Max number of VCCs/10) graphs, with 10 VCCs
     indicated on each graph.  The x-coordinate SHOULD be the test run
     time in either seconds, minutes or days depending on the total
     length of the test.  The x-coordinate time SHOULD be configurable.
     The y-coordinate SHOULD be the CRC-ER for each VCC.  There SHOULD
     be no more than 10 curves on each graph, one curve indicated and
     labeled for each VCC.  The integration time per point MUST be
     indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.6. Cell Transfer Delay (CTD)

3.2.6.1. Test Setup

  The cell transfer delay measurements assume that both the transmitter
  and receiver timestamp information is synchronized.  Synchronization
  SHOULD be achieved by supplying a common clock signal (minimum of 100
  Mhz or 10 ns resolution) to both the transmitter and receiver.  The
  maximum timestamp values MUST be recorded to ensure synchronization
  in the case of counter rollover.  The cell transfer delay
  measurements SHOULD utilize the O.191 cell (ITUT-O.191) encapsulated
  in a valid IP packet.  If the O.191 cell is not available, a test
  cell encapsulated in a valid IP packet MAY be used.  The test cell



Dunn & Martin                Informational                     [Page 76]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  MUST contain a transmit timestamp which can be correlated with a
  receive timestamp.  A description of the test cell MUST be included
  in the test results.  The description MUST include the timestamp
  length (in bits), counter rollover value, and the timestamp accuracy
  (in ns).

3.2.6.2. CTD/Steady Load/One VCC

  Objective: To determine the SUT variation in cell transfer delay with
  one VCC as defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with one VCC.  The VCC SHOULD
      contain one VPI/VCI.  The VCC MUST be configured as either a CBR,
      VBR, or UBR connection.  The VPI/VCI MUST not be one of the
      reserved ATM signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing timestamps at a
      specific constant rate through the SUT via the defined test VCC.
      Since this test is not a throughput test, the rate should not be
      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device.

  Reporting Format:

     The results of the CTD/Steady Load/One VCC test SHOULD be reported
     in a form of text, graph, and histogram.

     The text results SHOULD display the numerical values of the CTD.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, minimum,
     maximum, and mean CTD during the test in us.

     The graph results SHOULD display the cell transfer delay values.
     The x-coordinate SHOULD be the test run time in either seconds,



Dunn & Martin                Informational                     [Page 77]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the cell transfer delay in us.  The integration time per point
     MUST be indicated.

     The histogram results SHOULD display the cell transfer delay.  The
     x-coordinate SHOULD be the cell transfer delay in us with at least
     256 bins.  The y-coordinate SHOULD be the number of cells observed
     in each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     bearer class of the created VCC MUST also be indicated.

3.2.6.3. CTD/Steady Load/Twelve VCCs

  Objective: To determine the SUT variation in cell transfer delay with
  twelve VCCs as defined in RFC 2761 "Terminology for ATM
  Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCCs, using 1 VPI
      and 12 VCIs.  The VCC's MUST be configured as either a CBR, VBR,
      or UBR connection.  The VPI/VCIs MUST not be one of the reserved
      ATM signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing timestamps at a
      specific constant rate through the SUT via the defined test VCCs.
      All of the VPI/VCI pairs will generate traffic at the same
      traffic rate.  Since this test is not a throughput test, the rate
      should not be greater than 90% of line rate.  The IP PDUs MUST be
      encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device for all VCCs.






Dunn & Martin                Informational                     [Page 78]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  Reporting Format:

     The results of the CTD/Steady Load/Twelve VCCs test SHOULD be
     reported in a form of text, graph, and histograms.

     The text results SHOULD display the numerical values of the CTD.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI values, total number of cells transmitted and received on
     each VCC during the test in positive integers, maximum and minimum
     CTD on each VCC during the test in us, and mean CTD on each VCC in
     us.

     The graph results SHOULD display the cell transfer delay values.
     The x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the cell transfer delay for each VCC in ms.  There SHOULD be 12
     curves on the graph, one curves indicated and labeled for each
     VCC.  The integration time per point MUST be indicated.

     The histograms SHOULD display the cell transfer delay.  There will
     be one histogram for each VCC.  The x-coordinate SHOULD be the
     cell transfer delay in us with at least 256 bins.  The y-
     coordinate SHOULD be the number of cells observed in each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     bearer class of the created VCC MUST also be indicated.

3.2.6.4. CTD/Steady Load/Maximum VCCs

  Objective: To determine the SUT variation in cell transfer delay with
  the maximum number VCCs supported on the SUT as defined in RFC 2761
  "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with the maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per
      VPI.  The VCC's MUST be configured as either a CBR, VBR, or UBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).




Dunn & Martin                Informational                     [Page 79]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  3)  Send a specific number of IP packets containing timestamps at a
      specific constant rate through the SUT via the defined test VCCs.
      All of the VPI/VCI pairs will generate traffic at the same
      traffic rate.  Since this test is not a throughput test, the rate
      should not be greater than 90% of line rate.  The IP PDUs MUST be
      encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device for all VCCs.

  Reporting Format:

     The results of the CTD/Steady Load/Maximum VCCs test SHOULD be
     reported in a form of text, graphs, and histograms.

     The text results SHOULD display the numerical values of the CTD.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI values, total number of cells transmitted and received on
     each VCC during the test in positive integers, maximum and minimum
     CTD on each VCC during the test in us, and mean CTD on each VCC in
     us.

     The graph results SHOULD display the cell transfer delay values.
     There will be (Max number of VCCs/10) graphs, with 10 VCCs
     indicated on each graph.  The x-coordinate SHOULD be the test run
     time in either seconds, minutes or days depending on the total
     length of the test.  The x-coordinate time SHOULD be configurable.
     The y-coordinate SHOULD be the cell transfer delay for each VCC in
     us.  There SHOULD be no more than 10 curves on each graph, one
     curve indicated and labeled for each VCC.  The integration time
     per point MUST be indicated.

     The histograms SHOULD display the cell transfer delay.  There will
     be one histogram for each VCC.  The x-coordinate SHOULD be the
     cell transfer delay in us with at least 256 bins.  The y-
     coordinate SHOULD be the number of cells observed in each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     bearer class of the created VCC MUST also be indicated.





Dunn & Martin                Informational                     [Page 80]

RFC 3116            Methodology for ATM Benchmarking           June 2001


3.2.6.5. CTD/Bursty VBR Load/One VCC

  Objective: To determine the SUT variation in cell transfer delay with
  one VCC as defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with one VCC.  The VCC SHOULD
      contain one VPI/VCI.  The VCC MUST be configured as either a CBR
      or VBR connection.  The VPI/VCI MUST not be one of the reserved
      ATM signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing timestamps at a
      specific VBR through the SUT via the defined test VCC.  Since
      this test is not a throughput test, the rate should not be
      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device.

  Reporting Format:

     The results of the CTD/Bursty VBR Load/One VCC test SHOULD be
     reported in a form of text, graph, and histogram.

     The text results SHOULD display the numerical values of the CTD.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, minimum,
     maximum, and mean CTD during the test in us.

     The graph results SHOULD display the cell transfer delay values.
     The x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the cell transfer delay in us.  The integration time per point
     MUST be indicated.





Dunn & Martin                Informational                     [Page 81]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The histogram results SHOULD display the cell transfer delay.  The
     x-coordinate SHOULD be the cell transfer delay in us with at least
     256 bins.  The y-coordinate SHOULD be the number of cells observed
     in each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.6.6. CTD/Bursty VBR Load/Twelve VCCs

  Objective: To determine the SUT variation in cell transfer delay with
  twelve VCCs as defined in RFC 2761 "Terminology for ATM
  Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCCs, using 1 VPI
      and 12 VCIs.  The VCC's MUST be configured as either a CBR or VBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing timestamps at a
      specific VBR through the SUT via the defined test VCCs.  All of
      the VPI/VCI pairs will generate traffic at the same traffic rate.
      Since this test is not a throughput test, the rate should not be
      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device for all VCCs.

  Reporting Format:

     The results of the CTD/Bursty VBR Load/Twelve VCCs test SHOULD be
     reported in a form of text, graph, and histograms.





Dunn & Martin                Informational                     [Page 82]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The text results SHOULD display the numerical values of the CTD.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI values, total number of cells transmitted and received on
     each VCC during the test in positive integers, maximum and minimum
     CTD on each VCC during the test in us, and mean CTD on each VCC in
     us.

     The graph results SHOULD display the cell transfer delay values.
     The x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the cell transfer delay for each VCC in ms.  There SHOULD be 12
     curves on the graph, one curves indicated and labeled for each
     VCC.  The integration time per point MUST be indicated.

     The histograms SHOULD display the cell transfer delay.  There will
     be one histogram for each VCC.  The x-coordinate SHOULD be the
     cell transfer delay in us with at least 256 bins.  The y-
     coordinate SHOULD be the number of cells observed in each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.6.7. CTD/Bursty VBR Load/Maximum VCCs

  Objective: To determine the SUT variation in cell transfer delay with
  the maximum number VCCs supported on the SUT as defined in RFC 2761
  "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with the maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per
      VPI.  The VCC's MUST be configured as either a CBR or VBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing timestamps at a
      specific VBR through the SUT via the defined test VCCs.  All of
      the VPI/VCI pairs will generate traffic at the same traffic rate.
      Since this test is not a throughput test, the rate should not be



Dunn & Martin                Informational                     [Page 83]

RFC 3116            Methodology for ATM Benchmarking           June 2001


      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device for all VCCs.

  Reporting Format:

     The results of the CTD/Bursty VBR Load/Maximum VCCs test SHOULD be
     reported in a form of text, graphs, and histograms.

     The text results SHOULD display the numerical values of the CTD.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI values, total number of cells transmitted and received on
     each VCC during the test in positive integers, maximum and minimum
     CTD on each VCC during the test in us, and mean CTD on each VCC in
     us.

     The graph results SHOULD display the cell transfer delay values.
     There will be (Max number of VCCs/10) graphs, with 10 VCCs
     indicated on each graph.  The x-coordinate SHOULD be the test run
     time in either seconds, minutes or days depending on the total
     length of the test.  The x-coordinate time SHOULD be configurable.
     The y-coordinate SHOULD be the cell transfer delay for each VCC in
     us.  There SHOULD be no more than 10 curves on each graph, one
     curve indicated and labeled for each VCC.  The integration time
     per point MUST be indicated.

     The histograms SHOULD display the cell transfer delay.  There will
     be one histogram for each VCC.  The x-coordinate SHOULD be the
     cell transfer delay in us with at least 256 bins.  The y-
     coordinate SHOULD be the number of cells observed in each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.








Dunn & Martin                Informational                     [Page 84]

RFC 3116            Methodology for ATM Benchmarking           June 2001


3.2.6.8. CTD/Bursty UBR Load/One VCC

  Objective: To determine the SUT variation in cell transfer delay with
  one VCC as defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with one VCC.  The VCC SHOULD
      contain one VPI/VCI.  The VCC MUST be configured as a UBR
      connection.  The VPI/VCI MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing timestamps at a
      specific UBR through the SUT via the defined test VCC.  Since
      this test is not a throughput test, the rate should not be
      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device.

  Reporting Format:

     The results of the CTD/Bursty UBR Load/One VCC test SHOULD be
     reported in a form of text, graph, and histogram.

     The text results SHOULD display the numerical values of the CTD.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, minimum,
     maximum, and mean CTD during the test in us.

     The graph results SHOULD display the cell transfer delay values.
     The x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the cell transfer delay in us.  The integration time per point
     MUST be indicated.





Dunn & Martin                Informational                     [Page 85]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The histogram results SHOULD display the cell transfer delay.  The
     x-coordinate SHOULD be the cell transfer delay in us with at least
     256 bins.  The y-coordinate SHOULD be the number of cells observed
     in each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     bearer class of the created VCC MUST also be indicated.

3.2.6.9. CTD/Bursty UBR Load/Twelve VCCs

  Objective: To determine the SUT variation in cell transfer delay with
  twelve VCCs as defined in RFC 2761 "Terminology for ATM
  Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCCs, using 1 VPI
      and 12 VCIs.  The VCC's MUST be configured as a UBR connection.
      The VPI/VCIs MUST not be one of the reserved ATM signaling
      channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing timestamps at a
      specific UBR through the SUT via the defined test VCCs.  All of
      the VPI/VCI pairs will generate traffic at the same traffic rate.
      Since this test is not a throughput test, the rate should not be
      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device for all VCCs.

  Reporting Format:

     The results of the CTD/Bursty UBR Load/Twelve VCCs test SHOULD be
     reported in a form of text, graph, and histograms.

     The text results SHOULD display the numerical values of the CTD.
     The values given SHOULD include: time period of test in s, test



Dunn & Martin                Informational                     [Page 86]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     VPI/VCI values, total number of cells transmitted and received on
     each VCC during the test in positive integers, maximum and minimum
     CTD on each VCC during the test in us, and mean CTD on each VCC in
     us.

     The graph results SHOULD display the cell transfer delay values.
     The x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the cell transfer delay for each VCC in ms.  There SHOULD be 12
     curves on the graph, one curves indicated and labeled for each
     VCC.  The integration time per point MUST be indicated.

     The histograms SHOULD display the cell transfer delay.  There will
     be one histogram for each VCC.  The x-coordinate SHOULD be the
     cell transfer delay in us with at least 256 bins.  The y-
     coordinate SHOULD be the number of cells observed in each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     bearer class of the created VCC MUST also be indicated.

3.2.6.10. CTD/Bursty UBR Load/Maximum VCCs

  Objective: To determine the SUT variation in cell transfer delay with
  the maximum number VCCs supported on the SUT as defined in RFC 2761
  "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with the maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per
      VPI.  The VCC MUST be configured as a UBR connection.  The
      VPI/VCIs MUST not be one of the reserved ATM signaling channels
      (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing timestamps at a
      specific UBR through the SUT via the defined test VCCs.  All of
      the VPI/VCI pairs will generate traffic at the same traffic rate.
      Since this test is not a throughput test, the rate should not be
      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.




Dunn & Martin                Informational                     [Page 87]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device for all VCCs.

  Reporting Format:

     The results of the CTD/Bursty UBR Load/Maximum VCCs test SHOULD be
     reported in a form of text, graphs, and histograms.

     The text results SHOULD display the numerical values of the CTD.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI values, total number of cells transmitted and received on
     each VCC during the test in positive integers, maximum and minimum
     CTD on each VCC during the test in us, and mean CTD on each VCC in
     us.

     The graph results SHOULD display the cell transfer delay values.
     There will be (Max number of VCCs/10) graphs, with 10 VCCs
     indicated on each graph.  The x-coordinate SHOULD be the test run
     time in either seconds, minutes or days depending on the total
     length of the test.  The x-coordinate time SHOULD be configurable.
     The y-coordinate SHOULD be the cell transfer delay for each VCC in
     us.  There SHOULD be no more than 10 curves on each graph, one
     curve indicated and labeled for each VCC.  The integration time
     per point MUST be indicated.

     The histograms SHOULD display the cell transfer delay.  There will
     be one histogram for each VCC.  The x-coordinate SHOULD be the
     cell transfer delay in us with at least 256 bins.  The y-
     coordinate SHOULD be the number of cells observed in each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     bearer class of the created VCC MUST also be indicated.

3.2.6.11. CTD/Mixed Load/Three VCC's

  Objective: To determine the SUT variation in cell transfer delay with
  three VCC's as defined in RFC 2761 "Terminology for ATM
  Benchmarking".






Dunn & Martin                Informational                     [Page 88]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with three VCC's.  Each VCC
      MUST be defined as a different Bearer class: one CBR, one UBR and
      one VBR.  Each VCC SHOULD contain one VPI/VCI.  The VPI/VCI MUST
      not be one of the reserved ATM signaling channels (e.g., [0,5],
      [0,16]).

  3)  Send a specific number of IP packets containing timestamps
      through the SUT via the defined test VCCs.  Each generated VCC
      stream MUST match the corresponding VCC Bearer class.  All of the
      VPI/VCI pairs will generate traffic at the same traffic rate.
      Since this test is not a throughput test, the rate should not be
      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device for all VCC's.

  Reporting Format:

     The results of the CTD/Mixed Load/Three VCC test SHOULD be
     reported in a form of text, graph, and histogram.

     The text results SHOULD display the numerical values of the CTD.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI value, total number of cells transmitted and received on
     the given VPI/VCI during the test in positive integers, minimum,
     maximum, and mean CTD during the test in us.

     The graph results SHOULD display the cell transfer delay values.
     The x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the cell transfer delay in us.  The integration time per point
     MUST be indicated.







Dunn & Martin                Informational                     [Page 89]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The histogram results SHOULD display the cell transfer delay.  The
     x-coordinate SHOULD be the cell transfer delay in us with at least
     256 bins.  The y-coordinate SHOULD be the number of cells observed
     in each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.6.12. CTD/Mixed Load/Twelve VCCs

  Objective: To determine the SUT variation in cell transfer delay with
  twelve VCCs as defined in RFC 2761 "Terminology for ATM
  Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCC's.  Each VCC
      MUST be defined as one of the Bearer classes for a total of four
      CBR, four UBR and four VBR VCC's.  Each VCC SHOULD contain one
      VPI/VCI.  The VPI/VCI MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing timestamps
      through the SUT via the defined test VCCs.  Each generated VCC
      stream MUST match the corresponding VCC Bearer class.  All of the
      VPI/VCI pairs will generate traffic at the same traffic rate.
      Since this test is not a throughput test, the rate should not be
      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device for all VCCs.

  Reporting Format:

     The results of the CTD/Mixed Load/Twelve VCCs test SHOULD be
     reported in a form of text, graph, and histograms.



Dunn & Martin                Informational                     [Page 90]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The text results SHOULD display the numerical values of the CTD.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI values, total number of cells transmitted and received on
     each VCC during the test in positive integers, maximum and minimum
     CTD on each VCC during the test in us, and mean CTD on each VCC in
     us.

     The graph results SHOULD display the cell transfer delay values.
     The x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the cell transfer delay for each VCC in ms.  There SHOULD be 12
     curves on the graph, one curves indicated and labeled for each
     VCC.  The integration time per point MUST be indicated.

     The histograms SHOULD display the cell transfer delay.  There will
     be one histogram for each VCC.  The x-coordinate SHOULD be the
     cell transfer delay in us with at least 256 bins.  The y-
     coordinate SHOULD be the number of cells observed in each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.

3.2.6.13. CTD/Mixed Load/Maximum VCCs

  Objective: To determine the SUT variation in cell transfer delay with
  the maximum number VCCs supported on the SUT as defined in RFC 2761
  "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per
      VPI.  Each VCC MUST be defined as one of the Bearer classes for a
      total of (max VCC/3) CBR, (max VCC/3) UBR and (max VCC/3) VBR
      VCC's.  If the maximum number of VCC's is not divisible by 3, the
      total for each bearer class MUST be within 3 VCC's of each other.
      The VPI/VCI MUST not be one of the reserved ATM signaling
      channels (e.g., [0,5], [0,16]).





Dunn & Martin                Informational                     [Page 91]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  3)  Send a specific number of IP packets containing timestamps
      through the SUT via the defined test VCCs.  Each generated VCC
      stream MUST match the corresponding VCC Bearer class.  All of the
      VPI/VCI pairs will generate traffic at the same traffic rate.
      Since this test is not a throughput test, the rate should not be
      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the packets timestamps at the transmitter and receiver
      ends of the test device for all VCCs.

  Reporting Format:

     The results of the CTD/Mixed Load/Maximum VCCs test SHOULD be
     reported in a form of text, graphs, and histograms.

     The text results SHOULD display the numerical values of the CTD.
     The values given SHOULD include: time period of test in s, test
     VPI/VCI values, total number of cells transmitted and received on
     each VCC during the test in positive integers, maximum and minimum
     CTD on each VCC during the test in us, and mean CTD on each VCC in
     us.

     The graph results SHOULD display the cell transfer delay values.
     There will be (Max number of VCCs/10) graphs, with 10 VCCs
     indicated on each graph.  The x-coordinate SHOULD be the test run
     time in either seconds, minutes or days depending on the total
     length of the test.  The x-coordinate time SHOULD be configurable.
     The y-coordinate SHOULD be the cell transfer delay for each VCC in
     us.  There SHOULD be no more than 10 curves on each graph, one
     curve indicated and labeled for each VCC.  The integration time
     per point MUST be indicated.

     The histograms SHOULD display the cell transfer delay.  There will
     be one histogram for each VCC.  The x-coordinate SHOULD be the
     cell transfer delay in us with at least 256 bins.  The y-
     coordinate SHOULD be the number of cells observed in each bin.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST also be indicated.



Dunn & Martin                Informational                     [Page 92]

RFC 3116            Methodology for ATM Benchmarking           June 2001


3.3. ATM Adaptation Layer (AAL) Type 5 (AAL5)

3.3.1. IP Packet Loss due to AAL5 Re-assembly Errors

  Objective: To determine if the SUT will drop IP packets due AAL5 Re-
  assembly Errors as defined in RFC 2761 "Terminology for ATM
  Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the uni-directional
      configuration.

  2)  Send a specific number of cells at a specific rate through the
      SUT.  Since this test is not a throughput test, the rate should
      not be greater than 90% of line rate.  The cell payload SHOULD
      contain valid IP PDUs.  The IP PDUs MUST be encapsulated in AAL5.

  3)  Count the cells that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  4)  Inject one error in the first bit of the AAL5 payload.  Verify
      that the SUT does not drop any AAL5 PDU's.

  5)  Discontinue the AAL5 payload error.

  6)  Inject one error in the first bit of the AAL5 header for 4
      consecutive IP PDUs in every 6 IP PDUs.  Verify that the SUT does
      drop the AAL5 PDU's.

  7)  Discontinue the AAL5 payload error.

  Reporting Format:

     The results of the AAL5 PDU Loss due to AAL5 PDU errors test
     SHOULD be reported in a form of a table.  The rows SHOULD be
     labeled single error, one error per second, and four consecutive
     errors every 6 IP PDUs.  The columns SHOULD be labeled AAL5 PDU
     loss and number of PDU's lost.  The elements of column 1 SHOULD be
     either True or False, indicating whether the particular condition
     was observed for each test.  The elements of column 2 SHOULD be
     non-negative integers.

     The table MUST also indicate the traffic rate in IP PDUs per
     second as generated by the test device.




Dunn & Martin                Informational                     [Page 93]

RFC 3116            Methodology for ATM Benchmarking           June 2001


3.3.2. AAL5 Reassembly Time.

  Objective: To determine the SUT AAL5 Reassembly Time as defined in
  RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the uni-directional
      configuration.

  2)  Send a specific number of IP packets at a specific rate through
      the SUT.  Since this test is not a throughput test, the rate
      should not be greater than 90% of line rate.  The IP PDUs MUST be
      encapsulated in AAL5.  The AAL5 PDU size is 65535 octets or 1365
      ATM cells.

  3)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  4)  Given an AAL5 reassembly timer of 'x' seconds, where 'x' is the
      actual value of the AAL5 reassembly timer on the SUT, sent
      traffic at 1365 cells per 'x' seconds.  The expected results are
      that no AAL5 PDU's will be dropped.

  5)  Send traffic at 1360 cells per 'x' seconds.  The expected results
      are that all AAL5 PDU's will be dropped.

  Reporting Format:

     The results of the IP packet loss due to AAL5 reassembly timeout
     test SHOULD be reported in a form of a table.  The rows SHOULD be
     labeled 1365 cells per 'x' seconds and 1360 cells per 'x' seconds.
     The columns SHOULD be labeled packet loss and number of packets
     lost.  The elements of column 1 SHOULD be either True or False,
     indicating whether the particular condition was observed for each
     test.  The elements of column 2 SHOULD be non-negative integers.

     The table MUST also indicate the packet size in octets and traffic
     rate in packets per second as generated by the test device,
     including the value of









Dunn & Martin                Informational                     [Page 94]

RFC 3116            Methodology for ATM Benchmarking           June 2001


3.3.3. AAL5 CRC Error Ratio.

3.3.3.1. Test Setup

  The AAL5 CRC error ratio measurements assume that both the
  transmitter and receiver payload information is synchronized.
  Synchronization MUST be achieved by supplying a known bit pattern to
  both the transmitter and receiver.  If this bit pattern is longer
  than the packet size, the receiver MUST synchronize with the
  transmitter before tests can be run.

3.3.3.2. AAL5-CRC-ER/Steady Load/One VCC

  Objective: To determine the SUT ratio of AAL5 CRC PDU errors on one
  VCC in a transmission in relation to the total AAL5 PDU's sent as
  defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with one VCC.  The VCC SHOULD
      contain one VPI/VCI.  The VCC MUST be configured as either a CBR,
      VBR, or UBR connection.  The VPI/VCI MUST not be one of the
      reserved ATM signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a constant rate through the SUT via the
      defined test VCC.  Since this test is not a throughput test, the
      rate should not be greater than 90% of line rate.  The IP PDUs
      MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the number of AAL5 CRC errors at the receiver end of the
      test device.

  Reporting Format:

     The results of the AAL5-CRC-ER/Steady Load/One VCC test SHOULD be
     reported in a form of text and graph.






Dunn & Martin                Informational                     [Page 95]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The text results SHOULD display the numerical values of the AAL5-
     CRC-ER.  The values given SHOULD include: time period of test in
     s, test VPI/VCI value, total number of AAL5 PDU's transmitted and
     received on the given VPI/VCI during the test in positive
     integers, and the AAL5-CRC-ER for the entire test.

     The graph results SHOULD display the AAL5 CRC error ratio values.
     The x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the AAL5-CRC-ER.  The integration time per point MUST be
     indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST be indicated.  The generated bit pattern MUST
     also be indicated.

3.3.3.3. AAL5-CRC-ER/Steady Load/Twelve VCCs

  Objective: To determine the SUT ratio of AAL5 CRC PDU errors on
  twelve VCC's in a transmission in relation to the total AAL5 PDU's
  sent as defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCCs, using 1 VPI
      and 12 VCIs.  The VCC's MUST be configured as either a CBR, VBR,
      or UBR connection.  The VPI/VCIs MUST not be one of the reserved
      ATM signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a constant rate through the SUT via the
      defined test VCCs.  All of the VPI/VCI pairs will generate
      traffic at the same traffic rate.

      Since this test is not a throughput test, the rate should not be
      greater than 90% of line rate.  The IP PDUs MUST be encapsulated
      in AAL5.







Dunn & Martin                Informational                     [Page 96]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of AAL5 CRC errors at the receiver end of the
      test device for all VCCs.

  Reporting Format:

     The results of the AAL5-CRC-ER/Steady Load/Twelve VCCs test SHOULD
     be reported in a form of text and graph.

     The text results SHOULD display the numerical values of the AAL5-
     CRC-ER.  The values given SHOULD include: time period of test in
     s, test VPI/VCI value, total number of AAL5 PDU's transmitted and
     received on the given VPI/VCI during the test in positive
     integers, and the AAL5-CRC-ER for the entire test.

     The graph results SHOULD display the AAL5 CRC error ratio values.
     The x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the AAL5-CRC-ER for each VCC.  There should be 12 curves on the
     graph, on curve indicated and labeled for each VCC.  The
     integration time per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST be indicated.  The generated bit pattern MUST
     also be indicated.

3.3.3.4. AAL5-CRC-ER/Steady Load/Maximum VCCs

  Objective: To determine the SUT ratio of AAL5 CRC PDU errors with the
  maximum number VCCs supported on the SUT in a transmission in
  relation to the total AAL5 PDU's sent as defined in RFC 2761
  "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with the maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs



Dunn & Martin                Informational                     [Page 97]

RFC 3116            Methodology for ATM Benchmarking           June 2001


      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per
      VPI.  The VCC's MUST be configured as either a CBR, VBR, or UBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a constant rate through the SUT via the
      defined test VCCs.  All of the VPI/VCI pairs will generate
      traffic at the same traffic rate.  Since this test is not a
      throughput test, the rate should not be greater than 90% of line
      rate.  The IP PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of AAL5 CRC errors at the receiver end of the
      test device for all VCCs.

  Reporting Format:

     The results of the AAL5-CRC-ER/Steady Load/Maximum VCCs test
     SHOULD be reported in a form of text and graph.

     The text results SHOULD display the numerical values of the AAL5-
     CRC-ER.  The values given SHOULD include: time period of test in
     s, test VPI/VCI value, total number of AAL5 PDU's transmitted and
     received on the given VPI/VCI during the test in positive
     integers, and the AAL5-CRC-ER for the entire test.

     The graph results SHOULD display the AAL5 CRC error ratio values.
     There will be (Max number of VCCs/10) graphs, with 10 VCCs
     indicated on each graph.  The x-coordinate SHOULD be the test run
     time in either seconds, minutes or days depending on the total
     length of the test.  The x-coordinate time SHOULD be configurable.
     The y-coordinate SHOULD be the AAL5-CRC-ER for each VCC.  There
     SHOULD be no more than 10 curves on each graph, one curve
     indicated and labeled for each VCC.  The integration time per
     point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST be indicated.  The generated bit pattern MUST
     also be indicated.




Dunn & Martin                Informational                     [Page 98]

RFC 3116            Methodology for ATM Benchmarking           June 2001


3.3.3.5. AAL5-CRC-ER/Bursty VBR Load/One VCC

  Objective: To determine the SUT ratio of AAL5 CRC PDU errors on one
  VCC in a transmission in relation to the total AAL5 PDU's sent as
  defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with one VCC.  The VCC SHOULD
      contain one VPI/VCI.  The VCC MUST be configured as either a CBR
      or VBR connection.  The VPI/VCI MUST not be one of the reserved
      ATM signaling channels (e.g., [0,5], [0,16]).  The PCR, SCR, and
      MBS must be configured using one of the specified traffic
      descriptors.

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific VBR rate through the SUT via
      the defined test VCC.  Since this test is not a throughput test,
      the rate should not be greater than 90% of line rate.  The IP
      PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the number of AAL5 CRC errors at the receiver end of the
      test device.

  Reporting Format:

     The results of the AAL5-CRC-ER/Bursty VBR Load/One VCC test SHOULD
     be reported in a form of text and graph.

     The text results SHOULD display the numerical values of the AAL5-
     CRC-ER.  The values given SHOULD include: time period of test in
     s, test VPI/VCI value, total number of AAL5 PDU's transmitted and
     received on the given VPI/VCI during the test in positive
     integers, and the AAL5-CRC-ER for the entire test.

     The graph results SHOULD display the AAL5 CRC error ratio values.
     The x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The





Dunn & Martin                Informational                     [Page 99]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the AAL5-CRC-ER.  The integration time per point MUST be
     indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST be indicated.  The generated bit pattern MUST
     also be indicated.

3.3.3.6. AAL5-CRC-ER/Bursty VBR Load/Twelve VCCs

  Objective: To determine the SUT ratio of AAL5 CRC PDU errors on
  twelve VCC's in a transmission in relation to the total AAL5 PDU's
  sent as defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCCs, using 1 VPI
      and 12 VCIs.  The VCC's MUST be configured as either a CBR or VBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).  The PCR, SCR, and MBS
      must be configured using one of the specified traffic
      descriptors.

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific VBR rate through the SUT via
      the defined test VCCs.  All of the VPI/VCI pairs will generate
      traffic at the same traffic rate.  Since this test is not a
      throughput test, the rate should not be greater than 90% of line
      rate.  The PCR, SCR, and MBS must be indicated.  The IP PDUs MUST
      be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of AAL5 CRC errors at the receiver end of the
      test device for all VCCs.







Dunn & Martin                Informational                    [Page 100]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  Reporting Format:

     The results of the AAL5-CRC-ER/Bursty VBR Load/Twelve VCCs test
     SHOULD be reported in a form of text and graph.

     The text results SHOULD display the numerical values of the AAL5-
     CRC-ER.  The values given SHOULD include: time period of test in
     s, test VPI/VCI value, total number of AAL5 PDU's transmitted and
     received on the given VPI/VCI during the test in positive
     integers, and the AAL5-CRC-ER for the entire test.

     The graph results SHOULD display the AAL5 CRC error ratio values.
     The x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the AAL5-CRC-ER for each VCC.  There should be 12 curves on the
     graph, on curve indicated and labeled for each VCC.  The
     integration time per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST be indicated.  The generated bit pattern MUST
     also be indicated.

3.3.3.7. AAL5-CRC-ER/Bursty VBR Load/Maximum VCCs

  Objective: To determine the SUT ratio of AAL5 CRC PDU errors with the
  maximum number VCCs supported on the SUT in a transmission in
  relation to the total AAL5 PDU's sent as defined in RFC 2761
  "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with the maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per
      VPI.  The VCC's MUST be configured as either a CBR or VBR
      connection.  The VPI/VCIs MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).  The PCR, SCR, and MBS
      must be configured using one of the specified traffic
      descriptors.





Dunn & Martin                Informational                    [Page 101]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  3)  Send a specific number of IP packets containing one of the
      specified bit patterns at a specific VBR rate through the SUT via
      the defined test VCCs.  All of the VPI/VCI pairs will generate
      traffic at the same traffic rate.  Since this test is not a
      throughput test, the rate should not be greater than 90% of line
      rate.  The IP PDUs MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of AAL5 CRC errors at the receiver end of the
      test device for all VCCs.

  Reporting Format:

     The results of the AAL5-CRC-ER/Bursty VBR Load/Maximum VCCs test
     SHOULD be reported in a form of text and graph.

     The text results SHOULD display the numerical values of the AAL5-
     CRC-ER.  The values given SHOULD include: time period of test in
     s, test VPI/VCI value, total number of AAL5 PDU's transmitted and
     received on the given VPI/VCI during the test in positive
     integers, and the AAL5-CRC-ER for the entire test.

     The graph results SHOULD display the AAL5 CRC error ratio values.
     There will be (Max number of VCCs/10) graphs, with 10 VCCs
     indicated on each graph.  The x-coordinate SHOULD be the test run
     time in either seconds, minutes or days depending on the total
     length of the test.  The x-coordinate time SHOULD be configurable.
     The y-coordinate SHOULD be the AAL5-CRC-ER for each VCC.  There
     SHOULD be no more than 10 curves on each graph, one curve
     indicated and labeled for each VCC.  The integration time per
     point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST be indicated.  The generated bit pattern MUST
     also be indicated.

3.3.3.8. AAL5-CRC-ER/Mixed Load/Three VCC's

  Objective: To determine the SUT ratio of AAL5 CRC PDU errors on three
  VCC's in a transmission in relation to the total AAL5 PDU's sent as
  defined in RFC 2761 "Terminology for ATM Benchmarking".



Dunn & Martin                Informational                    [Page 102]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with three VCC's.  Each VCC
      MUST be defined as a different Bearer class; one CBR, one UBR and
      one VBR.  Each VCC SHOULD contain one VPI/VCI.  The VPI/VCI MUST
      not be one of the reserved ATM signaling channels (e.g., [0,5],
      [0,16]).

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns through the SUT via the defined test VCCs.
      Each generated VCC stream MUST match the corresponding VCC Bearer
      class.  All of the VPI/VCI pairs will generate traffic at the
      same traffic rate.  Since this test is not a throughput test, the
      rate should not be greater than 90% of line rate.  The IP PDUs
      MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT to verify
      connectivity and load.  If the count on the test device is the
      same on the SUT, continue the test; else lower the test device
      traffic rate until the counts are the same.

  5)  Record the number of AAL5 CRC errors at the receiver end of the
      test device for all VCCs.

  Reporting Format:

     The results of the AAL5-CRC-ER/Bursty Mixed Load/Three VCCs test
     SHOULD be reported in a form of text and graph.

     The text results SHOULD display the numerical values of the AAL5-
     CRC-ER.  The values given SHOULD include: time period of test in
     s, test VPI/VCI value, total number of AAL5 PDU's transmitted and
     received on the given VPI/VCI during the test in positive
     integers, and the AAL5-CRC-ER for the entire test.

     The graph results SHOULD display the AAL5 CRC error ratio values.
     The x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the AAL5-CRC-ER for each VCC.  There should be 12 curves on the
     graph, on curve indicated and labeled for each VCC.  The
     integration time per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the



Dunn & Martin                Informational                    [Page 103]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST be indicated.  The generated bit pattern MUST
     also be indicated.

3.3.3.9. AAL5-CRC-ER/Mixed Load/Twelve VCCs

  Objective: To determine the SUT ratio of AAL5 CRC PDU errors on
  twelve VCC's in a transmission in relation to the total AAL5 PDU's
  sent as defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with twelve VCC's.  Each VCC
      MUST be defined as one of the Bearer classes for a total of four
      CBR, four UBR and four VBR VCC's.  Each VCC SHOULD contain one
      VPI/VCI.  The VPI/VCI MUST not be one of the reserved ATM
      signaling channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns through the SUT via the defined test VCCs.
      Each generated VCC stream MUST match the corresponding VCC Bearer
      class.  All of the VPI/VCI pairs will generate traffic at the
      same traffic rate.  Since this test is not a throughput test, the
      rate should not be greater than 90% of line rate.  The IP PDUs
      MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.

  5)  Record the number of AAL5 CRC errors at the receiver end of the
      test device for all VCCs.

  Reporting Format:

     The results of the AAL5-CRC-ER/Bursty Mixed Load/Twelve VCCs test
     SHOULD be reported in a form of text and graph.

     The text results SHOULD display the numerical values of the AAL5-
     CRC-ER.  The values given SHOULD include: time period of test in
     s, test VPI/VCI value, total number of AAL5 PDU's transmitted and
     received on the given VPI/VCI during the test in positive
     integers, and the AAL5-CRC-ER for the entire test.



Dunn & Martin                Informational                    [Page 104]

RFC 3116            Methodology for ATM Benchmarking           June 2001


     The graph results SHOULD display the AAL5 CRC error ratio values.
     The x-coordinate SHOULD be the test run time in either seconds,
     minutes or days depending on the total length of the test.  The
     x-coordinate time SHOULD be configurable.  The y-coordinate SHOULD
     be the AAL5-CRC-ER for each VCC.  There should be 12 curves on the
     graph, on curve indicated and labeled for each VCC.  The
     integration time per point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST be indicated.  The generated bit pattern MUST
     also be indicated.

3.3.3.10. AAL5-CRC-ER/Mixed Load/Maximum VCCs

  Objective: To determine the SUT ratio of AAL5 CRC PDU errors with the
  maximum number VCCs supported on the SUT in a transmission in
  relation to the total AAL5 PDU's sent as defined in RFC 2761
  "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Configure the SUT and test device with maximum number of VCCs
      supported on the SUT.  For example, if the maximum number of VCCs
      supported on the SUT is 1024, define 256 VPIs with 4 VCIs per
      VPI.  Each VCC MUST be defined as one of the Bearer classes for a
      total of (max VCC/3) CBR, (max VCC/3) UBR and (max VCC/3) VBR
      VCC's.  The VPI/VCI MUST not be one of the reserved ATM signaling
      channels (e.g., [0,5], [0,16]).

  3)  Send a specific number of IP packets containing one of the
      specified bit patterns through the SUT via the defined test VCCs.
      Each generated VCC stream MUST match the corresponding VCC Bearer
      class.  All of the VPI/VCI pairs will generate traffic at the
      same traffic rate.  Since this test is not a throughput test, the
      rate should not be greater than 90% of line rate.  The IP PDUs
      MUST be encapsulated in AAL5.

  4)  Count the IP packets that are transmitted by the SUT on all VCCs
      to verify connectivity and load.  If the count on the test device
      is the same on the SUT, continue the test; else lower the test
      device traffic rate until the counts are the same.




Dunn & Martin                Informational                    [Page 105]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  5)  Record the number of AAL5 CRC errors at the receiver end of the
      test device for all VCCs.

  Reporting Format:

     The results of the AAL5-CRC-ER/Bursty Mixed Load/Maximum VCCs test
     SHOULD be reported in a form of text and graph.

     The text results SHOULD display the numerical values of the AAL5-
     CRC-ER.  The values given SHOULD include: time period of test in
     s, test VPI/VCI value, total number of AAL5 PDU's transmitted and
     received on the given VPI/VCI during the test in positive
     integers, and the AAL5-CRC-ER for the entire test.

     The graph results SHOULD display the AAL5 CRC error ratio values.
     There will be (Max number of VCCs/10) graphs, with 10 VCCs
     indicated on each graph.  The x-coordinate SHOULD be the test run
     time in either seconds, minutes or days depending on the total
     length of the test.  The x-coordinate time SHOULD be configurable.
     The y-coordinate SHOULD be the AAL5-CRC-ER for each VCC.  There
     SHOULD be no more than 10 curves on each graph, one curve
     indicated and labeled for each VCC.  The integration time per
     point MUST be indicated.

     The results MUST also indicate the packet size in octets, traffic
     rate in packets per second, and bearer class as generated by the
     test device.  The VCC and VPI/VCI values MUST be indicated.  The
     PCR, SCR, and MBS MUST be indicated.  The bearer class of the
     created VCC MUST be indicated.  The generated bit pattern MUST
     also be indicated.

3.4. ATM Service: Signaling

3.4.1. CAC Denial Time and Connection Establishment Time

  Objective: To determine the CAC rejection time and Connection
  Establishment Time on the SUT as defined in RFC 2761 "Terminology for
  ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Create a UNI signaling setup message, as described in Appendix C,
      specifying a PCR which will not allow CAC to reject the call.





Dunn & Martin                Informational                    [Page 106]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  3)  Send the UNI signaling setup message.  Note the time the setup
      message was sent.  Verify that the SVC has been setup with the
      correct parameters.  Note the time the connect message was
      received

  4)  Create a UNI signaling setup message, as described in Appendix C,
      specifying a PCR which will allow CAC to reject the call.

  5)  Send the UNI signaling setup message.  Note the time the setup
      message was sent.  Verify that the SVC has been rejected with the
      correct cause code.  Note the time the release complete message
      was received.

  6)  Compute the rejection time as the difference between the time the
      release complete message was received and the time setup message
      was send.

  Reporting Format:

     The results of the CAC Denial Time and Connection Establishment
     Time tests SHOULD be reported in a form of a table.  The rows
     SHOULD be labeled call accepted and call rejected.  The columns
     SHOULD be labeled time setup sent, time response received, and
     correct response.  The elements of the columns 1 and 2 SHOULD be
     in seconds.  The elements of column 3 SHOULD be be either True or
     False, indicating whether the particular condition was observed
     for each test.

     The table MUST also indicate the packet size in octets and traffic
     rate in packets per second as generated by the test device.

3.4.2. Connection Teardown Time

  Objective: To determine the Connection Teardown Time on the SUT as
  defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Create a UNI signaling setup message, as described in Appendix C,
      specifying a PCR which will not allow CAC to reject the call.

  3)  Send the UNI signaling setup message.  Note the time the setup
      message was sent.  Verify that the SVC has been setup with the
      correct parameters.  Note the time the connect message was
      received



Dunn & Martin                Informational                    [Page 107]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  4)  Create a UNI signaling release message, as described in Appendix
      C, specifying a cause code of normal call clearing.

  5)  Send the UNI signaling release message.  Note the time the
      release message was sent.  Verify that the SVC has been
      terminated with the correct cause code.  Note the time the
      release complete message was received.

  6)  Compute the release time as the difference between the time the
      release complete message was received and the time release
      message was send.

  Reporting Format:

     The results of the Connection Teardown Time tests SHOULD be
     reported in a form of a table.  The rows SHOULD be labeled call
     accepted and call released.  The columns SHOULD be labeled time
     message sent, time response received, and correct response.  The
     elements of the columns 1 and 2 SHOULD be in seconds.  The
     elements of column 3 SHOULD be be either True or False, indicating
     whether the particular condition was observed for each test.

     The table MUST also indicate the packet size in octets and traffic
     rate in packets per second as generated by the test device.

3.4.3. Crankback Time

  Objective: To determine the Crankback Time on the SUT as defined in
  RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the uni-directional
      passthrough configuration.

  2)  Create a PNNI signaling setup message, as described in Appendix
      C, specifying a DTL which is not blocked by the far end SUT.

  3)  Send the PNNI signaling setup message.  Note the time the setup
      message was sent.  Verify that the connect message has been
      received by the near-end switch.  Note the time the connect
      message was received

  4)  Create a PNNI signaling setup message, as described in Appendix
      C, specifying a DTL which is blocked by the far end SUT.

  5)  Send the PNNI signaling release message.  Note the time the
      release message was sent.  Note the time the release complete



Dunn & Martin                Informational                    [Page 108]

RFC 3116            Methodology for ATM Benchmarking           June 2001


      message was received.  Note the time the near-end switch sends
      it's own PNNI setup message (referred to as the near-end setup
      message) specifying the non- blocked DTL.

  6)  Compute the crankback time as the difference between the time the
      near-end setup message was received and the time release message
      was send.

  Reporting Format:

     The results of the Crankback Time tests SHOULD be reported in a
     form of a table.  The rows SHOULD be labeled DTL call accepted and
     call released.  The columns SHOULD be labeled time message sent,
     time response received, and correct response.  The elements of the
     columns 1 and 2 SHOULD be in seconds.  The elements of column 3
     SHOULD be be either True or False, indicating whether the
     particular condition was observed for each test.

     The table MUST also indicate the packet size in octets and traffic
     rate in packets per second as generated by the test device.

3.4.4. Route Update Response Time

  Objective: To determine the Route Update Response Time on the SUT as
  defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the uni-directional
      passthrough configuration.

  2)  Create a PNNI PTSE as described in Appendix C, specifying a
      routing topology.  Verify that the routing tables on the far-end
      and near-end switches are empty.

  3)  Send the PTSE message to the far-end switch.  Note the time the
      PTSE message was sent.  Verify that the PTSE message has been
      received by the far-end switch.  Note the time the PTSE message
      was received.

  4)  Create another PNNI PTSE as described in Appendix C, specifying a
      change in the routing topology.  Verify that the routing tables
      on the far-end and near-end switches contain the previous PTSE
      routes.

  5)  Send the PTSE message to the far-end switch.  Note the time the
      PTSE message was sent.  Verify that the PTSE message has been
      received by the far-end switch.  Note the time the PTSE message



Dunn & Martin                Informational                    [Page 109]

RFC 3116            Methodology for ATM Benchmarking           June 2001


      was received.  Note the time the PTSE was sent to the near-end
      switch.  Note the time the PTSE message was received on the
      near-end switch.

  6)  Compute the Route Update Response time as the difference between
      the time the far-end PTSE message was sent and the time far-end
      PTSE message was received by the near-end.

  Reporting Format:

     The results of the Route Update Response Time tests SHOULD be
     reported in a form of a table.  The rows SHOULD be labeled PTSE
     call accepted, far-end PTSE message send, and near-end message
     received.  The columns SHOULD be labeled time message sent, time
     response received, and correct response.  The elements of the
     columns 1 and 2 SHOULD be in seconds.  The elements of column 3
     SHOULD be be either True or False, indicating whether the
     particular condition was observed for each test.

     The table MUST also indicate the packet size in octets and traffic
     rate in packets per second as generated by the test device.

3.5. ATM Service: ILMI

3.5.1. MIB Alignment Time

  Objective: To determine the MIB Alignment Time on the SUT as defined
  in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Send a Cold Start message to the SUT.  Note the time the message
      was sent to the SUT.  Verify that the Cold Start message has been
      received by the SUT.  Note the time the message was received.

  3)  Send a Get Request message to the SUT.  Note the time the message
      was sent to the SUT.  Verify that the Get Request message has
      been received by the SUT.  Note the time the message was
      received.

  4)  After all MIB elements are exchanged, verify that the final Get
      Request message has been received by the SUT.  Note the time the
      message was send and received by the SUT.





Dunn & Martin                Informational                    [Page 110]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  5)  Compute the MIB Alignment Time as the difference between the time
      the Cold Start message was sent and the time the final Get
      Request was received by the SUT.

  Reporting Format:

     The results of the MIB Alignment Time tests SHOULD be reported in
     a form of a table.  The rows SHOULD be labeled Cold Start Send,
     Cold Start accepted, Final Get Request send, and Final Get Request
     received.  The columns SHOULD be labeled time message sent, time
     response received, and correct response.  The elements of the
     columns 1 and 2 SHOULD be in seconds.  The elements of column 3
     SHOULD be be either True or False, indicating whether the
     particular condition was observed for each test.

     The table MUST also indicate the packet size in octets and traffic
     rate in packets per second as generated by the test device.

3.5.2. Address Registration Time

  Objective: To determine the Address Registration Time on the SUT as
  defined in RFC 2761 "Terminology for ATM Benchmarking".

  Procedure:

  1)  Set up the SUT and test device using the bi-directional
      configuration.

  2)  Send a Set Request message to the SUT.  Note the time the message
      was sent to the SUT.  Verify that the Set Request message has
      been received by the SUT.  Note the time the message was
      received.

  3)  Send a Get Request message to the SUT.  Note the time the message
      was sent to the SUT.  Verify that the Get Request message has
      been received by the SUT.  Note the time the message was
      received.

  4)  After all MIB elements are exchanged, verify that the final Get
      Request message has been received by the SUT.  Note the time the
      message was send and received by the SUT.

  5)  Compute the Address Registration Time as the difference between
      the time the Set Request message was sent and the time the final
      Get Request was received by the SUT.






Dunn & Martin                Informational                    [Page 111]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  Reporting Format:

     The results of the Address Registration Time tests SHOULD be
     reported in a form of a table.  The rows SHOULD be labeled Set
     Request Send, Set Request accepted, Final Get Request send, and
     Final Get Request received.  The columns SHOULD be labeled time
     message sent, time response received, and correct response.  The
     elements of the columns 1 and 2 SHOULD be in seconds.  The
     elements of column 3 SHOULD be be either True or False, indicating
     whether the particular condition was observed for each test.

     The table MUST also indicate the packet size in octets and traffic
     rate in packets per second as generated by the test device.

4. Security Considerations

  As this document is solely for the purpose of providing methodology
  and describes neither a protocol nor an implementation, there are no
  security considerations associated with this document.

5. Notices

  The IETF takes no position regarding the validity or scope of any
  intellectual property or other rights that might be claimed to
  pertain to the implementation or use of the technology described in
  this document or the extent to which any license under such rights
  might or might not be available; neither does it represent that it
  has made any effort to identify any such rights.  Information on the
  IETFs procedures with respect to rights in standards-track and
  standards-related documentation can be found in BCP-11.  Copies of
  claims of rights made available for publication and any assurances of
  licenses to be made available, or the result of an attempt made to
  obtain a general license or permission for the use of such
  proprietary rights by implementors or users of this specification can
  be obtained from the IETF Secretariat.

  The IETF invites any interested party to bring to its attention any
  copyrights, patents or patent applications, or other proprietary
  rights which may cover technology that may be required to practice
  this standard.  Please address the information to the IETF Executive
  Director.










Dunn & Martin                Informational                    [Page 112]

RFC 3116            Methodology for ATM Benchmarking           June 2001


6. References

  [RFC2544]      Bradner, S. and J. McQuaid, "Benchmarking Methodology
                 for Network Interconnect Devices", RFC 2544, March
                 1999.

  [RFC2225]      Laubach, M. and J. Halpern, "Classical IP and ARP over
                 ATM", RFC 2225, April 1998.

  [RFC2761]      Dunn, J. and C. Martin, "Terminology for ATM
                 Benchmarking", RFC 2761, February 2000.

  [AF-ILMI4.0]   ATM Forum Integrated Local Management Interface
                 Version 4.0, af-ilmi-0065.000, September 1996.

  [AF-TEST-0022] Introduction to ATM Forum Test Specifications, af-
                 test-0022.00, December 1994.

  [AF-TM4.1]     ATM Forum, Traffic Management Specification Version
                 4.1, af-tm-0121.00, April 1996.

  [AF-UNI3.1]    ATM Forum, User Network Interface Specification
                 Version 3.1, September 1994.

  [AF-UNI4.0]    ATM Forum, User Network Interface Specification
                 Version 4.0, July 1996.

7. Authors' Addresses

  Jeffrey Dunn
  Advanced Network Consultants, Inc.
  4214 Crest Place
  Ellicott City, MD 21043, USA

  Phone: +1 (410) 750-1700
  EMail: [email protected]


  Cynthia Martin
  Advanced Network Consultants, Inc.
  4214 Crest Place
  Ellicott City, MD 21043, USA

  Phone: +1 (410) 750-1700
  EMail: [email protected]






Dunn & Martin                Informational                    [Page 113]

RFC 3116            Methodology for ATM Benchmarking           June 2001


Appendix A: Ranges

  ATM NSAP Network Prefix.
    39 0000 0000 0000 0000 0000 0000-39 0000 0000 0000 0000 0000 00FF
    39 0000 0000 0000 0000 0001 0000-39 0000 0000 0000 0000 0001 00FF
    39 0000 0000 0000 0001 0000 0000
    39 0000 0000 0000 0002 0020 0000
    39 0000 0000 0300 0002 0030 0000
    39 0000 0000 4000 0002 0060 0000
    39 0000 0006 0060 0002 0030 0000
    39 0000 0006 0050 0002 0030 0000
    39 0000 0009 0300 0002 0030 0000
    39 0000 00A0 0300 0002 0030 0000
    39 0000 0B00 0300 0002 0030 0000
    39 0000 C000 0300 0002 0030 0000

  ATM NSAP End System Identifier.
    1111 1111 1111 00-1111 1111 11FF 00
    2222 2222 2000 00-2222 2222 2222 00
    9999 999A 0000 00-9999 999C 0000 00

Appendix B: Rates

  PNNI Routing Update Size.

  1) 1 PNNI routing entry update on non-aggregated addresses

  2) 2 PNNI routing entry updates on non-aggregated addresses

  3) 5 PNNI routing entry updates on non-aggregated addresses

  4) 1 % of total available bandwidth or 1 Mb/s, whichever is less on
     non- aggregated addresses

  5) 1 % of total available bandwidth or 1 Mb/s, whichever is less on
     of non-aggregated addresses and of aggregated addresses

  6) 1 % of total available bandwidth or 1 Mb/s, whichever is less on
     aggregated addresses

  7) 2 % of total available bandwidth or 2 Mb/s, whichever is less on
     non- aggregated addresses

  8) 2 % of total available bandwidth or 2 Mb/s, whichever is less on
     of non-aggregated addresses and of aggregated addresses

  9) 2 % of total available bandwidth or 2 Mb/s, whichever is less on
     aggregated addresses



Dunn & Martin                Informational                    [Page 114]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  PNNI Routing Update Repetition Interval.

  Repetition Interval begins after initial PNNI routing table
     stabilizes.

  1) 1 update every 1 hour, for 24 hours

  2) 1 update every 30 minutes, for 24 hours

  3) 1 update every 5 minutes, for 1 hour

  4) 1 update every 1 minute, for 15 minutes

  5) 1 update every 30 seconds, for 5 minutes

  6) 1 update every 30 seconds, for 1 minute

  7) 1 update every 1 second, for 30 seconds

  Maximum WAN Connection rates in packets per second (pps):

                   25.6        OC-3c       OC-12c
  IP Packet Size
  octets/cells
      44/2         30188       176603      706412
      64/2         30188       176603      706412
     128/3         20125       117735      470940
     256/6         10062        58867      235468
   1024/22          2744        16054      64216
   1518/32          1886        11037      44148
   2048/43          1404         8214      32856
   4472/94           642         3757      15028
  9180/192           314         1839       7356

  Maximum LAN Connection rates in packets per second (pps):

                   DS-1       DS-3       E1        E3
  IP Packet Size
  octets/cells
      44/2          1811      52133      2340     40000
      64/2          1811      52133      2340     40000
     128/3          1207      34755      1560     26666
     256/6           603      17377       780     13333
   1024/22           164       4739       212      3636
   1518/32           113       3258       146      2500
   2048/43            84       2424       108      1860
   4472/94            38       1109        49       851
   9180/192           18        543        24       416



Dunn & Martin                Informational                    [Page 115]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  Notes: 1.  PDU size in cells is computed based on ceiling( ( PDU size
  in octets + 16) / 48).  This assumes an 8 octet LLC/SNAP header and
  an 8 octet AAL/5 trailer.

  2.  Due to the number of possible configurations, IMA pps rates are
  not listed, but may be derived from the following formula: floor
  (IDCR/cells per packet), where cells per packet is computed as in
  note 1.

  3. The following cell rates were used: DS-1 = 3622 cps (using ATM TC)
  E1 = 4681 cps 25.6 Mb/s = 60377 cps E3 = 80000 cps (using ATM TC)
  DS-3 = 104266 cps (using ATM TC) OC-3c = 353207 cps OC-12c = 1412828
  cps

Appendix C: PDU's

TCP/IP over ATM Example 1.
   LLC:    DSAP                        0xAA (SNAP-SAP)
               SSAP                       0xAA (SNAP-SAP)
               Control                    0x03 (Unnumbered Information)
   SNAP: OUI                           0x00-00-00 (Ethertype)
                PID                       0x0800 (Internet Protocol)
   IP:      Version = 4
            Header length = 20
            Type of service = 0
                000. .... Precedence = Routine(0)
                ...0 .... Delay = Normal (0)
                .... 0... Throughput = Normal (0)
                .... .0.. Reliability = Normal (0)
            Packet length = 40
            Id = 0
            Fragmentation Info = 0x0000
                .0.. ....  .... .... Don't Fragment Bit = FALSE
                ..0. ....  .... .... More Fragments Bit = FALSE
                ...0 0000  0000 0000 Fragment offset = 0
            Time to live = 255
            Protocol = TCP (6)
            Header checksum = F9CF
            Source address = 15.19.209.236
            Destination address = 15.19.209.237
   TCP:     Source port = smtp (25)
            Destination port = smtp (25)
            Sequence number = 1
            Ack number = 0
            Data offset = 20
            Flags = 0x02
                ..0. .... URGENT Flag = FALSE
                ...0 .... ACK Flag = FALSE



Dunn & Martin                Informational                    [Page 116]

RFC 3116            Methodology for ATM Benchmarking           June 2001


                .... 0... PUSH Flag = FALSE
                .... .0.. RST Flag = FALSE
                .... ..1. SYN Flag = TRUE
                .... ...0 FIN Flag = FALSE
            Window = 0
            Checksum = EDAF
            Urgent pointer = 00000000

TCP/IP over ATM Example 2.
LLC:     DSAP                         0xAA (SNAP-SAP)
            SSAP                        0xAA (SNAP-SAP)
            Control                     0x03 (Unnumbered Information)
   SNAP:  OUI                        0x00-00-00 (Ethertype)
            PID                         0x0800 (Internet Protocol)
   IP:      Version = 4
            Header length = 20
            Type of service = 0
                000. .... Precedence = Routine(0)
                ...0 .... Delay = Normal (0)
                .... 0... Throughput = Normal (0)
                .... .0.. Reliability = Normal (0)
            Packet length = 40
            Id = 0
            Fragmentation Info = 0x0000
                .0.. ....  .... .... Don't Fragment Bit = FALSE
                ..0. ....  .... .... More Fragments Bit = FALSE
                ...0 0000  0000 0000 Fragment offset = 0
            Time to live = 255
            Protocol = TCP (6)
            Header checksum = F9CF
            Source address = 15.19.209.236
            Destination address = 15.19.209.237
   TCP:     Source port = ftp-data (20)
            Destination port = 2000
            Sequence number = 1
            Ack number = 0
            Data offset = 20
            Flags = 0x02
                ..0. .... URGENT Flag = FALSE
                ...0 .... ACK Flag = FALSE
                .... 0... PUSH Flag = FALSE
                .... .0.. RST Flag = FALSE
                .... ..1. SYN Flag = TRUE
                .... ...0 FIN Flag = FALSE
            Window = 0
            Checksum = E5FD
            Urgent pointer = 00000000




Dunn & Martin                Informational                    [Page 117]

RFC 3116            Methodology for ATM Benchmarking           June 2001


UDP/IP over ATM Example.
   LLC:    DSAP                        0xAA (SNAP-SAP)
           SSAP                        0xAA (SNAP-SAP)
           Control                     0x03 (Unnumbered Information)
   SNAP:   OUI                         0x00-00-00 (Ethertype)
           PID                         0x0800 (Internet Protocol)
   IP:      Version = 4
            Header length = 20
            Type of service = 0
                000. .... Precedence = Routine(0)
                ...0 .... Delay = Normal (0)
                .... 0... Throughput = Normal (0)
                .... .0.. Reliability = Normal (0)
            Packet length = 28
            Id = 0
            Fragmentation Info = 0x0000
                .0.. ....  .... .... Don't Fragment Bit = FALSE
                ..0. ....  .... .... More Fragments Bit = FALSE
                ...0 0000  0000 0000 Fragment offset = 0
            Time to live = 255
            Protocol = ICMP (1)
            Header checksum = F9E0
            Source address = 15.19.209.236
            Destination address = 15.19.209.237
   ICMP:    Type = Echo request (8)
            Code = 0
            Checksum = F7FF
            Identifier = 0 (0x0)
            Sequence Number = 0 (0x0)

RIP Routing Update over ATM.

   -- DATAGRAM HEADER
         offset data (hex)            description
         00     FF FF FF FF FF FF     dest MAC address is broadcast
         06     xx xx xx xx xx xx     source hardware address
         12     08 00                 type

         -- IP HEADER
         14     45                    IP version - 4, header length (4
        byte units) - 5
         15     00                    service field
         16     00 EE                 total length
         18     00 00                 ID
         20     40 00                 flags (3 bits) 4 (do not
        fragment),
                                      fragment offset-0
         22     0A                    TTL



Dunn & Martin                Informational                    [Page 118]

RFC 3116            Methodology for ATM Benchmarking           June 2001


         23     11                    protocol - 17 (UDP)
         24     C4 8D                 header checksum
         26     xx xx xx xx           source IP address
         30     xx xx xx              destination IP address
         33     FF                    host part = FF for broadcast

         -- UDP HEADER
         34     02 08                 source port 208 = RIP
         36     02 08                 destination port 208 = RIP
         38     00 DA                 UDP message length
         40     00 00                 UDP checksum

         -- RIP packet
         42     02                  command = response
         43     01                  version = 1
         44     00 00               0

         -- net 1
         46     00 02               family = IP
         48     00 00               0
         50     xx xx xx            net 1 IP address
         53     00                  net not node
         54     00 00 00 00         0
         58     00 00 00 00         0
         62     00 00 00 07         metric 7

         -- net 2

         66     00 02               family = IP
         68     00 00               0
         70     xx xx xx            net 2 IP address
         73     00                  net not node
         74     00 00 00 00         0
         78     00 00 00 00         0
         82     00 00 00 07         metric 7

         -- net 3
         86     00 02               family = IP
         88     00 00               0
         90     xx xx xx            net 3 IP address
         93     00                  net not node
         94     00 00 00 00         0
         98     00 00 00 00         0
         102    00 00 00 07         metric 7

         -- net 4
         106    00 02               family = IP
         108    00 00               0



Dunn & Martin                Informational                    [Page 119]

RFC 3116            Methodology for ATM Benchmarking           June 2001


         110    xx xx xx            net 4 IP address
         113    00                  net not node
         114    00 00 00 00         0
         118    00 00 00 00         0
         122    00 00 00 07         metric 7

         -- net 5
         126    00 02               family = IP
         128    00 00               0
         130    00                  net 5 IP address
         133    00                  net not node
         134    00 00 00 00         0
         138    00 00 00 00         0
         142    00 00 00 07         metric 7

         -- net 6
         146    00 02               family = IP
         148    00 00               0
         150    xx xx xx            net 6 IP address
         153    00                  net not node
         154    00 00 00 00         0
         158    00 00 00 00         0
         162    00 00 00 07         metric 7

  UNI  3.1 Signaling Setup Message Example.  PCR will not allow CAC to
  reject the call.

   Protocol Discriminator    : Q.93B UNI call control
   Call Reference Length     : 3
   Call Reference Flag       : orig
   Call Reference Value      : 0
   Message Type              : SETUP
   Ext                       : last octet
   Action Indicator          : clear call
   Message Length            : 50
   Information Element ID    : ATM Traffic Descriptor
   Ext                       : last octet
   Coding Standard           : ITU-T standard
   Action Indicator          : clear call
   IE Length                 : 9
   Cell Rate Subfield ID     : forward peak CR(CLP=0+1)
   Forward Peak Cell Rate    : 1
   Cell Rate Subfield ID     : backward peak CR(CLP=0+1)
   Backward Peak Cell Rate   : 1
   Cell Rate Subfield ID     : best effort indicator
   Information Element ID    : Broadband Bearer Capability
   Ext                       : last octet
   Coding Standard           : ITU-T standard



Dunn & Martin                Informational                    [Page 120]

RFC 3116            Methodology for ATM Benchmarking           June 2001


   Action Indicator          : clear call
   IE Length                 : 2
   Ext                       : last octet
   Bearer Class              : BCOB-X
   Ext                       : last octet
   Clipping Susceptibility   : not susceptible to clipping
   User Plane Connection CFG : point-to-point
   Information Element ID    : Called Party Number
   Ext                       : last octet
   Coding Standard           : ITU-T standard
   Action Indicator          : clear call
   IE Length                 : 21
   Ext                       : last octet
   Addressing/Numbering Plan : ISO NSAP addressing
   ISO NSAP Address Octets   : 3900000000000000000000000011111111111100
   Information Element ID    : Quality of Service Parameter
   Ext                       : last octet
   Coding Standard           : ITU-T standard
   Action Indicator          : clear call
   IE Length                 : 2
   QoS Class Forward         : QoS class 0 - unspecified
   QoS Class Backward        : QoS class 0 - unspecified

  UNI 3.1 Signaling Setup Message Reject Example.  PCR  will  allow
  CAC  to reject the call.

   Protocol Discriminator    : Q.93B UNI call control
   Call Reference Length     : 3
   Call Reference Flag       : orig
   Call Reference Value      : 0
   Message Type              : SETUP
   Ext                       : last octet
   Action Indicator          : clear call
   Message Length            : 50
   Information Element ID    : ATM Traffic Descriptor
   Ext                       : last octet
   Coding Standard           : ITU-T standard
   Action Indicator          : clear call
   IE Length                 : 8
   Cell Rate Subfield ID     : forward peak CR(CLP=0+1)
   Forward Peak Cell Rate    : 300000
   Cell Rate Subfield ID     : backward peak CR(CLP=0+1)
   Backward Peak Cell Rate   : 300000
   Information Element ID    : Broadband Bearer Capability
   Ext                       : last octet
   Coding Standard           : ITU-T standard
   Flag                      : not significant
   Action Indicator          : clear call



Dunn & Martin                Informational                    [Page 121]

RFC 3116            Methodology for ATM Benchmarking           June 2001


   IE Length                 : 3
   Ext                       : another octet
   Bearer Class              : BCOB-X
   Ext                       : last octet
   Traffic Type              : constant bit rate
   Timing Requirements       : end-to-end timing required
   Ext                       : last octet
   Clipping Susceptibility   : not susceptible to clipping
   User Plane Connection CFG : point-to-point
   Information Element ID    : Called Party Number
   Ext                       : last octet
   Coding Standard           : ITU-T standard
   Action Indicator          : clear call
   IE Length                 : 21
   Ext                       : last octet
   Addressing/Numbering Plan : ISO NSAP addressing
   ISO NSAP Address Octets   : 3900000000000000000000000011111111111100
   Information Element ID    : Quality of Service Parameter
   Ext                       : last octet
   Coding Standard           : ITU-T standard
   Action Indicator          : clear call
   IE Length                 : 2
   QoS Class Forward         : QoS class 0 - unspecified
   QoS Class Backward        : QoS class 0 - unspecified

  UNI  3.1 Signaling Release Message, specifying a cause code of normal
  call clearing.

   Protocol Discriminator   : Q.93B UNI call control
   Call Reference Length    : 3
   Call Reference Flag      : orig
   Call Reference Value     : 0
   Message Type             : RELEASE
   Ext                      : last octet
   Action Indicator         : clear call
   Message Length           : 6
   Information Element ID   : Cause
   Ext                      : last octet
   Coding Standard          : ITU-T standard
   Action Indicator         : clear call
   IE Length                : 2
   Ext                      : last octet
   Location                 : user
   Ext                      : last octet
   Cause Value              : NE:normal call clearing

  PNNI Signaling Setup Message, specifying a DTL which is not blocked
  by the far end SUT.



Dunn & Martin                Informational                    [Page 122]

RFC 3116            Methodology for ATM Benchmarking           June 2001


   Protocol Discriminator    : PNNI signalling
   Call Reference Length     : 3
   Call Reference Flag       : from
   Message Type              : SETUP
   Ext                       : last octet
   Pass Along Request        : no pass along request
   Action Indicator          : clear call
   Message Length            : 56
   Information Element ID    : ATM Traffic Descriptor
   Ext                       : last octet
   Coding Standard           : ITU-T standardized
   Pass Along Request        : no pass along request
   Action Indicator          : clear call
   IE Length                 : 0
   Information Element ID    : Broadband Bearer Capability
   Ext                       : last octet
   Coding Standard           : ITU-T standardized
   Pass Along Request        : no pass along request
   Action Indicator          : clear call
   IE Length                 : 3
   Ext                       : another octet
   Bearer Class              : BCOB-X
   Ext                       : last octet
   ATM Transfer Capability   : reserved for bwd compatibility
   Ext                       : last octet
   Clipping Susceptibility   : not susceptible to clipping
   User Plane Connection cfg : point-to-point
   Information Element ID    : Called Party Number
   Ext                       : last octet
   Coding Standard           : ITU-T standardized
   Pass Along Request        : no pass along request
   Action Indicator          : clear call
   IE Length                 : 8
   Ext                       : last octet
   Type of Number            : unknown
   Addressing/Numbering Plan : ATM endsystem address
   ATM Endsystem Address Oct : 11111111111101
   Information Element ID    : Designated Transit List
   Ext                       : last octet
   Coding Standard           : ATM Forum specific
   Pass Along Request        : no pass along request
   Action Indicator          : clear call
   IE Length                 : 29
   Current Transit Pointer   : 0
   Logical Node/Port Indicat : Logical Node/Port Indicator
   Logical Node Identifier   : 3900000000000000000000000011111111111100





Dunn & Martin                Informational                    [Page 123]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  PNNI  Signaling Setup Message Reject, specifying a DTL which is
  blocked by the far end SUT.

Protocol Discriminator      : PNNI signalling
   Call Reference Length   : 3
   Call Reference Flag     : from
   Call Reference Value    : 0
   Message Type            : SETUP
   Ext                     : last octet
   Pass Along Request      : no pass along request
   Action Indicator        : clear call
   Message Length          : 56
   Information Element ID  : ATM Traffic Descriptor
   Ext                     : last octet
   Coding Standard         : ITU-T standardized
   Pass Along Request      : no pass along request
   Action Indicator        : clear call
   IE Length               : 0
   Information Element ID  : Broadband Bearer Capability
   Ext                     : last octet
   Coding Standard         : ITU-T standardized
   Pass Along Request      : no pass along request
   Action Indicator        : clear call
   IE Length               : 3
   Bearer Class            : BCOB-X
   Ext                     : last octet
   ATM Transfer Capability : reserved for bwd compatibility
   Ext                     : last octet
   Clipping Susceptibility : not susceptible to clipping
   User Plane Connection cfg : point-to-point
   Information Element ID  : Called Party Number
   Ext                     : last octet
   Coding Standard         : ITU-T standardized
   Pass Along Request      : no pass along request
   Action Indicator        : clear call
   IE Length               : 8
   Ext                     : last octet
   Addressing/Numbering Plan : ATM endsystem address
   ATM Endsystem Address Oct : 11111111111101
   Information Element ID    : Designated Transit List
   Ext                       : last octet
   Coding Standard           : ATM Forum specific
   Pass Along Request        : no pass along request
   Action Indicator          : clear call
   IE Length                 : 29
   Current Transit Pointer   : 0
   Logical Node/Port Indicat : Logical Node/Port Indicator
   Logical Node Identifier   : 3900000000000000000000000011111111111100



Dunn & Martin                Informational                    [Page 124]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  PNNI Far End Request Message.

Header:  Packet Type                    5 (PTSE REQUEST)
            Packet Length             40
            Protocol Version           1
            Newest Version Supported   1
            Oldest Version Supported   0
            Reserved                   0
   IG:      Information Group Type   513 (Requested PTSE Header)
            Information Group Length  32
            Originating Node ID
                  00013900-00000000-00000000-00000011-11111111-1100
            PTSE Request Count     1
            PTSE Identifier        0

  PNNI PTSE, specifying a routing topology.

Header:  Packet Type                    4 (DATABASE SUMMARY)
            Packet Length             76
            Protocol Version           1
            Newest Version Supported   1
            Oldest Version Supported   0
            Reserved                   0
            Initialize (I)Bit          1 (during init. of DB syn
                                          process)
            More (M)Bit                1 (PTSEs to summarize)
            Master (MS)Bit             1 (both nodes)
            Reserved                   0
            Reserved                   0
            DS Sequence Number         0
   IG:      Information Group Type   512 (Nodal PTSE Summaries)
            Information Group Length  60
            Originating Node ID
                00013900-00000000-00000000-00000011-11111111-1100
            Originating Node's Peer Group 00000000-00000000-00000000-
                                          0001
            Reserved                    0
            PTSE Summary Count          1
            PTSE Type                   0
            Reserved                    0
            PTSE Identifier             0
            PTSE Sequence Number        0
            PTSE Checksum               0
            PTSE Remaining Lifetime     0







Dunn & Martin                Informational                    [Page 125]

RFC 3116            Methodology for ATM Benchmarking           June 2001


  PNNI PTSE Update, specifying a change in the routing topology.

Header:  Packet Type                    2 (PTSP)
            Packet Length             96
            Protocol Version           1
            Newest Version Supported   1
            Oldest Version Supported   0
            Reserved                   0
            Originating Node ID
                00013900-00000000-00000000-00000011-11111111-1100
            Originating Node's Peer Group 00000000-00000000-00000000-
                                          0001
   IG:      Information Group Type     64 (PTSE)
            Information Group Length   52
            PTSE Type                   0
            Reserved                    0
            PTSE Identifier             0
            PTSE Sequence Number        0
            PTSE Checksum           42252
            PTSE Remaining Lifetime  3600
   IG:       Information Group Type   224 (Internal Reachable ATM
                                           Addresses)
            Information Group Length   32
            VP Capability Flag          1 (VPCs supported)
            Reserved                    0
            Reserved                    0
            Port ID                     0
            Scope of Advertisement     96
            Address Information Length 14
            Address Information Count   1
            Prefix Length              13
            Reachable Address Prefix   39000000-00000000-00000000-01



















Dunn & Martin                Informational                    [Page 126]

RFC 3116            Methodology for ATM Benchmarking           June 2001


Full Copyright Statement

  Copyright (C) The Internet Society (2001).  All Rights Reserved.

  This document and translations of it may be copied and furnished to
  others, and derivative works that comment on or otherwise explain it
  or assist in its implementation may be prepared, copied, published
  and distributed, in whole or in part, without restriction of any
  kind, provided that the above copyright notice and this paragraph are
  included on all such copies and derivative works.  However, this
  document itself may not be modified in any way, such as by removing
  the copyright notice or references to the Internet Society or other
  Internet organizations, except as needed for the purpose of
  developing Internet standards in which case the procedures for
  copyrights defined in the Internet Standards process must be
  followed, or as required to translate it into languages other than
  English.

  The limited permissions granted above are perpetual and will not be
  revoked by the Internet Society or its successors or assigns.

  This document and the information contained herein is provided on an
  "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
  TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
  BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
  HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
  MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.

Acknowledgement

  Funding for the RFC Editor function is currently provided by the
  Internet Society.



















Dunn & Martin                Informational                    [Page 127]


You are viewing proxied material from gopher.fnord.one. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.