Network Working Group                                           J. Welch
Request for Comments: 4445                        IneoQuest Technologies
Category: Informational                                         J. Clark
                                                          Cisco Systems
                                                             April 2006


                A Proposed Media Delivery Index (MDI)

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 (2006).

IESG Note

  This RFC is not a candidate for any level of Internet Standard.
  There are IETF standards which are highly applicable to the space
  defined by this document as its applicability, in particular, RFCs
  3393 and 3611, and there is ongoing IETF work in these areas as well.
  The IETF also notes that the decision to publish this RFC is not
  based on IETF review for such things as security, congestion control,
  MIB fitness, or inappropriate interaction with deployed protocols.
  The RFC Editor has chosen to publish this document at its discretion.
  Readers of this document should exercise caution in evaluating its
  value for implementation and deployment.  See RFC 3932 for more
  information.

Abstract

  This memo defines a Media Delivery Index (MDI) measurement that can
  be used as a diagnostic tool or a quality indicator for monitoring a
  network intended to deliver applications such as streaming media,
  MPEG video, Voice over IP, or other information sensitive to arrival
  time and packet loss.  It provides an indication of traffic jitter, a
  measure of deviation from nominal flow rates, and a data loss
  at-a-glance measure for a particular flow.  For instance, the MDI may
  be used as a reference in characterizing and comparing networks
  carrying UDP streaming media.

  The MDI measurement defined in this memo is intended for Information
  only.




Welch & Clark                Informational                      [Page 1]

RFC 4445         A Proposed Media Delivery Index (MDI)        April 2006


1.  Introduction

  There has been considerable progress over the last several years in
  the development of methods to provide for Quality of Service (QoS)
  over packet-switched networks to improve the delivery of streaming
  media and other time-sensitive and packet-loss-sensitive applications
  such as [i1], [i5], [i6], [i7].  QoS is required for many practical
  networks involving applications such as video transport to assure the
  availability of network bandwidth by providing upper limits on the
  number of flows admitted to a network, as well as to bound the packet
  jitter introduced by the network.  These bounds are required to
  dimension a receiver`s buffer to display the video properly in real
  time without buffer overflow or underflow.

  Now that large-scale implementations of such networks based on RSVP
  and Diffserv are undergoing trials [i3] and being specified by major
  service providers for the transport of streaming media such as MPEG
  video [i4], there is a need to diagnose issues easily and to monitor
  the real-time effectiveness of networks employing these QoS methods
  or to assess whether they are required.  Furthermore, due to the
  significant installed base of legacy networks without QoS methods, a
  delivery system`s transitional solution may be composed of networks
  with and without these methods, thus increasing the difficulty in
  characterizing the dynamic behavior of these networks.

  The purpose of this memo is to describe a set of measurements that
  can be used to derive a Media Delivery Index (MDI) that indicates the
  instantaneous and longer-term behavior of networks carrying streaming
  media such as MPEG video.

  While this memo addresses monitoring MPEG Transport Stream (TS)
  packets [i8] over UDP, the general approach is expected to be
  applicable to other streaming media and protocols.  The approach is
  applicable to both constant and variable bit rate streams though the
  variable bit rate case may be somewhat more difficult to calculate.
  This document focuses on the constant bit rate case as the example to
  describe the measurement, but as long as the dynamic bit rate of the
  encoded stream can be determined (the "drain rate" as described below
  in Section 3), then the MDI provides the measurement of network-
  induced cumulative jitter.  Suggestions and direction for calculation
  of MDI for a variable bit rate encoded stream may be the subject of a
  future document.

  Network packet delivery time variation and various statistics to
  characterize the same are described in a generic approach in [i10].
  The approach is capable of being parameterized for various purposes
  with the intent of defining a flexible, customizable definition that
  can be applied to a wide range of applications and further



Welch & Clark                Informational                      [Page 2]

RFC 4445         A Proposed Media Delivery Index (MDI)        April 2006


  experimentation.  Other approaches to characterizing jitter behavior
  are also captured such as in [i12].  A wide-ranging report format
  [i11] has been described to convey information including jitter for
  use with the RTP Control Protocol (RTCP) [i12].  The MDI is instead
  intended to specifically address the need for a scalable,
  economical-to-compute metric that characterizes network impairments
  that may be imposed on streaming media, independent of control plane
  or measurement transport protocol or stream encapsulation protocol.
  It is a targeted metric for use in production networks carrying large
  numbers of streams for the purpose of monitoring the network quality
  of the flows or for other applications intended to analyze large
  numbers of streams susceptible to IP network device impairments.  An
  example application is the burgeoning deployments of Internet
  Protocol Television (IPTV) by cable and telecommunication service
  providers.  As described below, MDI provides for a readily scalable
  per-stream measure focused on loss and the cumulative effects of
  jitter.

2.  Media Delivery Index Overview

  The MDI provides a relative indicator of needed buffer depths at the
  consumer node due to packet jitter as well as an indication of lost
  packets.  By probing a streaming media service network at various
  nodes and under varying load conditions, it is possible to quickly
  identify devices or locales that introduce significant jitter or
  packet loss to the packet stream.  By monitoring a network
  continuously, deviations from nominal jitter or loss behavior can be
  used to indicate an impending or ongoing fault condition such as
  excessive load.  It is believed that the MDI provides the necessary
  information to detect all network-induced impairments for streaming
  video or voice-over-IP applications.  Other parameters may be
  required to troubleshoot and correct the impairments.

  The MDI is updated at the termination of selected time intervals
  spanning multiple packets that contain the streaming media (such as
  transport stream packets in the MPEG-2 case).  The Maximums and
  Minimums of the MDI component values are captured over a measurement
  time.  The measurement time may range from just long enough to
  capture an anticipated network anomaly during a troubleshooting
  exercise to indefinitely long for a long-term monitoring or logging
  application.  The Maximums and Minimums may be obtained by sampling
  the measurement with adequate frequency.









Welch & Clark                Informational                      [Page 3]

RFC 4445         A Proposed Media Delivery Index (MDI)        April 2006


3.  Media Delivery Index Components

  The MDI consists of two components:  the Delay Factor (DF) and the
  Media Loss Rate (MLR).

3.1.  Delay Factor

  The Delay Factor is the maximum difference, observed at the end of
  each media stream packet, between the arrival of media data and the
  drain of media data.  This assumes the drain rate is the nominal
  constant traffic rate for constant bit rate streams or the piece-wise
  computed traffic rate of variable rate media stream packet data.  The
  "drain rate" here refers to the payload media rate; e.g., for a
  typical 3.75 Mb/s MPEG video Transport Stream (TS), the drain rate is
  3.75 Mb/s -- the rate at which the payload is consumed (displayed) at
  a decoding node.  If, at the sample time, the number of bytes
  received equals the number transmitted, the instantaneous flow rate
  balance will be zero; however, the minimum DF will be a line packet's
  worth of media data, as that is the minimum amount of data that must
  be buffered.

  The DF is the maximum observed value of the flow rate imbalance over
  a calculation interval.  This buffered media data in bytes is
  expressed in terms of how long, in milliseconds, it would take to
  drain (or fill) this data at the nominal traffic rate to obtain the
  DF.  Display of DF with a resolution of tenths of milliseconds is
  recommended to provide adequate indication of stream variations for
  monitoring and diagnostic applications for typical stream rates from
  1 to 40 Mb/s.  The DF value must be updated and displayed at the end
  of a selected time interval.  The selected time interval is chosen to
  be long enough to sample a number of TS packets and will, therefore,
  vary based on the nominal traffic rate.  For typical stream rates of
  1 Mb/s and up, an interval of 1 second provides a long enough sample
  time and should be included for all implementations.  The Delay
  Factor indicates how long a data stream must be buffered (i.e.,
  delayed) at its nominal bit rate to prevent packet loss.  This time
  may also be seen as a measure of the network latency that must be
  induced from buffering, which is required to accommodate stream
  jitter and prevent loss.  The DF`s max and min over the measurement
  period (multiple intervals) may also be displayed to show the worst
  case arrival time deviation, or jitter, relative to the nominal
  traffic rate in a measurement period.  It provides a dynamic flow
  rate balance indication with its max and min showing the worst
  excursions from balance.

  The Delay Factor gives a hint of the minimum size of the buffer
  required at the next downstream node.  As a stream progresses, the
  variation of the Delay Factor indicates packet bunching or packet



Welch & Clark                Informational                      [Page 4]

RFC 4445         A Proposed Media Delivery Index (MDI)        April 2006


  gaps (jitter).  Greater DF values also indicate that more network
  latency is necessary to deliver a stream due to the need to pre-fill
  a receive buffer before beginning the drain to guarantee no
  underflow.  The DF comprises a fixed part based on packet size and a
  variable part based on the buffer utilization of the various network
  component switch elements that comprise the switched network
  infrastructure [i2].

  To further detail the calculation of DF, consider a virtual buffer VB
  used to buffer received packets of a stream.  When a packet P(i)
  arrives during a calculation interval, compute two VB values,
  VB(i,pre) and VB(i,post), defined as:

  VB(i,pre) = sum (Sj) - MR * Ti; where j=1..i-1
  VB(i,post) = VB(i,pre) + Si

  where Sj is the media payload size of the jth packet, Ti is the
  relative time at which packet i arrives in the interval, and MR is
  the nominal media rate.

  VB(i,pre) is the Virtual Buffer size just before the arrival of P(i).
  VB(i,post) is the Virtual Buffer size just after the arrival of P(i).

  The initial condition of VB(0) = 0 is used at the beginning of each
  measurement interval.  A measurement interval is defined from just
  after the time of arrival of the last packet during a nominal period
  (typically 1 second) as mentioned above to the time just after the
  arrival of the last packet of the next nominal period.

  During a measurement interval, if k packets are received, then there
  are 2*k+1 VB values used in deriving VB(max) and VB(min).  After
  determining VB(max) and VB(min) from the 2k+1 VB samples, DF for the
  measurement interval is computed and displayed as:

  DF = [VB(max) - VB(min)]/ MR

  As noted above, a measurement interval of 1 second is typically used.
  If no packets are received during an interval, the last DF calculated
  during an interval in which packets did arrive is displayed.  The
  time of arrival of the last previous packet is always retained for
  use in calculating VB when the next packet arrives (even if the time
  of the last received packet spans measurement intervals).  For the
  first received measurement interval of a measurement period, no DF is
  calculated; however, packet arrival times are recorded for use in
  calculating VB during the following interval.






Welch & Clark                Informational                      [Page 5]

RFC 4445         A Proposed Media Delivery Index (MDI)        April 2006


3.2.  Media Loss Rate

  The Media Loss Rate is the count of lost or out-of-order flow packets
  over a selected time interval, where the flow packets are packets
  carrying streaming application information.  There may be zero or
  more streaming packets in a single IP packet.  For example, it is
  common to carry seven 188 Byte MPEG Transport Stream packets in an IP
  packet.  In such a case, a single IP packet loss would result in 7
  lost packets counted (if those 7 lost packets did not include null
  packets).  Including out-of-order packets is important, as many
  stream consumer-type devices do not attempt to reorder packets that
  are received out of order.

3.3.  Media Delivery Index

  Combining the Delay Factor and Media Loss Rate quantities for
  presentation results in the following MDI:

                                 DF:MLR
   Where:
                         DF is the Delay Factor
                       MLR is the Media Loss Rate

  At a receiving node, knowing its nominal drain bit rate, the DF`s max
  indicates the size of buffer required to accommodate packet jitter.
  Or, in terms of Leaky Bucket [i9] parameters, DF indicates bucket
  size b, expressed in time to transmit bucket traffic b, at the given
  nominal traffic rate, r.

3.4.  MDI Application Examples

  If a known, well-characterized receive node is separated from the
  data source by unknown or less well-characterized nodes such as
  intermediate switch nodes, the MDI measured at intermediate data
  links provides a relative indication of the behavior of upstream
  traffic flows.  DF difference indications between one node and
  another in a data stream for a given constant interval of calculation
  can indicate local areas of traffic congestion or possibly
  misconfigured QoS flow specification(s) leading to greater filling of
  measurement point local device buffers, resultant flow rate
  deviations, and possible data loss.










Welch & Clark                Informational                      [Page 6]

RFC 4445         A Proposed Media Delivery Index (MDI)        April 2006


  For a given MDI, if DF is high and/or the DF Max-Min captured over a
  significant measurement period of multiple intervals is high, jitter
  has been detected but the longer-term, average flow rate may be
  nominal.  This could be the result of a transient flow upset due to a
  coincident traffic stream unrelated to the flow of interest causing
  packet bunching.  A high DF may cause downstream buffer overflow or
  underflow or unacceptable latency even in the absence of lost data.

  Due to transient network failures or DF excursions, packets may be
  lost within the network.  The MLR component of the MDI shows this
  condition.

  Through automated or manual flow detection and identification and
  subsequent MDI calculations for real-time statistics on a flow, the
  DF can indicate the dynamic deterioration or increasing burstiness of
  a flow, which can be used to anticipate a developing network
  operation problem such as transient oversubscription.  Such
  statistics can be obtained for flows within network switches using
  available switch cpu resources due to the minimal computational
  requirements needed for small numbers of flows.  Statistics for all
  flows present on, say, a gigabit Ethernet network, will likely
  require dedicated hardware facilities, though these can be modest, as
  buffer requirements and the required calculations per flow are
  minimal.  By equipping network switches with MDI measurements, flow
  impairment issues can quickly be identified, localized, and
  corrected.  Until switches are so equipped with appropriate hardware
  resources, dedicated hardware tools can provide supplemental switch
  statistics by gaining access to switch flows via mirror ports, link
  taps, or the like as a transition strategy.

  The MDI figure can also be used to characterize a flow decoder's
  acceptable performance.  For example, an MPEG decoder could be
  characterized as tolerating a flow with a given maximum DF and MLR
  for acceptable display performance (acceptable on-screen artifacts).
  Network conditions such as Interior Gateway Protocol (IGP)
  reconvergence time then might also be included in the flow tolerance
  DF resulting in a higher-quality user experience.

4.  Summary

  The MDI combines the Delay Factor, which indicates potential for
  impending data loss, and Media Loss Rate as the indicator of lost
  data.  By monitoring the DF and MLR and their min and max excursions
  over a measurement period and at multiple strategic locations in a
  network, traffic congestion or device impairments may be detected and
  isolated for a network carrying streaming media content.





Welch & Clark                Informational                      [Page 7]

RFC 4445         A Proposed Media Delivery Index (MDI)        April 2006


5.  Security Considerations

  The measurements identified in this document do not directly affect
  the security of a network or user.  Actions taken in response to
  these measurements that may affect the available bandwidth of the
  network or the availability of a service is out of scope for this
  document.

  Performing the measurements described in this document only requires
  examination of payload header information (such as MPEG transport
  stream headers or RTP headers) to determine nominal stream bit rate
  and sequence number information.  Content may be encrypted without
  affecting these measurements.  Therefore, content privacy is not
  expected to be a concern.

6.  Informative References

  [i1]  Braden, R., Zhang, L., Berson, S., Herzog, S., and S. Jamin,
        "Resource ReSerVation Protocol (RSVP) -- Version 1 Functional
        Specification", RFC 2205, September 1997.

  [i2]  Partridge, C., "A Proposed Flow Specification", RFC 1363,
        September 1992.

  [i3]  R. Fellman, `Hurdles to Overcome for Broadcast Quality Video
        Delivery over IP` VidTranS 2002.

  [i4]  CableLabs `PacketCable Dynamic Quality-of-Service
        Specification`, PKT-SP-DQOS-I06-030415, 2003.

  [i5]  Shenker, S., Partridge, C., and R. Guerin, "Specification of
        Guaranteed Quality of Service", RFC 2212, September 1997.

  [i6]  Wroclawski, J., "Specification of the Controlled-Load Network
        Element Service", RFC 2211, September 1997.

  [i7]  Braden, R., Clark, D., and S. Shenker, "Integrated Services in
        the Internet Architecture: an Overview", RFC 1633, June 1994.

  [i8]  ISO/IEC 13818-1 (MPEG-2 Systems)

  [i9]  V. Raisanen, "Implementing Service Quality in IP Networks",
        John Wiley & Sons Ltd., 2003.

  [i10] Demichelis, C. and P. Chimento, "IP Packet Delay Variation
        Metric for IP Performance Metrics (IPPM)", RFC 3393, November
        2002.




Welch & Clark                Informational                      [Page 8]

RFC 4445         A Proposed Media Delivery Index (MDI)        April 2006


  [i11] Friedman, T., Caceres, R., and A. Clark, "RTP Control Protocol
        Extended Reports (RTCP XR)", RFC 3611, November 2003.

  [i12] Schulzrinne, H.,  Casner, S., Frederick, R., and V. Jacobson,
        "RTP: A Transport Protocol for Real-Time Applications", STD 64,
        RFC 3550, July 2003.

7.  Acknowledgements

  The authors gratefully acknowledge the contributions of Marc Todd and
  Jesse Beeson of IneoQuest Technologies, Inc., Bill Trubey and John
  Carlucci of Time Warner Cable, Nishith Sinha of Cox Communications,
  Ken Chiquoine of SeaChange International, Phil Proulx of Bell Canada,
  Dr Paul Stallard of TANDBERG Television, Gary Hughes of Broadbus
  Technologies, Brad Medford of SBC Laboratories, John Roy of Adelphia
  Communications, Cliff Mercer, PhD of Kasenna, Mathew Ho of Rogers
  Cable, and Irl Duling of Optinel Systems for reviewing and evaluating
  early versions of this document and implementations for MDI.

Authors' Addresses

  James Welch
  IneoQuest Technologies, Inc
  170 Forbes Blvd
  Mansfield, Massachusetts 02048

  Phone: 508 618 0312
  EMail: [email protected]


  James Clark
  Cisco Systems, Inc
  500 Northridge Road
  Suite 800
  Atlanta, Georgia 30350

  Phone: 678 352 2726
  EMail: [email protected]













Welch & Clark                Informational                      [Page 9]

RFC 4445         A Proposed Media Delivery Index (MDI)        April 2006


Full Copyright Statement

  Copyright (C) The Internet Society (2006).

  This document is subject to the rights, licenses and restrictions
  contained in BCP 78 and at www.rfc-editor.org/copyright.html, and
  except as set forth therein, the authors retain all their rights.

  This document and the information contained herein are provided on an
  "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS
  OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY AND THE INTERNET
  ENGINEERING TASK FORCE DISCLAIM 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.

Intellectual Property

  The IETF takes no position regarding the validity or scope of any
  Intellectual Property Rights 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; nor does it represent that it has
  made any independent effort to identify any such rights.  Information
  on the procedures with respect to rights in RFC documents can be
  found in BCP 78 and BCP 79.

  Copies of IPR disclosures made to the IETF Secretariat 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 implementers or users of this
  specification can be obtained from the IETF on-line IPR repository at
  http://www.ietf.org/ipr.

  The IETF invites any interested party to bring to its attention any
  copyrights, patents or patent applications, or other proprietary
  rights that may cover technology that may be required to implement
  this standard.  Please address the information to the IETF at
  [email protected].

Acknowledgement

  Funding for the RFC Editor function is provided by the IETF
  Administrative Support Activity (IASA).







Welch & Clark                Informational                     [Page 10]