Network Working Group                                        K. Murakami
Request for Comments: 2175                                   M. Maruyama
Category: Informational                                 NTT Laboratories
                                                              June 1997


              MAPOS 16 - Multiple Access Protocol over
                   SONET/SDH with 16 Bit Addressing

Status of this Memo

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

Authors' note

  This memo documents MAPOS 16, a multiple access protocol for
  transmission of network-protocol datagrams, encapsulated in HDLC
  frames with 16 bit addressing, over SONET/SDH.  The primary
  difference with MAPOS version 1 is that it has 16 bit address field
  that offers significant wide address space. This document is NOT the
  product of an IETF working group nor is it a standards track
  document.  It has not necessarily benefited from the widespread and
  in depth community review that standards track documents receive.

Abstract

  This document describes the protocol MAPOS 16, Multiple Access
  Protocol over SONET/SDH with 16 Bit Addressing, for transmitting
  network-protocol datagrams over SONET/SDH.  The primary difference
  with MAPOS version 1 is that it has 16 bit address field that offers
  significant wide address space. It first describes the major
  differences between MAPOS and MAPOS 16 briefly. Then, it explains the
  header format and its 16 bit address format.

1. Introduction

  MAPOS is a multiple access protocol for transmission of High-level
  Datalink Control (HDLC) frames over the Synchronous Optical Network /
  Synchronous Digital Hierarchy(SONET/SDH)[1][2][3][4]. It provides
  multiple access capability to SONET/SDH, an inherently point-to-point
  link.

  MAPOS version 1[5] focuses on the frame format compatibility with the
  conventional PPP[6], resulting with its narrow 8 bit address field.
  In contrast to MAPOS version 1, MAPOS 16 has a 16 bit address space.




Murakami & Maruyama          Informational                      [Page 1]

RFC 2175                        MAPOS 16                       June 1997


  In this document, header format and its 16 bit format are described.
  It also explains that 16 bit addressing has minimal influence on the
  conventional MAPOS protocol family such as Node-Switch Protocol[7]
  and Switch-Switch Protocol[8].

2. MAPOS 16 Frame Format

  Like MAPOS version 1, MAPOS 16 framing is based on the HDLC-like
  framing used in PPP-over-SONET/SDH, described in RFC-1662[6].
  However, the address field is extended to 16 bits, and the control
  field in the MAPOS version 1 is omitted for maintain the 32bit
  alignment of the header.

  Figure 2 shows the MAPOS 16 frame format.  Logical Link Control
  (LLC), and Sublayer/Sub-Network Access Protocol (SNAP) are not used.
  It does not include the bytes for transparency.  The fields are
  transmitted from left to right.


          +----------+---------------------+----------+
          |          |                     |          |
          |   Flag   |       Address       | Protocol |
          | 01111110 |        16bits       |  16 bits |
          +----------+---------------------+----------+
             +-------------+------------+----------+-----------
             |             |            |          | Inter-frame
             | Information |    FCS     |   Flag   | fill or next
             |             | 16/32 bits | 01111110 | address
             +-------------+------------+----------+------------

                       Figure 2.  Frame format

    Flag Sequence

    Flag sequence is used for frame synchronization.  Each frame begins
    and ends with a flag sequence 01111110 (0x7E).  If a frame
    immediately follows another, one flag sequence may be treated as
    the end of the preceding frame and the beginning of the immediately
    following frame.  When the line is idle, the flag sequence is to be
    transmitted continuously on the line.

    Address

    The address field contains the destination HDLC address.  A frame
    is forwarded by a switch based on this field.  It is 16 bits wide.
    The LSB of the first byte indicates the continuation of this field,
    and must always be 0. The LSB of the second byte indicates the end
    of this field, and must always be 1.  The MSB of the first byte is



Murakami & Maruyama          Informational                      [Page 2]

RFC 2175                        MAPOS 16                       June 1997


    used to indicate if the frame is a unicast or multicast frame.  The
    MSB of 0 means unicast, with the remaining thirteen bits indicating
    the destination node address including two E/A bits. MSB of 1 means
    multicast, with the remaining thirteen bits indicating the group
    address.  The address 0xFEFF means that the frame is a broadcast
    frame. The address (0x0001) is reserved to identify the control
    processor inside a switch.  Frames with an invalid address should
    be silently discarded.

            +-------------+-+-------------+-+
            | | | | | | | | | | | | | | | | |
            | | node addr |0|  node addr  |1|
            +-+-----------+-+-------------+-+
             ^             ^               ^
             |             |               |
             |             |               +------- EA bit (always 1)
             |             +------- EA bit (always 0)
             1 : broadcast, multicast
             0 : unicast

                       Figure 3 Address format



    Protocol

    The protocol field indicates the protocol to which the datagram
    encapsulated in the information field belongs.  It conforms to the
    ISO 3309 extension mechanism, and the value for this field may be
    obtained from the most recent ``Assigned Numbers'' [9] and ``MAPOS
    Version 1 Assigned Numbers'' [10].

    Information

    The information field contains the datagram for the protocol
    specified in the protocol field.  The length of this field may
    vary, but shall not exceed 65,280 (64K - 256) octets.

    Frame Check Sequence (FCS)

    By default, the frame check sequence (FCS) field is 16-bits long.
    Optionally, 32 bit FCS may be used instead.  The FCS is calculated
    over all bits of the address, protocol, and information fields
    prior to escape conversions.  The least significant octet of the
    result is transmitted first as it contains the coefficient of the
    highest term.





Murakami & Maruyama          Informational                      [Page 3]

RFC 2175                        MAPOS 16                       June 1997


    Inter-frame fill

    A sending station must continuously transmit the flag sequence as
    inter-frame fill after the FCS field.  The inter-frame flag
    sequences must be silently discarded by the receiving station.
    When an under-run occurs during DMA in the sending station, it must
    abort the frame transfer and continuously send the flag sequence to
    indicate the error.

3.2 Octet-Synchronous Framing

  MAPOS 16 uses the same octet stuffing procedure[6] as MAPOS version
  1[5]. Since SONET/SDH provides transparency, Async-Control-
  Character-Map (ACCM) is not used.  HDLC frames are mapped into the
  SONET/SDH payload as follows.

  Each HDLC frame is separated from another frame by one or more flag
  sequence, 01111110 (0x7E).  An escape sequence is defined to escape
  the flag sequence and itself.  Prior to sending the frame, but after
  the FCS computation, every occurrence of 01111110 (0x7E) other than
  the flags is to be converted to the sequence 01111101 01011110 (0x7D
  0x5E), and the sequence 01111101 (0x7D) is to be converted to the
  sequence 01111101 01011101 (0x7D 0x5D).  Upon receiving a frame, this
  conversion must be reversed prior to FCS computation.

4. Influence on MAPOS ARP, UNARP, NSP, and SSP

  Since all of the MAPOS protocol family, ARP[11], UNARP[11], NSP[7],
  and SSP[8], use 32-bit address field for 8-bit MAPOS address, the
  16-bit addressing has no influence on them.  Each protocol only have
  to place a 16 bit address in the least significant place in their 32
  bit address fields as follows;

  (1) MAPOS ARP and UNARP
   16-bit addresses are placed in the least significant places of the
   32-bit sender and target HDLC addresses.

  (2) NSP
   In address assignment packet, a 16-bit address is placed in the
   least significant bytes of the 32-bit address field. The rest of the
   field is padded with zeros.

  (3) SSP
   In route entry of an SSP packet, the 16-bit MAPOS address is placed
   in the least significant bytes of the 32-bit address field.






Murakami & Maruyama          Informational                      [Page 4]

RFC 2175                        MAPOS 16                       June 1997


5. Mapping IP Multicast Address to MAPOS 16 Address

  When transmitting IP multicast[11], the thirteen bits of the HDLC
  address must contain the lowest-order thirteen bits of the IP
  multicast group address.  Exceptions arise when these thirteen bits
  are either all zeros or all ones, in which case the address 1111 1110
  1111 1101 should be used.

6. Security Considerations

  Security issues are not discussed in this memo.

References

  [1]  CCITT Recommendation G.707: Synchronous Digital Hierarchy Bit
       Rates (1990).

  [2]  CCITT Recommendation G.708: Network Node Interface for
       Synchronous Digital Hierarchy (1990).

  [3]  CCITT Recommendation G.709: Synchronous Multiplexing Structure
       (1990).

  [4]  American National Standard for Telecommunications - Digital
       Hierarchy - Optical Interface Rates and Formats Specification,
       ANSI T1.105-1991.

  [5]  Murakami, K. and M. Maruyama, " MAPOS - Multiple Access Protocol
       over SONET/SDH version 1", RFC2171, June, 1997.

  [6]  Simpson, W., editor, "PPP in HDLC-like Framing," RFC1662, July
       1994.

  [7]  Murakami, K. and M. Maruyama, "A MAPOS version 1 Extension -
       Node Switch Protocol," RFC2173, June, 1997.

  [8]  Murakami, K. and M. Maruyama, "A MAPOS version 1 Extension -
       Switch Switch Protocol," RFC2174, June, 1997.

  [9]  Reynolds, J. and J. Postel, "ASSIGNED NUMBERS," RFC1700, Oct.
       1994.

  [10] Maruyama, M. and K. Murakami, "MAPOS Version 1 Assigned
       Numbers," RFC2172, June, 1997.

  [11] Murakami, K. and M. Maruyama, "IPv4 over MAPOS Version 1,"
       RFC2176, June, 1997.




Murakami & Maruyama          Informational                      [Page 5]

RFC 2175                        MAPOS 16                       June 1997


Acknowledgements

  The authors would like to acknowledge the contributions and
  thoughtful suggestions of John P. Mullaney, Clark Bremer, Masayuki
  Kobayashi, Paul Francis, Toshiaki Yoshida, and Takahiro Sajima.

Author's Address

            Ken Murakami
            NTT Software Laboratories
            3-9-11, Midori-cho
            Musashino-shi
            Tokyo-180, Japan
            E-mail: [email protected]

            Mitsuru Maruyama
            NTT Software Laboratories
            3-9-11, Midori-cho
            Musashino-shi
            Tokyo-180, Japan
            E-mail: [email protected]






























Murakami & Maruyama          Informational                      [Page 6]