Network Working Group                                        B. Thompson
Request for Comments: 3337                                      T. Koren
Category: Standards Track                                  Cisco Systems
                                                               B. Buffam
                                                         Seaway Networks
                                                           December 2002


                     Class Extensions for PPP over
          Asynchronous Transfer Mode Adaptation Layer 2 (AAL2)

Status of this Memo

   This document specifies an Internet standards track protocol for the
   Internet community, and requests discussion and suggestions for
   improvements.  Please refer to the current edition of the "Internet
   Official Protocol Standards" (STD 1) for the standardization state
   and status of this protocol.  Distribution of this memo is unlimited.

Copyright Notice

   Copyright (C) The Internet Society (2002).  All Rights Reserved.

Abstract

   The Point-to-Point Protocol (PPP) over Asynchronous Transfer Mode
   (ATM) Adaptation Layer 2 defines the encapsulation that allows a PPP
   session to be transported over an ATM virtual circuit using the ATM
   Adaptation Layer 2 (AAL2) adaptation layer.  This document defines a
   set of class extensions to PPP over AAL2 that implement equivalent
   functionality to Multi Class Multi Link PPP over a single ATM virtual
   circuit.  Instead of using Multi Link PPP as the basis for
   fragmentation functionality, this document uses the functionality of
   the Segmentation and Reassembly Service Specific Convergence Sublayer
   that is already required as the basic encapsulation format of PPP
   over AAL2.

1. Introduction

   Using AAL2 as an adaptation layer for PPP transport over ATM provides
   a bandwidth efficient transport for IP applications that generate
   small packets.  An example IP application that generates small
   packets is RTP encapsulated voice (Voice over IP).

   In addition to bandwidth efficiency, real-time applications such as
   voice require low latency.  RFC 2689 [2] describes an architecture
   for providing transport services for real time applications on low
   bit rate links.  The main components of the architecture are: a
   real-time encapsulation format for asynchronous and synchronous low-
   bitrate links, a header compression architecture optimized for real-
   time flows, elements of negotiation protocols used between routers
   (or between hosts and routers), and announcement protocols used by
   applications to allow this negotiation to take place.

   Multi Class Multi Link PPP [3] defines a fragment-oriented solution
   for the real-time encapsulation format part of the architecture
   defined in [2], i.e., for the queues-of-fragments type sender.  As
   described in more detail in the architecture document, a real-time
   encapsulation format is required to guarantee low latency in the
   presence of large non real time packets. For example, a 1500 byte
   packet on a 128 kbit/s ATM virtual circuit makes this link
   unavailable for the transmission of real-time information for about
   100 ms.  This adds a worst-case delay that causes real-time
   applications to operate with round-trip delays that are too high for
   many interactive tasks.  Multi Class Multi Link PPP defines a set of
   extensions of Multi Link PPP [4] that enable the sender to fragment
   the packets of various priorities into multiple classes of fragments,
   allowing high-priority packets to be sent between fragments of lower
   priorities.

   This document defines a set of class extensions to PPP over AAL2 [1]
   that implement equivalent functionality to Multi Class Multi Link PPP
   over a single ATM virtual circuit.  Instead of using Multi Link PPP
   as the basis for fragmentation functionality, this document uses the
   functionality of the Service Specific Segmentation and Reassembly
   Sublayer (SSSAR) [5] that is already required as the basic
   encapsulation format of PPP over AAL2.

   In addition to providing fragmentation, the real time transport
   service must allow high priority fragments to be sent between
   fragments of lower priorities.  This can be accomplished in PPP over
   AAL2 by allowing a single PPP session to span multiple AAL2 CPS [6]
   Channel Identifiers.  Once a PPP session spans multiple AAL2 Channel
   IDs, the Channel ID can be used to identify the class that a fragment
   belongs to.  Fragments belonging to a high priority class can be sent
   using a particular AAL2 Channel ID.  Fragments of lower priority
   classes can be sent using different AAL2 Channel IDs.  Once multiple
   fragment classes are identified using different AAL2 Channel IDs, the
   AAL2 CPS layer can be used to send fragments belonging to a high
   priority class between fragments of lower priorities.

   The class based extensions to PPP over AAL2 use existing services of
   the AAL2 SSCS and CPS layers already specified in PPP over AAL2.
   Because of this, the extensions described in this document may be
   viewed as a desirable alternative to Multi Class Multi Link PPP in
   providing a class based transport service with PPP over AAL2.

1.1. Specification Language

   The keywords MUST, MUST NOT, REQUIRED, SHALL, SHALL NOT, SHOULD,
   SHOULD NOT, RECOMMENDED, MAY, and OPTIONAL, when they appear in this
   document, are to be interpreted as described in [7].

2. Requirements

   This document assumes the same service requirements as defined in
   Multi Class Multi Link PPP [3].  The reader is referred to section 2
   of Multi Class Multi Link PPP for the general requirements of a multi
   class fragmentation / preemption service.

3. Class Extensions for PPP over AAL2

   PPP over AAL2 uses the Service Specific Segmentation and Reassembly
   Sublayer (SSSAR) [5] for the AAL type 2.  The SSSAR sub-layer is used
   to segment PPP packets into frames that can be transported using the
   AAL2 CPS.  The SSSAR sub-layer uses different AAL2 UUI code-points to
   indicate whether a segment is the last segment of a packet or not.
   SSSAR provides basic fragmentation functionality for all packets
   encapsulated using PPP over AAL2.  The SSSAR layer fragments all
   packets into 64 byte fragments.

   The AAL2 CPS layer defines a Channel ID that is used to identify
   multiple streams of packets within a single ATM Virtual Circuit.  In
   this document, the AAL2 CPS Channel ID is used to identify the
   preemption class that a packet fragment belongs to.  Since the
   Channel ID is used to identify different preemption classes, packet
   fragments from each class of traffic MUST be assigned to different
   Channel IDs.  In addition, each PPP session MUST have at least as
   many Channel IDs assigned as there are different classes of
   preemptible traffic.

   To allow PPP packets to be assigned to different preemption classes,
   PPP packets must be classified into multiple preemption classes as
   they are fragmented using SSSAR.  Many classification methods may be
   used to determine the class that a particular PPP packet belongs to.
   The architecture document [2] describes possible alternatives that
   MAY be used to implement a real time classification scheme.

   Once packets have been classified into different preemption classes,
   each class of traffic is then assigned a different Channel ID.  Since
   fragments from each traffic class are now transmitted using separate
   Channel ID, the AAL2 CPS layer can be used to schedule fragments from
   the different classes.  The AAL2 CPS specification [6] does not
   specify a method for scheduling AAL2 CPS payloads from different
   Channel IDs.  The scheduling method required at the AAL2 CPS layer
   depends upon the real time requirements of applications using this
   service.  Some real-time applications MAY require the use of a
   priority based CID scheduler.  Other applications MAY only require a
   fair or weighted fair CID scheduler.  Implementations of PPP over
   AAL2 real time transport extensions SHOULD implement AAL2 CPS CID
   schedulers that meet the requirements of multi-class real time
   applications.

4. Example Implementation: Class Based Extensions for Voice Service

   When PPP over AAL2 is used to transport both voice and non-voice
   packets over low bandwidth ATM virtual circuits, it may be necessary
   to preempt the transmission of a large data packet in order to
   transmit a voice packet with minimal delay.  The example
   implementation described below shows an example of how the class
   extensions for PPP over AAL2 can be used to support a real time voice
   transport service over low bandwidth AAL2 virtual circuits.  To
   guarantee low latency and loss for voice transport, the ATM virtual
   circuit in this example must be provisioned using a real time traffic
   class such as VBRnrt or VBRrt.

   For the simple voice service described above, 2 classes are
   sufficient to guarantee low latency for voice packets.  The PPP over
   AAL2 session in this case can be configured to run across 2 AAL2 CPS
   Channel IDs.  One channel ID is used to transport large data packets
   while the other channel ID is used to transport real time voice
   packets.

   Packets that arrive at the PPP interface must first be classified as
   either belonging to the real time class or belonging to the data
   class.  A simple classifier that can be used to classify packets at
   this layer is packet size.

   Large packets are assigned to the non-real time (or data) traffic
   class and small packets are assigned to the real time traffic class.
   The packet size used to discriminate between real time and non-real
   time packets may vary based on the application and transmission rate
   of the virtual circuit.

   Once packets have been classified, they are now fragmented using the
   SSSAR layer of PPP over AAL2.  Separate instances of the SSSAR
   fragmentation function run on each of the 2 Channel IDs assigned to
   the PPP session.

   Fragments coming from the SSSAR functions are now scheduled into the
   AAL2 virtual circuit using the AAL2 CPS layer.  Most AAL2 SAR
   implementations currently implement fair scheduling across multiple
   AAL2 Channel IDs.  Since the AAL2 CPS scheduler implements fair
   scheduling, real time fragments will wait for at most one non-real
   time fragment to be transmitted on the AAL2 virtual circuit before
   being scheduled.

5.  Security Considerations

   Operation of this protocol is believed to be no more and no less
   secure than operation of PPP over AAL2 [1].

6. Acknowledgements

   The authors would like to thank James Carlson for his contributions
   to this proposal.

7. References

   [1] Thompson, B., Koren, T. and B. Buffam, "PPP Over Asynchronous
       Transfer Mode Adaptation Layer 2", RFC 3336, December 2002.

   [2] Bormann, C., "Providing Integrated Services over Low-bitrate
       Links", RFC 2689, September 1999.

   [3] Bormann, C., "The Multi-Class Extension to Multi-Link PPP", RFC
       2686 September 1999.

   [4] Sklower, K., Lloyd, B., McGregor, G., Carr, D. and T. Coradetti,
       "The PPP Multilink Protocol (MP)", RFC 1990, August 1996.

   [5] International Telecommunications Union, "Segmentation
       and Reassembly Service Specific Convergence Sublayer for the AAL
       type 2", ITU-T Recommendation I.366.1, June 1998.

   [6] International Telecommunications Union, "BISDN ATM Adaptation
       layer specification: Type 2 AAL(AAL2)", ITU-T Recommendation
       I.363.2, September 1997.

   [7] Bradner, S., "Key words for use in RFCs to Indicate Requirement
       Levels", BCP 14, RFC 2119, March 1997.

8. Authors' Addresses

   Bruce Thompson
   Cisco Systems, Inc.
   170 West Tasman Drive
   San Jose, CA 95134
   USA

   Phone: +1 408 527-0446
   EMail: brucet@cisco.com


   Bruce Buffam
   Seaway Networks
   One Chrysalis Way,
   Suite 300,
   Ottawa, Canada
   K2G-6P9

   Phone: +1 613 723-9161
   EMail: bruce@seawaynetworks.com


   Tmima Koren
   Cisco Systems, Inc.
   170 West Tasman Drive
   San Jose, CA 95134
   USA

   Phone: +1 408 527-6169
   EMail: tmima@cisco.com

9.  Full Copyright Statement

   Copyright (C) The Internet Society (2002).  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.