Network Working Group                                        R. Friend
Request for Comments: 2395                                  R. Monsour
Category: Informational                                    Hi/fn, Inc.
                                                        December 1998


                   IP Payload Compression Using LZS

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 (1998).  All Rights Reserved.

Abstract

  This document describes a compression method based on the LZS
  compression algorithm. This document defines the application of the
  LZS algorithm to the IP Payload Compression Protocol [IPCOMP].
  [IPCOMP] defines a method for applying lossless compression to the
  payloads of Internet Protocol datagrams.

Table of Contents

  1. Introduction...................................................2
     1.1 General....................................................2
     1.2 Background of LZS Compression..............................2
     1.3 Licensing..................................................3
     1.4 Specification of Requirements..............................3
  2. Compression Process............................................3
     2.1 Compression History........................................3
     2.2 Compression Encoding Format................................3
     2.3 Padding....................................................4
  3. Decompression Process..........................................4
  4. IPComp Association (IPCA) Parameters...........................4
     4.1 ISAKMP Transform ID........................................5
     4.2 ISAKMP Security Association Attributes.....................5
     4.3 Manual configuration.......................................5
     4.4 Minimum packet size threshold..............................5
     4.5 Compressibility test.......................................5
  5. Security Considerations........................................5
  6. Acknowledgements...............................................5
  7. References.....................................................6
  8. Authors' Addresses.............................................7



Friend & Monsour             Informational                      [Page 1]

RFC 2395            IP Payload Compression Using LZS       December 1998


  9. Appendix: Compression Efficiency versus Datagram Size..........8
  10. Full Copyright Statement......................................9

1. Introduction

1.1 General

  This document specifies the application of LZS compression, a
  lossless compression algorithm, to IP datagram payloads. This
  document is to be used in conjunction with the IP Payload Compression
  Protocol [IPCOMP].  This specification assumes a thorough
  understanding of the IPComp protocol.

1.2 Background of LZS Compression

  Starting with a sliding window compression history, similar to [LZ1],
  Hi/fn developed a new, enhanced compression algorithm identified as
  LZS. The LZS algorithm is a general purpose lossless compression
  algorithm for use with a wide variety of data types.  Its encoding
  method is very efficient, providing compression for strings as short
  as two octets in length.

  The LZS algorithm uses a sliding window of 2,048 bytes.  During
  compression, redundant sequences of data are replaced with tokens
  that represent those sequences. During decompression, the original
  sequences are substituted for the tokens in such a way that the
  original data is exactly recovered. LZS differs from lossy
  compression algorithms, such as those often used for video
  compression, that do not exactly reproduce the original data.

  The details of LZS compression can be found in [ANSI94].

  The efficiency of the LZS algorithm depends on the degree of
  redundancy in the original data.  A table of compression ratios for
  the [Calgary] Corpus file set is provided in the appendix in Section
  7.















Friend & Monsour             Informational                      [Page 2]

RFC 2395            IP Payload Compression Using LZS       December 1998


1.3 Licensing

  Hi/fn, Inc. holds patents on the LZS algorithm. Licenses for a
  reference implementation are available for use in IPPCP, IPSec, TLS
  and PPP applications at no cost.  Source and object licenses are
  available on a non-discriminatory basis. Hardware implementations are
  also available.  For more information, contact Hi/fn at the address
  listed with the authors' addresses.

1.4 Specification of Requirements

  The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
  "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
  document are to be interpreted as described in [RFC-2119].

2. Compression Process

2.1 Compression History

  The sender MUST reset the compression history prior to processing
  each datagram's payload. This ensures that each datagram's payload
  can be decompressed independently of any other, as is needed when
  datagrams are received out of order.

  The sender MUST flush the compressor each time it transmits a
  compressed datagram.  Flushing means that all data going into the
  compressor is included in the output, i.e., no data is held back in
  the hope of achieving better compression.  Flushing is necessary to
  prevent a datagram's data from spilling over into a later datagram.

2.2 Compression Encoding Format

  The input to the payload compression algorithm is an IP datagram
  payload. The output of the algorithm is a new (and hopefully smaller)
  payload. The output payload contains the input payload's data in
  either compressed or uncompressed format. The input and output
  payloads are each an integral number of bytes in length.

  If the uncompressed form is used, the output payload is identical to
  the input payload and the IPComp header is omitted.  If the
  compressed form is used, the output payload is prepended with the
  IPComp header and encoded as defined in [ANSI94], which is repeated
  here for informational purposes ONLY.

  <Compressed Stream> := [<Compressed String>] <End Marker>
  <Compressed String> := 0 <Raw Byte> | 1 <Compressed Bytes>
  <Raw Byte> := <b><b><b><b><b><b><b><b>          (8-bit byte)
  <Compressed Bytes> := <Offset> <Length>



Friend & Monsour             Informational                      [Page 3]

RFC 2395            IP Payload Compression Using LZS       December 1998


  <Offset> := 1 <b><b><b><b><b><b><b> |           (7-bit offset)
              0 <b><b><b><b><b><b><b><b><b><b><b> (11-bit offset)
  <End Marker> := 110000000

  <b> := 1 | 0

  <Length> :=
  00        = 2     1111 0110      = 14
  01        = 3     1111 0111      = 15
  10        = 4     1111 1000      = 16
  1100      = 5     1111 1001      = 17
  1101      = 6     1111 1010      = 18
  1110      = 7     1111 1011      = 19
  1111 0000 = 8     1111 1100      = 20
  1111 0001 = 9     1111 1101      = 21
  1111 0010 = 10    1111 1110      = 22
  1111 0011 = 11    1111 1111 0000 = 23
  1111 0100 = 12    1111 1111 0001 = 24
  1111 0101 = 13     ...

2.3 Padding

  A datagram payload compressed using LZS always ends with the last
  compressed data byte (also known as the <end marker>), which is used
  to disambiguate padding.  This allows trailing bits as well as bytes
  to be considered padding.

  The size of a compressed payload MUST be in whole octet units.

3. Decompression Process

  If the received datagram is compressed, the receiver MUST reset the
  decompression history prior to processing the datagram. This ensures
  that each datagram can be decompressed independently of any other, as
  is needed when datagrams are received out of order. Following the
  reset of the decompression history, the receiver decompresses the
  Payload Data field according to the encoding specified in section 3.2
  of [ANSI94].

  If the received datagram is not compressed, the receiver needs to
  perform no decompression processing and the Payload Data field of the
  datagram is ready for processing by the next protocol layer.

4. IPComp Association (IPCA) Parameters

  ISAKMP MAY be used to negotiate the use of the LZS compression method
  to establish an IPCA, as defined in [IPCOMP].




Friend & Monsour             Informational                      [Page 4]

RFC 2395            IP Payload Compression Using LZS       December 1998


4.1 ISAKMP Transform ID

  The LZS Transform ID as IPCOMP_LZS, as specified in The Internet IP
  Security Domain of Interpretation [SECDOI].  This value is used to
  negotiate the LZS compression algorithm under the ISAKMP protocol.

4.2 ISAKMP Security Association Attributes

  There are no other parameters required for LZS compression negotiated
  under ISAKMP.

4.3 Manual configuration

  The CPI value IPCOMP_LZS is used for a manually configured IPComp
  Compression Associations.

4.4 Minimum packet size threshold

  As stated in [IPCOMP], small packets may not compress well.  Informal
  tests using the LZS algorithm over the Calgary Corpus data set show
  that the average payload size that may produce expanded data is
  approximately 90 bytes.  Thus implementations may not want to attempt
  to compress payloads smaller than 90 bytes.

4.5 Compressibility test

  There is no adaptive algorithm embodied in the LZS algorithm, for
  compressibility testing, as referenced in [IPCOMP].

5. Security Considerations

  This document does not add any further security considerations that
  [IPCOMP] and [Deutsch96] have already declared.

6. Acknowledgments

  The LZS details presented here are similar to those in PPP LZS-DCP
  Compression Protocol (LZS-DCP), [RFC-1967].

  The author wishes to thank the participants of the IPPCP working
  group mailing list whose discussion is currently active and is
  working to generate the protocol specification for integrating
  compression with IP.








Friend & Monsour             Informational                      [Page 5]

RFC 2395            IP Payload Compression Using LZS       December 1998


7. References

  [AH]       Kent, S., and R., Atkinson, "IP Authentication Header",
             RFC 2402, November 1998.

  [ANSI94]   American National Standards Institute, Inc., "Data
             Compression Method for Information Systems," ANSI X3.241-
             1994, August 1994.

  [Calgary]  Text Compression Corpus, University of Calgary, available
             at ftp://ftp.cpsc.ucalgary.ca/pub/projects/text.
             compression.corpus.

  [IPCOMP]   Shacham, A., "IP Payload Compression Protocol (IPComp)",
             RFC 2393, December 1998.

  [LZ1]      Lempel, A., and Ziv, J., "A Universal Algorithm for
             Sequential Data Compression", IEEE Transactions On
             Information Theory, Vol.  IT-23, No. 3, May 1977.

  [RFC-1962] Rand, D., "The PPP Compression Control Protocol (CCP)",
             RFC 1962, June 1996.

  [RFC-1967] Schneider, K., and R. Friend, "PPP LZS-DCP Compression
             Protocol (LZS-DCP)", RFC 1967, August 1996.

  [RFC-2003] Perkins, C., "IP Encapsulation within IP", RFC 2003,
             October 1996.

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

  [SECDOI]   Piper, D., "The Internet IP Security Domain of
             Interpretation for ISAKMP", RFC 2407, November 1998.

















Friend & Monsour             Informational                      [Page 6]

RFC 2395            IP Payload Compression Using LZS       December 1998


8. Authors' Addresses

  Robert Friend
  Hi/fn Inc.
  5973 Avenida Encinas
  Suite 110
  Carlsbad, CA  92008

  EMail: [email protected]


  Robert Monsour
  Hi/fn Inc.
  2105 Hamilton Avenue
  Suite 230
  San Jose, CA 95125

  EMail: [email protected]

































Friend & Monsour             Informational                      [Page 7]

RFC 2395            IP Payload Compression Using LZS       December 1998


9. Appendix: Compression Efficiency versus Datagram Size

  The following table offers some guidance on the compression
  efficiency that can be achieved as a function of datagram size.  Each
  entry in the table shows the compression ratio that was achieved when
  LZS was applied to a test file using datagrams of a specified size.

  The test file was the University of Calgary Text Compression Corpus
  [Calgary].  The Calgary Corpus consists of 18 files with a total size
  (all files) of 3.278MB.

   Datagram size,|
   bytes         |  64   128   256   512  1024  2048  4096  8192 16384
   --------------|----------------------------------------------------
   Compression   |1.18  1.28  1.43  1.58  1.74  1.91  2.04  2.11  2.14
   ratio         |



































Friend & Monsour             Informational                      [Page 8]

RFC 2395            IP Payload Compression Using LZS       December 1998


10.  Full Copyright Statement

  Copyright (C) The Internet Society (1998).  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.
























Friend & Monsour             Informational                      [Page 9]