CaltechAUTHORS
  A Caltech Library Service

Lossless and near-lossless source coding for multiple access networks

Zhao, Qian and Effros, Michelle (2003) Lossless and near-lossless source coding for multiple access networks. IEEE Transactions on Information Theory, 49 (1). pp. 112-128. ISSN 0018-9448. http://resolver.caltech.edu/CaltechAUTHORS:ZHAieeetit03

[img]
Preview
PDF
See Usage Policy.

1278Kb

Use this Persistent URL to link to this item: http://resolver.caltech.edu/CaltechAUTHORS:ZHAieeetit03

Abstract

A multiple access source code (MASC) is a source code designed for the following network configuration: a pair of correlated information sequences {X-i}(i=1)(infinity), and {Y-i}(i=1)(infinity) is drawn independent and identically distributed (i.i.d.) according to joint probability mass function (p.m.f.) p(x, y); the encoder for each source operates without knowledge of the other source; the decoder jointly decodes the encoded bit streams from both sources. The work of Slepian and Wolf describes all rates achievable by MASCs of infinite coding dimension (n --> infinity) and asymptotically negligible error probabilities (P-e((n)) --> 0). In this paper, we consider the properties of optimal instantaneous MASCs with finite coding dimension (n < infinity) and both lossless (P-e((n)) = 0) and near-lossless (P-e((n)) --> 0) performance. The interest in near-lossless codes is inspired by the discontinuity in the limiting rate region at P-e((n)) = 0 and the resulting performance benefits achievable by using near-lossless MASCs as entropy codes within lossy MASCs. Our central results include generalizations of Huffman and arithmetic codes to the MASC framework for arbitrary p(x, y), n, and P-e((n)) and polynomial-time design algorithms that approximate these optimal solutions.


Item Type:Article
Additional Information:© Copyright 2003 IEEE. Reprinted wih permission. Manuscript received June 23, 2001; revised October 2, 2002. [Posted online: 2003-01-14] This work was supported by NSF under Awards CCR-9909026 and CCR-0220039 and by the Caltech Lee Center for Advanced Networking. The material in this paper was presented in part at the Data Compression Conference, Snowbird, UT, March 2001 and the IEEE International Symposium on Information Theory, Washington, DC, June 2001. Communicated by R. Zamir, Associate Editor for Source Coding. The authors gratefully acknowledge the detailed critiques and many helpful comments from the anonymous reviewers.
Subject Keywords:arithmetic codes; distributed source coding; Huffman codes; lossless coding; near-lossless coding; network source coding; polynomial complexity; Slepian-Wolf; Wyner-Ziv codes; source code design; source coding theory; network information theory
Record Number:CaltechAUTHORS:ZHAieeetit03
Persistent URL:http://resolver.caltech.edu/CaltechAUTHORS:ZHAieeetit03
Alternative URL:http://dx.doi.org/10.1109/TIT.2002.806145
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:4769
Collection:CaltechAUTHORS
Deposited By: Archive Administrator
Deposited On:06 Sep 2006
Last Modified:26 Dec 2012 09:01

Repository Staff Only: item control page