Joint Source/Channel Coding with Low Density Parity Check Matrices

The objectives of this paper are two-fold, first to present the problem of joint source and channel coding from a graphical model perspective and second to propose a structure that uses a new graphical model for jointly encoding and decoding of a redundant source. In the first part of the paper, relevant contributions to joint source and channel coding, ranging from the Slepian-Wolf problem to joint decoding of variable length codes with state-of-the-art source codes, are reviewed and summarized. In the second part, a double low-density parity-check (LDPC) code for joint source and channel coding is proposed. The double LDPC code can be decoded as a single bipartite graph using standard belief propagation and its limiting performance is analyzed by using extrinsic information transfer (EXIT) chart approximations. M. Fresia was with Princeton University. She is now with Infineon Technologies AG, Munich, Germany. H. Vincent Poor and S. Verdú are with the Department of Electrical Engineering, Princeton University, Princeton, NJ (USA). F. Pérez-Cruz is with Universidad Carlos III de Madrid (Spain). E-mail: [email protected], [email protected], [email protected] and [email protected] This work was partially supported by National Science Foundation under Grants CNS-06-25637, NCR-0074277 CCR-0312879, CCF-0728445 and CCF-0635154. This work was also partially funded by Spanish government (Ministerio de Educación y Ciencia TEC2009-14504-C02-01, Consolider-Ingenio 2010 CSD2008-00010). F. Pérez-Cruz was supported by Marie Curie Fellowship 040883-AI-COM. December 31, 2009 DRAFT