Compression of Correlated Sources Using LDPC Codes

We consider the problem of compressing correlated binary sources when the correlation between sources is defined by a Hidden Markov Model (HMM). Specifically, the HMM describes the correlation pattern, i.e. the modulo-2 addition of the two sources. In this framework, we develop a density evolution analysis of a compression system using irregular LDPC codes as source codes. In order to achieve this goal, we modify the standard density evolution approach to incorporate the HMM. The proposed scheme is applied to the design of irregular LDPC codes that optimize the system performance. Theoretical results agree with simulations, and show that it is possible to achieve a performance close to the theoretical Slepian-Wolf limit.