Rateless Distributed Source-Channel Coding of Correlated Binary Sources Using Raptor Codes

Abstract

The problem of distributed source-channel coding of a binary source with correlated side information at the decoder is addressed in this paper. We implement and evaluate the performance of rateless Raptor codes for distributed source-channel coding in binary symmetric (BSC) channel and additive white Gaussian noise (AWGN) channel. We adapt the conventional distributed source codec for error-free channel using Raptor codes to noisy channels. The proposed implementation is rateless, that is, the rate of the code can be adapted to varying channel conditions with incremental redundancy. This is achieved with built in rateless property of Raptor codes.