Low-Delay Distributed Source Coding: Bounds and Performance of Practical Codes

The performance of distributed source coding methods based on scalar quantization followed by scalar entropy coding is investigated. For a fixed quantizer, the problem is converted to that of zero-error entropy coding for bipartite graphs with a special structure. It is then experimentally shown that for bipartite graphs corresponding to jointly Gaussian sources and high-resolution uniform quantization, (i) the asymptotic rate region for zero-error entropy coding is significantly above the rates that can be achieved by Slepian-Wolf coding of the quantization indices, and (ii) a very simple scalar zero-error entropy coding scheme comes very close to the asymptotic performance. Also, a new type of non-uniform quantization based on reversing the order of compressors and expandors is introduced, and it is shown that under the proposed framework, non-uniform quantization can outperform uniform quantization, in contrast with point-to-point communication scenarios.