Optimized Bit Rate Allocation for Iterative Source-Channel Decoding and its Extension towards Multi-Mode Transmission

Recent advancements in iterative source-channel decoding (ISCD) focus on the optimization of a single processing step of source coding (like the index assignment) or of channel coding. In this paper, we propose a new design guideline for a more general optimization of the overall ISCD scheme. In particular, we discuss how to allocate a given gross bit rate to source and channel coding. The new bit rate allocation reveals (at least) two benefits: Firstly, the proposed ISCD scheme using a rate r = 1 channel encoder yields improved error concealing/correcting capabilities if compared to the prior ISCD approaches. Secondly, the new scheme can easily be extended to a flexible multi-mode system. While all modes use the same r = 1 channel code, the joint adaptation of quantizer resolution and index assignment allows to manage the trade-off between quantization noise and error robustness.