Adaptive rate allocation in a joint source/channel coding framework for wireless channels

We propose an adaptive rate allocation scheme to choose the rates of the source coder and channel coder pair optimally. We express the end-to-end expected distortion in terms of the source coder distortion and the channel distortion. Also, we propose a modified Gilbert (1960) noise channel to model the combination of channel encoder, channel, and channel decoder. Based on our new model, we estimate the channel distortion and select the best source coder and channel coder pair to encode and transmit the source. Simulation results show the optimality of the rate allocation scheme and the advantage of the adaptation.