Iterative Joint Source-Channel Decoding of H.264 Compressed Video

Abstract

This paper proposes an iterative joint source-channel decoding (IJSCD) scheme for the transmission of H.264 compressed video over a noisy channel. It uses channel coding along with H.264 semantic verification. The structure, selection of design parameters, and performance of the proposed IJSCD based on a rate-1/2 recursive systematic convolutional (RSC) code over an AWGN channel are described and discussed as an illustrative example. In the example, for the same PSNR, the proposed IJSCD scheme offers a significant saving of 2.1dB in required channel SNR as compared to a system using the same RSC code alone. Furthermore, the performance can be improved by iterative decoding at the cost of increased delay. Hence, a tradeoff can be made between performance improvement and delay.