The University of British Columbia

The task of broadcasting video in wireless environments requires high coding efficiency in addition to reliable error resilience techniques. Existing solutions are set to tackle either coding efficiency and bandwidth utilization on the one hand, or error resilience and recovery from packet delay or loss on the other. We propose a novel combined approach to deal with the problem of video broadcast over wireless networks by offering a layered multiple description scalable coding (LMDSC) technique using the subband extension of H.264. This approach combines the high coding efficiency and layered structure of the subband extension of H.264/AVC along with the highly error resilient performance of multiple description coding (MDC) while virtually eliminating all redundancy between the transmitted video streams. Performance evaluations show that when faced with the same amount of packet loss, our approach achieves significant improvement in PSNR over existing methods.