Multiple Description Coding versus Transport Layer FEC for Resilient Video Transmission

Video content delivery is a challenging task due to its large bandwidth requirements and its sensitivity to transmission errors. In this context, simultaneously providing scalability and resilience against transmission errors is of paramount importance. Layered video coding coupled with multicast video transmission employing no feedback provides the required scalability features and reduces the network burden. However, it is not easy to provide reliability in such environments. This is crucial to mobile users as error rates on wireless links can be high. To compensate for transmission errors, redundancy must be added to the video content. This can be added at the application layer by the employed video coding system, or can be added transparently by the transport layer. In this paper, we compare the performance obtained with a scalable wavelet-based video codec that adds redundancy using multiple description coding with the equivalent system adding redundancy at the transport level, using forward error correction