Unequal loss-protected multiple description coding of scalable source streams using a progressive approach

An analysis-based approach for unequal loss-protected multiple description coding (packetization) of the scalable (prioritized / progressive) source code streams is proposed. For a given number of packets (descriptions) of the known size, unequal loss-protected packetization leads to segment the scalable code stream, such that the source can be reconstructed with the maximum possible fidelity at the decoder side. Here, we find an analytical relation between optimal sizes of any two consecutive segments. This idea yields a low-complexity progressive solution with a performance close to that of local search, which has been approved as an efficient method to solve the segmentation problem. Simulation results are used to confirm the efficiency of the proposed method as compared with the local search algorithm.