Hierarchical random linear network coding for multicast scalable video streaming

Network coding (NC) has been exploited to minimize delay, energy per bit and to maximize throughput over lossy networks. However, applying network coding for scalable video transmission over packet-lossy networks is still an active research topic because of the challenge in trading off between rank deficiency problem and scalability of the video bitstream. Scalable video coding promises providing flexible broadcasting to heterogeneous user equipments, but the rank deficiency problem in NC limits the scalable capability in video streaming since the time of waiting for full rank causes long delay. In this paper, we exploit the combination of hierarchical and random linear network coding for solving the above issue. The hierarchical network coding provides unequal erasure protection for scalable layers and addresses the process of decoding layer by layer according to importance. Otherwise, random linear network coding provides error-correcting capability and makes the scheme resilient to packet loss. Our proposed network coding scheme benefits from these two schemes for multicast scalable video streaming. The simulation results and performance analysis are provided to evaluate the efficiency in terms of averaged decoding probability and peak signal-to-noise ratio. The proposed scheme achieves same performance of random linear network coding in good channel and approximately hierarchical network coding scheme in bad channel.