A Framework for Video Network Coding with Multi-generation Mixing

Abstract—Multi-Generation Mixing (MGM) is a generalized approach of Network Coding (NC) that has its improvements when applied in packet loss networks. In this paper we apply MGM network coding in networks communicating video contents. NC has shown viable improvements when applied in packet loss networks. With practical network coding (generation based network coding), packets are grouped in chunks called generations. Generation is the unit of network coding encoding and decoding. The generation grouping of packets is necessary for the practical deployment of network coding. On the other hand it increases the cost of NC losses. NC losses reduce the ability of receiver to decode packets, and hence can severely degrade the quality of recovered video. Video is encoded in layers with Scalable Video Coding (SVC) a base layer and one or more enhancement layers. MGM employs the layering of scalable video to enhance the reliability of communication. MGM provides different levels of reliable communication for the different video layers to improve the overall reliability of video communication. SVC enhancement layers are dependent on lower layers to be recovered. With MGM enhancement layers support the recovery of lower layers. This is done by network encoding lower video layers in higher layers. Through extensive simulations, we show that MGM highly improves the quality of recovered vide