Video Multicast Over Wireless Ad Hoc Networks

Existing video multicast routing protocols in wireless ad hoc networks have been developed under the assumption that destination nodes wish to receive all the information sent by the multicast source, i.e., they do not support heterogeneous destinations. This paper addresses the problem of video multicast for heterogeneous destinations in wireless ad hoc networks. Multiple Description Coding (MDC) is used for video coding. MDC generates multiple independent bit-streams, where the multiple bit-streams are referred to as multiple descriptions (MD). Furthermore, MDC enables a useful reproduction of the video when any description is correctly received. Specifically, we propose three novel multiple multicast trees routing protocols. The first protocol constructs multiple disjoint multicast trees and assigns MD video in a centralized fashion, and is referred to as Centralized MDMTR (Multiple Disjoint Multicast Trees Routing). The second protocol is a variant of Centralized MDMTR. We refer to it as Sequential MDMTR. The main difference between Sequential MDMTR and Centralized MDMTR is that, Sequential MDMTR sequentially assigns MD video to the destination nodes. In order to reduce construction delay and routing overhead, we further propose Distributed MDMTR protocol. Both protocols, Centralized MDMTR and Distributed MDMTR, exploit the independent-description property of MDC along with multiple disjoint paths to increase the number of assigned video descriptions to each destination. We extensively evaluate our proposed protocols by simulations and show that they outperform the existing work. Key-Words: Video multicast, wireless ad hoc networks, multiple description coding, heterogeneous destinations