Heuristic algorithms for packing of multiple-group multicasting

Abstract

Multicast communication is an efficient routing method for multimedia data distribution, since it can save network bandwidth during the communication session. Thus, the multicast routing problems have received much attention from many researchers. In this paper, we consider a multicast routing problem with multiple multicast sessions under a capacity limited constraint. This problem is formulated as a tree packing problem. We propose two heuristic algorithms, Steiner-tree-based heuristic (STH) algorithm and cut-set-based heuristic (CSH) algorithm, for solving this problem. The simulation results show that the STH algorithm can find a better approximate solution in a shorter computation time compared to CSH. In addition, if the available bandwidth for the service is just enough, the STH and CSH algorithms may fail to find a solution even if the solution exists. The simulation results also indicate that CSH has a higher probability than STH to find a solution. Thus, it is suggested that one can apply the STH algorithm first to solve the tree packing problem. In case STH fails, CSH algorithm will be used instead. Scope and purpose Several new applications in multimedia networks have been developed in recent years, such as video conferencing, video on demand and network TV. The routing problem for these multimedia services is primarily concerned with efficiently delivering data to multiple destinations. Multicast communication which delivers data along a tree is a preferred routing method because it can reduce network traffic and save network resources, e.g., bandwidth. Most of the current researches for the multicast routing problems only consider a single multicast session. However, several multicast sessions may occur simultaneously in a network and these multicast sessions will contend for the limited network bandwidth. This creates a new network optimization problem. In this research, we studied an optimal tree packing problem. The goal of this study is to arrange the multiple multicast trees in the network such that the bandwidth constraint is maintained and the overall transmission cost is minimized. We believe that the results are useful for improving network utilization in multimedia networks which have multicast services.