A tabu-search based algorithm for the multicast-streams distribution problem

• Authors: Wen-Lin Yang
• Venue: Computer Networks: The International Journal of Computer and Telecommunications Networking
• Year: 2002

Consider a video library system on campus, there are usually many classes demanding on the same and/or different educational video programs provided by the video server at the beginning of a session. A near video on demand system with multicast communication technology would be a proper implementation for this video library system. As a multicast stream carries one video program, more than one multicast streams may flow on the network simultaneously. At the network resource reservation step, however, it is possible that the link bandwidths are not large enough to accommodate all the multicast streams if we allocate link bandwidths to each multicast stream one at a time. As a result, some clients cannot be served because of the lack of link bandwidths. Hence, it is important to learn how to utilize these limited bandwidths more efficiently by taking all the multicast streams into consideration altogether, such that the number of clients is served as large as possible. In the past, a branch-and-bound based heuristic algorithm has been proposed. However, this algorithm is not stable and only efficient for a small-scale video library system.In this study, we propose a tabu-search based algorithm to solve this multicast-streams distribution problem. For the acyclic digraphs and general digraphs with simple cycles, the experimental results show that the approximation solutions found by our method are very close to the exact solutions obtained by the branch-and-bound based algorithm. As for the executing performance, our approach is shown to be much more stable and efficient than the branch-and-bound based algorithm. Hence, the procedure proposed in this paper is a practical and highly effective procedure for the problem.