Multicast Network Design by the Use of Heuristic Algorithms

In the near future, there will be many multicast services, and they will be a large part of tra c in the network. The characteristics of multicast connections are quite di erent from those of point-to-point connections. However, the conventional network topology is designed for point-to-point connection. If we use the conventional network topology for multicast services, the qualities of services will be bad, or the network construction costs will be high. Therefore, it is required to nd the topology that is t for the future network, where both multicast tra c and point-to-point tra c exist. In this paper, the multicast network design methods using heuristic algorithms are proposed for the area assumed to be the lowest layer of the hierarchical network. Regarding the network as a graph, a network design problem is converted into a combinatorial optimization problem. To solve this problem, two heuristic algorithms, which are Greedy and Simulated Annealing, are used. The results show that heuristic algorithms are better than the conventional methods, which use the basic topologies, such as star, mesh, ring from the viewpoint of the network construction cost.

[1]  Aaas News,et al.  Book Reviews , 1893, Buffalo Medical and Surgical Journal.

[2]  Takumi Miyoshi,et al.  Optimal Hierarchical Structure of Nationwide Multicast Network , 1997 .

[3]  Takumi Miyoshi,et al.  Application of delta-star topology to multicast network , 2000 .

[4]  Hao Chen,et al.  Parallel Genetic Simulated Annealing: A Massively Parallel SIMD Algorithm , 1998, IEEE Trans. Parallel Distributed Syst..

[5]  Takumi Miyoshi,et al.  Optimal hierarchical structure of broadcast network , 1999 .

[6]  C. Reeves Modern heuristic techniques for combinatorial problems , 1993 .

[7]  Shivendra S. Panwar,et al.  Topological design of interconnected LAN-MAN networks , 1992, [Proceedings] IEEE INFOCOM '92: The Conference on Computer Communications.