Heuristic copy-nodes allocation algorithm for multicast network design using neural computing

An efficient scheme for establishing the multicast route on the packet-type networks is to connect the tree-shaped path in which the multicast packets are replicated at specific branching nodes with the packet copy function, called “copy-nodes”, and forwarded to the downstream nodes. This scheme leads to the efficient usage of the network resources, but needs the multicast routing control, especially. In order to realize the rapid multicast routing, it is suitable and realistic to restrict the number of copy-nodes in calculating multicast route. However, even in this approach we need to adequately select the restricted number of copy-nodes which satisfy the conditions of both the priority for multicasting and the geographical distribution at the same time. In this paper, a new copy-nodes allocation algorithm is proposed, which greatly improves the neural approach called the “elastic network”. This heuristic algorithm makes some “copy-node functions” move on the network map sequentially according to the attraction and repulsion determined by several conditions, and finally makes them converge on the switching nodes suitable for the copy-nodes. In addition, the performance of this algorithm is evaluated through the computer simulation