Research Of QoS Multicast Routing Based On Genetic And Ant Colony Algorithm

With the development of the network techniques, some real-time multimediaapplications appear continuously, and gain a fast development. The"best-effort"serviceswhich the current Internet provede can not satisfy the users’ demands proposed intransferring business flow. Like the fast development of current Internet, the network nodesincrease rapidly at exponent level. The NGI is a new large-scale networks for high speed,high capacity and high reliability. In this network, no doubt many real-time multimediaapplications gain a development space. This application usually have various(a mixture of)date type, not a singl datetype; Like video on demand, the datetype it transmitted includingimages, sound and text and so on. There are many real-time multimedia applications likevideo on demand, such as network conferencing, IP telephony, Multiplayer online games.all of this application have a common characteristic is that their data is from asource node totransfer date to multiple destination nodes. In orde to support these real-time multimediaapplications better, it is necessary to determine a multicast tree of minimal cost to connectthe source node to the destination nodes subject to delay constraints. How to use multicasttechnology to provide QoS to this application? It’s a major intrest of network routing. baseon genetic algorithm and ant colony algorithm, we presents a new algorithm to solve theQoS multicast routing problem. Recent studies of QoS multicast routing prove ant colonyalgorithm is a good heuristic algorithm to solve the QoS mutlicast routing problem.In this paper, we design a new QoS multicast Routing algorithm based on geneticalgorithm and ant colony algorithm for real-time multimedia applications communication.At the first of the algorithm, we use genetic algorithm to determine the k shortest paths withbandwidth constraints from a single source node to multiple destinations nodes. Thealgorithm uses the connection matrix of a given network, and the bandwidth of the links toobtain the k shortest paths. Then with these k shortest paths, we use the ant algorithm wepresent to find the multicast tree that minimizes the total cost. By comparing the results theexpermintal results show that the algorithm can find optimal solution quickly and has agood scalability.