Research On Multicast QoS Routing Algorithms Based On GA In Ad Hoc Networks

Ad Hoc networks are a self-organized multi-hop network with highly dynamic topology and random node. It has broad prospects in the field of military and civil affairs. With the popularization of multi-media and real-time projects, QoS technology is becoming more and more essential in Ad Hoc networks. Multicast can reduce the load of network and server, and improve the transmission performance effectively. However, because of the characteristic that network topology changes frequently and node energy is limited in Ad Hoc networks, the way how to design a multicast routing which meets the requirement of QoS becomes the key and nodus in research of Ad Hoc networks.Genetic algorithm (GA) is a new search-optimization algorithm which is highly parallel and seeks global optimization randomly. Recently, it has been widely used to solve the problems of QoS multicast routing. Due to the frequent mobility of nodes and poor stability of routing in Ad Hoc networks, this paper designs a QoS multicast algorithm (EMQGA) based on information entropy and GA. It utilizes information entropy as the criterion parameter of routing stability to select routing which is long-living and stable. Depending on GA, it shapes a stable multicast tree to meet the requirements of QoS. Besides, it improves the coding mechanism of traditional GA and designs proper operation of interlocking and variation which suits coding scheme. The utilization of parameter control method helps the algorithm converge steadily while seeking the global optimization. Regarding to the problem of node energy limitation of Ad Hoc networks, a QoS multicast genetic algorithm (BMQGA) is brought out. The application of GA can optimize the trees which is required by little cost and strong energy. The coding mechanism applies improved ST sequence topology code which reduces the complexity of operation on multicast tree coding. The interlocking scheme of one point and double points cross utilization increases the diversity of GA population and enhances the searching capability of the algorithm. Based on variable manipulation of energy-fix function, this algorithm balances the cost of node energy in multicast trees and extends the existing time of networks.The research implements experiments and analyzes performances through simulation. Results prove that the algorithm converges rapidly, performs well and are able to find stable multicast trees which meet the requirements of QoS and have large amount of excess energy.