Research On Routing Optimization And Flow Balancing Based On PSO

As the network in many fields of application,quality of service provided by network which presents the fast growth is put forward higher request. QoS routing and Flow balance has been proved a NP complete problems, therefore, many scholars apply such as particle swarm algorithm, the ant colony algorithm and genetic algorithm of intelligent algorithm in such problems to seek the optimum solution of the problem. Intelligent algorithm in network applications has become a hotspot, and particle swarm algorithm of application is also a study of important aspects.PSO is a kind of swarm intelligence algorithm and heuristic global optimization technique, whose individual will flock with no quality and no volume, is a potential solution to solve the problem, and rules stipulated algorithm can show complex characteristics in the simple behavior. PSO has fewer set parameters and computing speeder and easier to realize, etc. These make it become a kind of simple and effective in the treatment of random algorithm, constraint condition problem than traditional algorithm to more flexible. At present more and more network application needs to ensure the quality of service, the target by routing algorithm of traditional find a shortest path for more change more optimal path constraint. Based on the minimal hop due to delay or simple routing algorithm in network cannot have satisfied with the constant change of the quality requirements and different types of traffic demand, the application must be sought by routing optimization path will meet the constraints of grouping nodes, which aim to push the performance requirements can realize network load balance, etc.Before the related information of PSO and the routing algorithm based on PSO are proposed, relationship matrix as a kind of encoding method is put forward to deal with routing optimization and flow optimization by PSO. Therefore the position of a particle must be relationship matrix in order that PSO could performance well while dealing with the problems. The kind of encoding method can solve the problem of routing optimization and load balancing in the application of PSO , that is , it can reduce the redundant space generating and redundant searching with making little change of using that coding method.