Ant Colony Algorithm in QoS Network Routing Optimization
|School||East China University of Science and Technology|
|Keywords||QoS network routing ant colony algorithm pheromone Network simulation|
As a result of the rapid development of modern networks, the multimedia applications emerging in endlessly on the Internet, and therefore the network performance requirements are higher. Traditional best-effort delivery service can not fully meet this demand for multimedia data transmission services. Therefore, the quality of service (QoS) routing becomes a trend, which is the future of the Internet core. However, under many constraints, QoS routing problem is NP-hard to solve with conventional methods.Ant colony algorithm is derived from a large new heuristic search algorithm in nature, with the robustness, parallelism, flexibility, robustness and fast convergence characteristics. Thus, although it appears later than the genetic algorithms, simulated annealing, tabu search algorithm, but shows a strong advantage in solving complex optimization problems, especially when discrete optimization.In this thesis,we describe the characteristics of QoS routing algorithm first, and then details the basic principles of ant colony algorithm, the algorithm processes, parameters and characteristics of ant colony system algorithm advantages and disadvantages.Base on the conclusion of the ideas and methods improved by the predecessors, put forward improvement strategy, this are outlined as follows:1. the improvements about pheromone; the setting initial threshold of pheromone, the update strategy and the improvement about the concentration of pheromone on the path around the destination node. 2 the share information mechanism design when ants meet enhanced ant cooperation.3. Improved the original node selection strategy is proposed to limit the maximum and minimum concentration of pheromone, to ensure the diversity of the solution. Generate random network through simulation and testify the feasibility and effectiveness of this algorithm in the QoS routing.