Research And Design On QoS Routing Algorithm Based On Ant Colony Algorithm

With the development of the Internet, people have more diverse needs with immediate transmission. In order to improve the quality of network's service, many scholars and technicians began to study the algorithm of QoS routing. The Solution of QoS routing is a NP-C problem. The traditional algorithm has been unable to solve this problem. In this case, the ant colony algorithm are suggested to solve this problem. Ant colony algorithm is a kind of swarm intelligence heuristic approach that inspired from ants finding foods. The feature of ant colony algorithm is that one agent can only do some easy jobs, and very complicate tasks must accomplished by the cooperation of each agent in a whole colony. Ant colony algorithms have widely applied to solve many combinatory problems now. Ant colony algorithm supports distributed computing and multi-path, so it just meets the need of network routing.The article studies the QoS routing algorithm based on the ACO algorithm and the QoS routing algorithm based on the ACO-GA algorithm. It explores the shortages of the algorithms. The paper analyzes the reason for that the QoS routing algorithm based on the ACO algorithm is slowing and premature stagnation. The article analyzes and points out that although the QoS routing algorithm based on the ACO-GA algorithm is a combination of two intelligent algorithms,but the algorithm only search for routes in the non-linear integer programming model blindly and its intellectualized bionic technology is not combining with the feature of the problem. Two kind of intelligent algorithm's union cause the uncertainty of algorithm to increase and produces the dead turn easily in the routing process. To overcome the default of these algorithms, the boundary restriction ant colony algorithm and a new QoS routing algorithm named FALP is presented. The boundary restriction ant colony algorithm retains the random number adjustment switching rules technology and joins the boundary restriction parameter to prevent the information element density excessive growth, it has avoided the precocious stagnant occurring effectively. FALP is based on network topology's characteristic and combine the boundary restriction ant colony algorithm and floyd, it solves closed loop problem and accelerate the rate of solution, it raises the efficiency. Simulation and experiment results prove that FALP can get a better solution than some other algorithms based on the ant colony algorithm. FALP converges faster and more stable.