The Application Of Ant Colony Algorithm In Network Routing

With the rapid development of computer networks,network applications become more and more diversified,and the network structure also become more and more complex.People's requirements for the quality of the network are also getting higher and higher.So here comes the QoS(Quality of Service),which is recognized as a new indicator to measure the network transmission.In order to guarantee the QoS requirement of the network,it is the real key to choose the appropriate QoS routing algorithm.And QoS routing is completely a NP-C(Non-deterministic Polynomial complete problem)problem,which is difficult to figure out by using the traditional algorithm.Network congestion is also a big problem under the situation of today's large-scale multimedia applications.In the process of network routing discovery,it is also prone to the loss of data,and even the paralyzed network due to congestion.Those problems in these network routing are the hot spots and difficulties of our research today.According to the requirements of network routing,scholars have proposed the use of Ant Colony Optimization algorithm to solve such problems.Ant Colony Optimization algorithm is a swarm intelligence algorithm,which mainly makes use of ants' information transmission with the environment in the search process,thus finding the optimal path.It is a positive feedback mechanism,and can finally reach the optimal path through the constant update of pheromone.This algorithm has the characteristics of distributed,random and adaptive,and is suitable for solving NP-C problem.So it can be used very well in QoS routing.At the same time,the feedback mechanism of Ant Colony Optimization algorithm can be applied to the process of optimization.It can help to bypass the congestion section and reduce the load of congestion network,thus solving the problem of network congestion well.This thesis analyzes the situation of research and development of network routing and Ant Colony Optimization algorithm at home and abroad.This thesis also analyzes the existing technologies of network routing such as transmission mode,QoS routing,routing model,basic algorithm of routing and some related protocols of network.Then,the principle,model,parameters,implementation steps,advantages and disadvantages of the Ant Colony Optimization algorithm are studied in depth.The main problems of this algorithm are also studied such as the slow convergence speed and the phenomenon of stagnation and local optimization.In this thesis,we study the applications of the ant colony algorithm to network routing,and analyze the stagnation and local optimization problems that exist mainly in these applications and the use the path finding in network congestion.the ant colony algorithm in the application of network routing proposed a corresponding improvement program: the first for the ant colony algorithm,there are some problems in the application of network QoS.The combination of queen ant and the ant colony system based on the optimized ranking is added to the basic ant colony algorithm,so as to increase the search range of the network and its randomness and reduce stagnation and local optimization in network routing.Secondly in view of the problem of network congestion,the basic ant colony algorithm is used to improve the state transition probability,The use of penalties and incentives to improve the pheromone update strategy to enhance the chances of reducing congestion,improve network communication efficiency,reduce packet loss rate,so that the entire network to achieve a certain network load balancing,will not There is a lot of data loss in multimedia applications.This thesis simulates the experimental design of the improved ant colony algorithm through the NS2 simulation platform and the basic ant colony algorithm and network congestion with the Dijkstra algorithm.From these,we can clearly see that the network delay,Jitter and price have been improved by improving the application in the network QoS routing,while reducing the local optimization and stagnation.As to the network congestion,it can clearly see that the network load and delay,packet loss,throughput have been improved.The loss of network data is significantly less than before at the same number of iterations and under load,without any effects on the transmission of data and the use of resources.