Research And Design Of Multipath Routing Algorithm Based On Ant Colony Optimization Algorithm

Now the requirements of the network communication are increasing, efficient data transmission which is ample flow, fast speed and low cost, is the key to the network communication data processing. With the increasingly high degree of military information for fault tolerance and speed of information transmission network are increasingly high requirements. In peacetime as well as early warning radar detection of wartime operational control of the battlefield surrounding environment, voice and video in multimedia transmission problems, network transmission latency and throughput have higher requirements.Traditional network routing algorithm is likely to cause traffic congestion and uneven distribution of network load, Resulting in the lack of support for quality of service, routing easily lead to shock, unable to make full use of network resources and network topology fault resilience poor, to solve the problems mentioned above, the concept of Balanced Routing and Multi Path Routing is proposed.Multi path Routing takes advantage of several paths which meet the service requirements (including the shortest path) to replace one of the shortest path, to achieve the flow equilibrium, so as to improve the network throughput. Due to the randomness of network, as well as the distribution of real physical network, the ant colony optimization algorithm (ACO) has a unique advantage in the network routing issue, closely matches the topology and traffic patterns, therefore, ACO algorithm has good redundancy and fault tolerance in solving the network routing problem. The AntNet is not the only one that can solve routing problems, but it is the only one ant colony algorithm that has reached the first-class computing performance in test and simulation field among numerous ACO algorithms.The existing research often pays attention to the implementation of the Multi Path Routing Algorithm, and lacks of certain effective rules. Existing multi-path routing algorithm research, specifically can be summarized into three:First, choose a number of high-path, use one of the optimal path; the second is selected from a plurality of high-path routing and continuously updated collection feasible, use one optimal path failure to replace the path immediately; Third, multi-path routing algorithm for traffic segmentation this research is now heavy hard. Most of the algorithm parameter values to set quantitative theoretical point of view, the lack of a valid set some rules, but it does not meet the actual network. The parameters of the algorithm are mostly set from the theoretical perspective, however, this does not conform to the actual situation of the network. For different network topology or different service requirements, the parameter values of the algorithm are also different, and the parameter values of the algorithm affect the actual performance of the network directly.This project is based on Ant Colony Optimization that is used for the design of mul ti path routing algorithm, uses ant colony pheromone routing table, data multipath proba bility forwarding and ant search process to measure actively QoS information such as ne twork latency, queue length, bandwidth, jitter, loss rate etc. By using feedback informati on, we can adjust the project dynamically to satisfy different optimization objective, mo difies pheromone table and routing table, makes an effective rule mechanism to guide a nt’s subsequent Iterative evolution. Particularly setting the specific parameter value by a mount of simulation experiment to control the algorithm performance. Makes full use of topology of networks, equilibrates load of each node and each link, avoids network con gestions, improves network throughputs.