Research Of The Peer-to-Peer Adaptive Routing Protocol Based On Ant Colony Algorithm

ABSTRACT: Communication is the core component of computer network antagonizingsystem, and routing protocol is the core component of communication. Now there are twoprotocols in Peer-to-peer Network: Structured protocol and unstructured protocol. For the routingof unstructured topology, the higher of its adaptive level，the smaller of its routing spending, thehigher of its hit rate. The using of adaptive routing based on biological principles can adapt to thenetwork’s randomness better.Ant colony algorithm is a sort of swarm intelligent search algorithm, which inspired fromsocial insects in natures. Swarm intelligence refers that a single intelligent individual can onlycomplete fairly simple task，while the cooperation of the entire agent population is able toaccomplish complex tasks well. A typical example of swarm intelligence is ants find food.Because Ant colony algorithms and peer-to-peer network have similar sell-organizationcharacteristics, so this paper’s focus is peer-to-peer adaptive routing protocols based on antcolony algorithm.Since the course of routing information finding object is similar to the course of the antsfinding food，so an ant peer-to-peer adaptive routing protocol model will be made in the paper. Inthis model, we will exchange information to body mapping between ants，and we will exchangethe different types of message maps for different ants, the pheromone routing table mappingtable. At the same time, we will copy the message routing process map for the cloning of antrouting process. Every times when we clone the ants, the number of ants (antNum)will beproduced of adaptive routing number of ants clone, while ttl will be produced of cloned adaptivelife cycle of ants, also we must control the size of the entire colony.Based on APPRM，an improved ant adaptive routing algorithm will be presented. On thisbasis，the design of ant adaptive routing protocol will be given, and using the object-orientedsoftware engineering ideas and methods to realize the system.Some test results of use cases show that for any valid routing on the peer are successful inthe AARPS system, and always can be find an optimum path in the effective time for the task ID,and the successful delivery of a message to the source peer.