Research On Small World Based P2P Self-Orgnized Resource Search Algorithm

In recent years P2P application grows rapidly and has become world wide hotspot of computer network research. Shared resource searching is the key-point of P2P technology. How to find resources rapidly and effectively and how to find QoS qualified resource with the lowest cost is the sticking point of P2P applications. Current searching algorithm such as Gossip is blindfold when inquiring resources. So the efficiency is not high and it may lead to network split when the network scope is large. Also most searching algorithms don't consider resource quality and don't select resources according to factors such as network bandwidth, node processing ability. So they are not proper for QoS application. It is necessary to research new resource searching algorithm.This paper researches on how to improve resource searching algorithm efficiency and speed against current algorithms. Network nodes become high congregated and have low characteristic path by creating self-organized P2P overlay network topology with small-world property. A hierarchy resource searching algorithm is brought forward upon this topology. Emulation experiment shows that hierarchy resource searching algorithm is good at Gossip and congregation priority algorithm in network cost, searching result satisfaction and responding time. Hereinto network cost is 80% and 70% lower compared to Gossip and congregation priority algorithm respectively. Searching result satisfaction is 20% and 10% higher respectively. Responding time is 40% and 15% lower respectively. So this hierarchy resource searching algorithm could be used in any scope unstructured topology P2P network.An improved ant cycle system is brought forward by dynamically adjusting hormone thickness. It solves the issue that standard ant cycle system has slow convergence speed and is apt to get local optimized resource. The result of standard TSP problem shows that improved ant cycle...