Date Added: Sep 2010
Efficient resource discovery and availability improvement are very important issues in unstructured P2P networks. In this paper, a bio-inspired resource discovery scheme inspired by the principle of elephants migration is proposed. A replication scheme based on Q-learning and erasure codes is also introduced. Simulation results show that the proposed schemes significantly increase query success rate and availability, and reduce the network traffic as the resources are effectively distributed to well-performing nodes. P2P networks are a recent addition to the already large number of distributed system models. P2P networking has spawned immense attention worldwide amongst both Internet users and computer professionals.