Expandable resource lookup method in peer-to-peer network with structure
暂无分享,去创建一个
The invention relates to a method for searching scalable resources in a structured peer-to-peer network which pertains to the technical field of computer network and solves the problems of high search cost and low learning mechanism efficiency of the existing method for blind search in the structured peer-to-peer network so as to reduce search cost and lead to more efficient resource search. Information nodes of the invention are interconnected by distributed hash tables, and each information node maintains a local knowledge base and the local knowledge bases conserve index record. The method of the invention comprises the steps of searching local resources, transmitting searched messages, index updating and feedback. The method of the invention makes full use of the peer-to-peer network technique and machine learning mechanism, therefore, the method has very good scalability, high learning algorithm efficiency, low index cost with space cost of 0(log N) magnitude and short training process and can continuously approach the distribution state of actual recourses, obviously reduce the network cost of scalable blind search algorithm and self-adapt to dynamic join or withdrawal of resources and other situations. Besides, the performance can recover after short fluctuation.