A Structured P2P System with Match Path and Probability Balance Tree

Excellent P2P network infrastructure should support scalability and efficient search simultaneity. The intelligent routing mechanisms have flexible query expression but hold low scalability and recall rate. Contrastively, DHTs are scalable but either difficult to organize or too expensive for maintenance. A new structured P2P model is proposed in which resource storages are organized by match path and random access method. When an approximately balance distributed character is obtained, nodes will keep high search efficiency and low maintenance cost. The system supports various query models, has load balancing capability and high scalability, and ensures search range of each node can cover entire network. The system runs without any center service while all nodes simultaneously take index storage, data storage and routing task to drive network by only maintaining a little local information