Layer Optimization for DHT-Based Peer-to-Peer Network
Hierarchical architecture has been found to facilitate effective search in P2P network and ensure system scalability in P2P application deployment. However, the lack of appropriate size ratio of super nodes layer and ordinary nodes layer makes the system search performance far from being optimal. Taking advantage of node heterogeneity, this paper presents a search delay model to characterize the two-layer P2P architecture using DHT at the top level. With the proposed models, optimal layer division can be achieved. Simulation analysis and numerical results validate the proposed model and solution.