Improving Unstructured P2P Systems by Partial Indexing

Resource discovery in structured and unstructured peer-to-peer (P2P) systems have been extensively studied. However, most of such systems are not designed to take advantage of node heterogeneity. Previous researchers have obtained the result that, to keep system performance high, the resources of nodes should serve queries in different ranges according to their different characters. Based on the above result, we propose a novel Partial-indexing architecture to improve search efficiency of P2P systems. Simulation results show that the suggested architecture can get high query success rate with low search cost.