Supporting Content-Based Music Retrieval in Structured Peer-to-Peer Overlays

Abstract

The advent of data-sharing applications has made peer-to-peer systems popular for the widespread exchange of resources and information among millions of users. Recently, structured peer-to-peer networks have been proposed to guarantee file availability, ease network load, and support keyword queries efficiently. However, keyword-only searches are insufficient for users. This study aimed to solve the problem of how to support content-based music retrieval in a structured peer-to-peer overlay. Based on a modified hypercube structure, a ternary cube model named Hyper-Ternary-Cube was proposed to support content-based music searches with error tolerance in a peer-to-peer overlay. Moreover, a set of simulations was performed to verify the efficiency and effectiveness of the proposed structure. The simulation results indicated that the proposed method processed pin queries efficiently.