Application of Fault-Tolerant Mechanism to Reduce Pollution Attacks in Peer-to-Peer Networks

Abstract

File pollution is a recent security threat to peer-to-peer (P2P) file sharing systems. By disseminating numerous polluted files with mismatched or partially tampered contents in the P2P system, the attacker causes users to download unexpected files. This attack is aimed at frustrating users and making them abandon the system. Present researches on combating file pollution have mostly focused on pollution modeling or evaluating the extent of pollution. Only a few researches have proposed effective methods to eliminate pollution attacks, and they are primarily based on reputation systems and blacklisting mechanisms. However, these methods require exchange of significant feedback among the peers in order to identify the malicious peers or polluted files in the system. In this paper, we describe the application of fault-tolerant mechanism used in the redundant arrays of independent disks system to suppress file pollution attacks based on the concept that P2P file sharing systems currently have global file storage systems. We have extended the previously developed Fluid Model to analyze and evaluate the proposed antipollution mechanism. The model accuracy has been demonstrated by performing several simulation experiments; the proposed mechanism could effectively suppress the pollution and successfully decrease the polluted-time exposure of a P2P file sharing system by approximately 40∼60%.