Diagnosis of Content Pollution in P2P Live Streaming Networks

Abstract

Content pollution is one of the challenges for massively deploying live streaming P2P networks in the Internet. As the peers themselves are responsible to retransmit data, there is no trivial solution to this problem. This work presents a new strategy to detect content pollution that employs comparison-based diagnosis to identify modifications on the data stream. A peer compares randomly selected chunks received from its neighbors. Based on the comparison results, peers that transmitted polluted content are identified. The proposed solution was implemented using Fireflies, a scalable and intrusion-tolerant overlay network. Experimental results show that the strategy represents a feasible solution to detect content pollution and causes a low overhead in terms of network bandwidth.