Loss-tolerant real-time content integrity validation for P2P video streaming

The use of peer-to-peer (P2P) mechanisms for content delivery is attractive to content and service providers alike. P2P data transfers offload the demand on servers and reduce the bandwidth requirements, with corresponding benefits of improved scalability and performance. This, however, poses interesting challenges in ensuring content integrity. Peers may be malicious and attempt to send corrupt and/or inappropriate content to disrupt the service. Consequently, service providers must provide clients with the capability to validate the integrity of content delivered from peers. This goal is particularly challenging in the context of streaming video because the content needs to be validated in real time. A practical solution must provide high integrity assurance while incurring low communication and computation overhead. In this paper, we present a packet-based validation approach for ensuring the integrity of data obtained from peers. Our proposed scheme randomly selects packets and validates their correctness. Through detailed experiments, we show that this mechanism is not only lightweight but is also able to detect content corruption with very high probability, thus protecting the viewing experience and the provider´s content delivery service.