A method of the Free rider Detection in peer-to-peer Video Streaming Networks Using Convolutional Neural Networks

Due to the effective distribution of multimedia content between peers, peer-to-peer networks are a suitable platform for video streaming and file sharing, and peer-to-peer networks distribute shared data throughout the network. The analysis of these networks shows that all of these peers do not participate in sharing common data and some of them avoid cooperating with the network, which causes a drop in network efficiency. With the increase in the number of peers as well as various network parameters in order to prevent the lack of communication between peers, we need a flexible method to deal with the complexities of the video streaming network. In the proposed method, peers are clustered based on common interests and a cluster head is determined for them. Each peer receives services according to its reputation in the network. Therefore, the proposed approach tries to identify selfish peers using convolutional neural network (CNN) and encourage them to cooperate with each other using reputation mechanism. The simulation results show a 10% improvement in the prediction and detection of selfish peers in peer-to-peer video streaming network.