Skip to Main Content
P2P based VoD (Video on Demand) system has low cost, high scalability and high video quality features, which make it become one of the most popular research hotspots and evolution trends. Content distribution policy is the key point of P2P VoD technology. Against the problems of data scheduling policy priority and node performance difference in content distribution, this paper researched the influence of scheduling policies for network performance information, the degree of content scarcity and broadcast emergency, and on the basis, the gradient model based data distribution policy is proposed. Through the utilization of distributed gradient model strategy, sender start strategy and receiver start strategy, each node interacts only with the part of nodes to reduce the conflict caused by the load balance. Experiment proved that our policy improves content delivery speeds and reduces the network load.