A Collaborative Cache Strategy Based on Utility Optimization

With the continuous development of network technology, the numberof streaming media videos is growing rapidly. More and more users are watchingvideos through the Internet, which leads to the increasing huge server load andthe increasing transmission cost across ISP domains. A feasible scheme to reducetransmission cost across ISP domains and alleviate the server load is to cachesome popular videos in a large number of terminal users. Therefore, in this paper,in order to utilize the idle resources of the terminal peers, some peers with goodperformance were selected from the fixed peers as the super peers, which wereaggregated into a super peer set (SPS). In addition, with the supply and demandrelation of streaming videos among ISP domains, a mathematical model was formulatedto optimize the service utility of ISP. Then, a collaborative cache strategywas proposed based on the utility optimization. The simulation results show thatthe strategy proposed can effectively improve the user playback fluency and hitrate while ensuring the optimal service utility.