Performance Evaluation of an Object Management Policy Approach for P2P Networks

Abstract

The increasing popularity of network-based multimedia applications poses many challenges for content providers to supply efficient and scalable services. Peer-to-peer (P2P) systems have been shown to be a promising approach to provide large-scale video services over the Internet since, by nature, these systems show high scalability and robustness. In this paper, we propose and analyze an object management policy approach for video web cache in a P2P context, taking advantage of object's metadata, for example, video popularity, and object's encoding techniques, for example, scalable video coding (SVC). We carry out trace-driven simulations so as to evaluate the performance of our approach and compare it against traditional object management policy approaches. In addition, we study as well the impact of churn on our approach and on other object management policies that implement different caching strategies. A YouTube video collection which records over 1.6 million video's log was used in our experimental studies. The experiment results have showed that our proposed approach can improve the performance of the cache substantially. Moreover, we have found that neither the simply enlargement of peers' storage capacity nor a zero replicating strategy is effective actions to improve performance of an object management policy.