Enhanced Chunk Regulation Algorithm for Superior QoS in Heterogeneous P2P Video on Demand

Abstract

Traditionally, Video-on-Demand (VoD) systems were based on servers that streamed video contents to the end users. Later, data replication using Content Delivery Networks (CDN) came into existence; this reduced the overall network load by effectively distributing contents to end user. However, replication using CDN is relatively costly. Today, Peer-to-Peer (P2P) based VoD systems have gained wide popularity due to its robustness, scalability and availability. P2P networks consist of a large number of nodes, including mobile devices that join and leave the network frequently. The challenge is to provide better video Quality of Service (QoS) to the end user by making requested copies of video chunks available on demand. In this paper, we propose an Enhanced Chunk Regulation Algorithm (ECRA) for superior video QoS in P2P networks. We propose a chunk selection algorithm called Local Demand based Chunk Download (LDCD) that fulfills the temporal demand of chunks. Our proposed Peer Ranking Factor (PRF) algorithm uses ranking factor to choose the best suitable peers to download video chunks. Additionally, we present Local Demand based Chunk Replication (LDCR) algorithm that considers the demand for particular videos (i.e. movies) in the local cluster. Based on the available bandwidth and cache space the replication of chunks is accomplished such as the QoS of VoD is superior. Our evaluation shows that the proposed algorithm improves the overall end user’s QoS by minimizing missed chunks, stalling delay and startup delay. Moreover, it provides more effective bandwidth utilization and minimizes the load on the source server