Network Performance Analysis of Startup Buffering for Live Streaming in P2P VOD Systems for Mesh-Based Topology

Abstract

This paper explores mesh-based clustering for different start video streaming in P2P systems and estimates the performance of noncluster and clustered models. These models are based on mesh-based topology of P2P streaming consisting of peer join/leave. A new approach by way of “clustering” peers is proposed to tackle P2P VOD streaming. The proposed models were simulated and verified using OMNET++ V.4. A clustered model for video streaming is proposed and simulated to consider the performance of network under startup buffering for frame loss, startup delay, and end-to-end delay parameters. The results obtained from simulations are compared for both noncluster versus cluster models. The results show the impact of startup buffering on both models is also bounded due to time limits of release buffer and playing buffer under the proposed models, which causes reduction in wait time to view video improving the overall VOD system performance. The proposed model is also able to provide missing parts (of video) to late viewers, which gives the facilities of both live and stored streaming from user’s point of view, therefore it serves to be functionally hybrid and is most useful.