Network Resource Allocation for Scalable Video Streaming over P2P Networks Based on Game Theory

Abstract

In order to adapt to the heterogeneity of terminals and networks, Scalable Video Coding (SVC) encodes raw video stream with different scales of temporal, spatial and quality into layers. Considering the P2P network characteristic, it is a challenging task to design an appropriate P2P steaming network resource allocation mechanism combining with SVC. In this paper, SVC is applied in P2P streaming based on game theory; considering free-riding, bandwidth conflicts in P2P multi-overlay and one chunk with multiple providers, we design a bidirectional serial auction model that jointly optimize the bandwidth allocation, the data scheduling and the incentive mechanism, then optimized allocation for scalable video streaming over P2P networks is achieved. With extensive theoretical analysis, we show that these games converge to an optimal topology for each overlay.