Cloud Assisted P2P Media Streaming for Bandwidth Constrained Mobile Subscribers

Abstract

Multimedia streaming applications have disruptively occupied bandwidth in wire line Internet, yet today’s fledging mobile media streaming still poses many challenges in efficient content distribution due to the form of mobile devices. At the same time, cloud computing is gaining power as a promising technology to transform IT industry and many eminent enterprises are developing their own cloud infrastructures. However, the lack of applications hinders clouds’ large-scale implementation. In this paper, we envision a cloud-assisted power-efficient mobile P2P media streaming architecture that addresses the weakness of today’s wireless access technologies. Clouds are responsible for storage and computing demanding tasks, and mobile devices colocating with each other share bandwidth and cooperatively stream media content to distribute the load. We first model interactions among mobile devices as a coalition game, and then discuss the optimal chunk retrieval scheduling. Finally, we draw on realistic mobile phone data and utilize an ARIMA model for colocation duration prediction among mobile devices.