Peer-assisted video on-demand streaming system in practical WiFi-based mobile opportunistic networks

Network convergence paradigm will substantially increase the pervasive use of WiFi-enabled mobile devices such as smart phones (e.g., Android-based mobile and iPhone), e-book viewers (e.g., iPad and Kindle) and personal media players. Although various on-demand streaming services are already available over mobile WiFi-enabled devices, it remains a challenging problem due to WiFi's limited communication range, mobility and user population density issues. In this paper, we extend and enhance our previous work on MOVi (Mobile Opportunistic Video-on-demand). In particular, we exploit the use of cooperative content sharing concept in gaining performance advantages. We reinvestigate the previous version of MOVi in more realistic operational environments and propose an improved scheduling algorithm which incorporates H-hop blocking and opportunistic download skipping schemes. The scalability of the extended MOVi system is verified by extensive simulations and several experiments with prototype implementation over OMF (cOntrol and Management Framework)-enabled wireless network testbeds. In terms of the number of supported user, an average of 30% improvement can be achieved while at the same time consuming the same amount of battery power.