Fog Computing Aided Multi-View Video in Mobile Social Networks

Multi-view video (MVV) consists of multiple video streams captured simultaneously by multiple closely spaced cameras, and it enables users to freely change their viewpoints by playing different video streams. However, the network transmission delay of multiple video streams from certain video sources to the base station via the core network are different, which results in the asynchronous among the video streams when users switch streams. It tremendously degrades user Quality of Experience (QoE). Considering the social characteristics of MVV users in terms of spatially clustering and the similarity of interests on MVV streams, we introduce the edge caching technology in fog computing into the application of MVV in mobile social networks (MSNs), with which the MVV streams can be synchronized with the assistance of edge caching among local users. Besides, we model the spatial distribution of edge caching users to calculate their capability of edge caching and D2D communication, as well as the coverage probability and Ergodic rate of the multicast groups. Moreover, the edge caching user selection is formulated as an optimization problem to maximize the system throughput, and a greedy based edge caching algorithm is proposed to find the suboptimal solution. Simulation results indicate that the proposed edge caching scheme can significantly increase the QoE of MVV and the system throughput.