Improving Mobile Video Streaming with Mobility Prediction and Prefetching in Integrated Cellular-WiFi Networks

Abstract: We present and evaluate a procedure that utilizes mobility and throughput prediction to prefetch video streaming data in integrated cellular and WiFi networks. The effective integration of such heterogeneous wireless technologies will be significant for supporting high performance and energy efficient video streaming in ubiquitous networking environments. Our evaluation is based on trace-driven simulation considering empirical measurements and shows how various system parameters influence the performance, in terms of the number of paused video frames and the energy consumption; these parameters include the number of video streams, the mobile, WiFi, and ADSL backhaul throughput, and the number of WiFi hotspots. Also, we assess the procedure's robustness to time and throughput variability. Finally, we present our initial prototype that implements the proposed approach.