Proxy-driven Rate-distortion Optimized Video Streaming over Wireless Network Using Asynchronous Clocks

Video streaming over wireless network is challenging due to node mobility and high channel error rate. In this paper, we propose a joint server/peer, proxy-driven video streaming architecture to support high quality video streaming service over IEEE 802.11-based wireless networks. The video proxy tracks peer mobility pattern, coordinates among multiple senders and performs rate-distortion optimized streaming. We formulate the joint server /peer streaming as a combinatorial optimization problem and introduce the concept of asynchronous clocks to decouple the problem into three steps: first deciding the candidate set of potential senders among the wireless peers based on their mobility pattern, then selecting the sender in each optimization instance using asynchronous clocks, and finally applying point-to-point rate-distortion optimization framework between the selected sender-receiver pair. In addition, we consider two different system setups, no cache and limited cache, to investigate how caching affects streaming performance. To design more realistic simulation models, we use empirical results from corporate wireless networks to model peer mobility. Simulation results show that our proposed joint streaming scheme has better performance than the server-only streaming scheme. Simulation results also show that having cache in the proxy can potentially improve the streaming performance.