Matching Capabilities of Prediction to Communication and Computing for Proactive VR Video Streaming
share
Proactive tile-based video streaming can avoid motion-to-photon latency of wireless virtual reality (VR) by computing and delivering the predicted tiles in a segment to be requested before playback. However, all existing works either focus on the tile prediction or focus on tile computing and delivering, overlooking the important fact that prediction, computing and communication have to share the same duration. Since the quality of experience (QoE) of proactive tile-based streaming depends on the worst performance of prediction, computing and communication, it is vital to match the prediction capability to the computing and communication capability. In this paper, we jointly optimize the duration of the observation window for tile prediction and the duration used for computing and communication, to maximize the QoE of watching a VR video.We find the global optimal solution with closed-form expression by decomposing the original problem equivalently into subproblems. From the optimal solution we find two regions where tile prediction and computing and communication capabilities respectively play the dominant role, and reveal the tradeoff between the performance of tile prediction and the capability of computing and communication. Simulation results using two existing tile prediction methods with a real dataset demonstrate the gain of the optimized duration over the non-optimized duration of observation window.
READ FULL TEXT