Exploiting Caching and Prediction to Promote User Experience for a Real-Time Wireless VR Service.

In this paper, we propose a novel wireless virtual reality scheme by caching resource on head-mounted displays (HMD) and exploiting head movement prediction to improve user experience. We render images for the future request based on the useru0027s head movement prediction at the server-end to reduce the motion-to- photon latency and cache those rendered images on the HMD to form an image pool. We then develop a corresponding algorithm to select and warp the exact image from the image pool for display. Furthermore, a performance metric - warping distance is defined and used to evaluate the image quality of the proposed scheme. Finally, real dataset-driven results show that the proposed scheme is able to provide higher image quality as well as experience consistency compared with existed schemes.