EAF: Energy-aware adaptive free viewpoint video wireless transmission

By transmitting views captured using two adjacent cameras in texture plus depth format, any middle view in between could be synthesized with depth image based rendering (DIBR) technique. This provides users with continuous viewing angles for selection and is a key technology for a number of emerging applications such as free viewpoint TV. How to deliver the free viewpoint video contents to clients is one fundamental problem for these applications and the challenges exist in the large bandwidth requirement and frequent change of network environments. Source encoding rates could be varied to fit different network conditions by changing frame rates, with extra battery life consumption due to skipped frames interpolation before synthesizing the requested middle view. Note that battery development is lagging behind mobile devices function and computation capability development. How to trade off the source encoding rate reduction and extra energy consumption has not been formally solved as far as we understand. In this paper, we proposed EAF: an energy-aware adaptive free viewpoint video wireless transmission system, where we jointly considered the source encoding rate reduction and extra energy consumption with network constraints. Simulation results showed that the proposed scheme advantages over competing schemes significantly in typical network scenarios.