Dynamic source-channel rate-distortion control under time-varying complexity constraint for wireless video transmission

In recent years, a number of rate-distortion (R-D) optimization schemes have been proposed. These works discussed the joint source-channel rate allocation problem for achieving best end-to-end video transmission quality under limited transmission bit rate. However, encoding complexity, a critical concern for video transmission from mobile devices, was hardly considered. Especially, the time-varying nature of computational resources for mobile devices was largely ignored. In this paper, an on-line adaptive video encoder parameters control scheme is proposed. Simulation results show that the proposed scheme, by dynamically adjusting the encoder parameters in H.264 using the precomputed rate-complexity-distortion (R-C-D) tables, achieves low end-to-end distortion under time-varying channel condition and dynamic computational complexity constraints.