Low-complexity and optimal resource allocation scheme for scalable video transmission over realistic noisy MIMO channels

Abstract

International audienceWe investigate a joint source-channel coding solution for H.264/SVC video transmission over noisy MIMO channels. To exploit the channel diversity, we use a closed-loop MIMO scheme, which allows the decomposition of the MIMO channel into several hierarchical SISO subchannels. The scalable video bit stream is divided into hierarchical quality layers (base layer and enhancement layers) which are passing through the SISO subchannels. For each subchannel, an adaptive modulation, an unequal error protection and an unequal power allocation are provided according to the channel state information. The contribution of our work is to provide an optimal and low-complexity method that jointly parameterizes the system elements connected to the video contents and the physical layer. The performance of our proposal is evaluated over a realistic noisy MIMO channel. The results disclose that the proposed solution optimizes the complexity and performs an optimal resource allocation to achieve the best QoS in term of PSNR regardless of the severity of channel conditions