Research On Error Resilience In Wireless Video Communications

Wireless video communications play a significant role in the applications of wireless communication. To obtain a high quality video transmission over the wireless channel is still quite a challenge. In one hand, the video data should be compressed with the residual excess information removed in order to fix the limited bandwidth of wireless channel. In the other hand, because of the high bit error ratio of the wireless channel, the compressed video data should be equipped with error protection for reliable transmission. The improvement of video data compression ratio will decrease the performance when suffering with bit errors, while improving the performance of error-robust will decrease the video data compression ratio. In order to get a balance between the video data compression ratio and the reliability of transmission, this paper focuses on the error resilience technology in wireless video communications.In chapter 1, the importance of the research work is illustrated and the current research statuses of error resilience technologies are surveyed. Then the research work and the architecture of the thesis are introduced.In chapter 2, the key technologies in wireless video communication are studied. Firstly, the various techniques of the prevalent technologies and the standard of the video coding are presented. Secondly, the simulation model and channel coding techniques of wireless channel are illustrated. Finally, the structure of wireless video communication for the paper is introduced. This chapter prepares necessary information for the following chapter in order to study error resilience in wireless video communications.In order to improve the performance of error control in wireless video communications, an error control method based on expected perceptual distortion in video coding is proposed in chapter 3. The proposed method considers the characteristics of the human visual system, the error-sensitivity of the bit-stream, and the state of the time-varying wireless channel jointly. An expected perceptual distortion model is used to adjust the intra refresh rate adaptively. As a consequence, the error propagation is bounded and the video quality is improved. Firstly, the expected perceptual distortion model which consists of motion, skin tone and space position, is utilized to produce the perceptual weight map. Then, the bit-error probability map is calculated based on bit-stream size and the bit-error ratio of the wireless channel. Secondly, the expected perceptual distortion is calculated according to the perceptual weight map and the bit-error probability map. Finally, to get the balance between video data compressed ration and the effect of error control, the intra refresh rate is adaptively adjusted according to the accumulation of each macroblock's expected perceptual distortion. As a result, the video quality of wireless communication is improved.In chapter 4, an adaptive joint source-channel rate allocation method for wireless video communications is proposed to get a smooth perceptual video quality in wireless video communications. According to the expected end-to-end distortion, the proposed method does the optimal rate allocation between the source coding and the channel coding。Meanwhile, the target rate allocation of video regions is also adjusted in order to get the best end-to-end video quality. Firstly, the expected end-to-end distortion is calculated based on the variance of video data and the channel state information which is fed back form the wireless channel. Secondly, the rate allocation between the source coding and the channel coding is adjusted to get the minimal expected end-to-end distortion. Finally, according to the perceptual weight map, the regions with higher perceptual weight are allocated with more target bit-rate. Then these regions can be encoded with constant ratio and can keep smoothly when the bandwidth of the wireless channel varies.In chapter 5, in order to improve the performance of error concealment in wireless video communications, a network-aware perceptual error concealment method with side information is proposed. The proposed method considers both the state information of the network and the perceptual weight of the video content, and the spatial and temporal error concealment is also performed according the video encoder type. The quality of video data over wireless channel with packet loss is improved as a result. Firstly, the reliability weight of the reconstructed macroblocks is derived by the channel state information, and the neighboring macroblocks are picked up by the reliability weight. Secondly, the perceptual weight map is extracted from the side information sent by the encoder side, and the reconstruct weight of the picked neighboring macroblocks is calculated depending on the perceptual weight map. Finally, according to the reconstruct weight and the video encoder type, pixel interpolation is utilized for spatial error concealment and edge-match based motion compensation is utilized for temporal error concealment, then the lost video data is recovered, and the video quality is improved.In the final chapter, the thesis is concluded and the prospect of the future research is also presented.