An Adaptive Unequal Error Proctection Mechanism For H.264 Video Transmission In Wireless Networks

Video application is becoming an important application under the next-generation wireless networks. However, supporting video communications over wireless networks poses many challenges due to fluctuations of wireless channel conditions. Advanced video codec, like H.264, can obtain much higher compression, but only source coding cannot achieve optimal video transmission performance. The joint source-channel coding is considered to be a promising solution.In this dissertation, we proposed a new unequal error protection (UEP, Unequal Error Protection) mechanism based on joint source-channel coding, according to the data partitioning method of H.264 and the characteristics of wireless networks. This mechanism encodes the data partitioning A using RS (Reed-Solomon) coder and for the data partitioning B/C, using a repeat-packet coder. A joint source and channel RD (Rate Distortion) model was established for P frames in order to optimize the bit allocation between source coding and channel coding. Using this mechanism, the optimal parameters such as QP (Quantization Parameter), the RS coding strength and the repeat-packet coding strength, can be calculated in real time.The joint source-channel RD model was verified by testing typical standard video sequences under varied network bandwidth and loss rate. Extensive wireless network simulations were performed to evaluate the proposed UEP mechanism. The simulation results show that, compared with that without any protection, the video stream can adapt to the fluctuation of the wireless channel conditions, and the PSNR (Peak Signal-to-Noise Ratio) of the recovered video in the receiver is increased up to about 6-10 dB.