An Adaptive Error Resilient Scheme for Packet-Switched H.264 Video Transmission

When applying conventional standard video codecs in wireless video applications, error resilience and coding efficiency are the two main issues need to be considered. Since it is difficult to corroborate robust quality of service (QoS) in wireless networks, transmitted video packets are sometime lost or corrupted due to fading and shadowing effect of wireless channel. Providing robust video transmission in wireless packet-switched networks is therefore a challenging task as compressed video is very vulnerable to channel error. In recent years, many error resilient tools have been proposed to enhance the robust performance of video transmission in wireless environment (Chen et al., 2008)(Stockhammer et al., 2003) (Stockhammer et al., 2005) (Vetra et al., 2005) (Wang et al., 2000) (Wiegand et al. 2003). Coding efficiency is one of the important issues to be taken into account for the limited bandwidth of wireless networks (Etoh and Yoshimura, 2005). To achieve a robust video transmission over wireless channels, the video codec on one hand should have supreme error resilient performance, on the other hand, it should also maintain a good coding efficiency by limiting the overhead information introduced by the error resilient tools. Hence, a good compromise between the error resilience performance and coding efficiency should be made. Interactive error control is one of the effective error resilient techniques adopted in video codec. In this category, some error resilient techniques based on data hiding are proposed (Zeng, 2003) (Yilmaz and Alatan, 2003) (Kang and Leou, 2005). In such techniques, the important information for error concealment is extracted and embedded into video bitstream at the video encoder. When some packets are lost or corrupted, their corresponding embedded data at proper positions can enhance the error concealment effect at the video decoder. Although data hiding methods can obtain satisfied error resilient effect, their notable increase of bits overhead is disadvantageous for coding efficiency. Since the principle of embedding important information (Yin el al., 2001) they adopted modifies the original AC coefficients, not only video quality is degraded, but also coding overhead will be increased significantly. In wireless channels, as the transmission rate is limited, an obvious increase on coding overhead results in inevitable delay. Moreover, in wireless