Video packetization techniques for enhancing H.264 video transmission over 3G networks

Transmission of video over a wireless network is a challenging task due to the error prone characteristics of the wireless link. This application requires the video encoding process to offer error resilience features in order to deal with the high packet loss probability of wireless networks. H.264, the newest and most efficient video compression standard, offers several error resiliency features. H.264 also offers network support by utilizing a Network Abstraction Layer (NAL) and encapsulating individually decodable video slices in NAL units. The size of the video slices, or packets, has a significant effect on the overall performance of a video application in wireless networks. In this paper we study the possibility of determining the optimal packet size of encoded video that minimizes the packet loss rate and maximizes the video quality during transmission over 3G networks. Through simulation experiments, we show that smaller slices are more favorable, but encoding inefficiency and increase in overhead set a limit on the minimum acceptable packet size.