MPEG 2 video error resilience experiments: : The importance considering the impact of the systems layer

Abstract With increasing interest in the transport of video traffic over lossy networks, several techniques for improving the quality of video services in the presence of loss have been proposed, often using the MPEG 2 video coding algorithm as a basis. Many of the papers describing this work show experimental results using only the MPEG 2 video layer, and do not take into account the effect of the MPEG 2 systems layer, which will be required for carrying synchronization information and other data associated with the video, including audio. In this paper, the importance of including the impact of the systems layer on the quality of decoded video is demonstrated by comparing the degradation in quality, with and without the systems layer, of decoded MPEG 2 video being carried in a simulated ATM network with cell loss. This holds true in cases where it is desired to compare the improvement in quality gained by different techniques, and in particular where it is desired to carry out subjective or task-based tests to determine if the quality provided by particular techniques is adequate for a given application.