Channel Aware Joint Optimization of Latency and Distortion for Image Communications

In some online transcoding applications, such as web browsing, web contents are transcoded into the same or a different format to reduce bandwidth consumption and speed up web access for narrow bands, such as wireless channels. In addition to the rate-distortion performance of the transcoder, transcoder complexity plays a crucial role in real time transcoding applications, since it directly affects the total access latency. This paper proposes and formulates the joint optimization problem of latency and distortion, which is related to both the rate-distortion performance and the complexity of the transcoder. A practical solution is presented to choose the right transcoder based on the available bandwidth at the time of transcoding, so as to minimize the total latency for a given distortion constraint. This is especially useful for wireless communications where the available bandwidth keeps changing due to different contributing factors.