Combined source-channel coding schemes for video transmission overan additive white Gaussian noise channel

摘要

There has been an increased interest in the transmission ofndigital video over real-world transmission media, such as the directnbroadcast satellite (DBS) channel. Video transmitted over such a channelnis subject to degradation due, in part, to additive white Gaussian noisen(AWGN). Some form of forward error-control (FEC) coding may be appliednin order to reduce the effect of the noise on the transmitted bitstream;nhowever, determination of the appropriate level of FEC coding isngenerally an unwieldy and computationally intensive problem, as it mayndepend upon a variety of parameters such as the type of video, thenavailable bandwidth, and the channel SNR. More specifically, a combinednsource-channel coding approach is necessary in optimally allocating ratenbetween source and channel coding subject to a fixed constraint onnoverall transmission bandwidth. In this paper we develop a method ofnoptimal bit allocation under the assumption that the distortion isnadditive and independent on a frame-by-frame basis. A set of universalnoperational distortion-rate characteristics is developed which balancesnthe tradeoff between source coding accuracy and channel error protectionnfor a fixed overall transmission rate and provides the basis for thenoptimal bit allocation approach. The results for specific source andnchannel coding schemes show marked improvement over suboptimum choicesnof channel error protection. In addition, we show that our resultsnapproach information-theoretic performance bounds which are developed innthis work