Title :
Rate-distortion function when side-information may be present at the decoder
Author :
Kaspi, Amiram H.
Author_Institution :
RAFAEL, Haifa, Israel
fDate :
11/1/1994 12:00:00 AM
Abstract :
A discrete memoryless source {Xk} is to be coded into a binary stream of rate R bits/symbol such that {Xk} can be recovered with minimum possible distortion. The system is to be optimized for best performance with two decoders, one of which has access to side-information about the source. For given levels of average distortion for these two decoders, the minimum achievable rate R (in the usual Shannon theory sense) is given as a per-letter minimization of information theoretic quantities. The problem is solved for two cases: where the encoder is and is not informed of the side-information
Keywords :
decoding; rate distortion theory; source coding; Shannon theory; average distortion; binary stream; decoder; decoders; discrete memoryless source; information theory; per-letter minimization; performance; rate distortion function; side-information; Convergence; Convolution; Decoding; Distortion measurement; Entropy; Lattices; Quantization; Random variables; Rate distortion theory; Rate-distortion;
Journal_Title :
Information Theory, IEEE Transactions on