Title :
Nonlinear Sparse-Graph Codes for Lossy Compression of Discrete Nonredundant Sources
Author :
Gupta, Ankit ; Verdu, Sergio
Author_Institution :
Princeton Univ., Princeton
Abstract :
We propose a scheme to implement lossy data compression for discrete equiprobable sources using block codes based on sparse matrices. We prove asymptotic optimality of the codes for a Hamming distortion criterion. We also present a sub-optimal decoding algorithm, which has near optimal performance for moderate blocklengths.
Keywords :
Hamming codes; block codes; data compression; decoding; graph theory; nonlinear codes; sparse matrices; Hamming distortion criterion; block codes; discrete equiprobable sources; discrete nonredundant sources; lossy data compression; nonlinear sparse-graph codes; sparse matrices; sub-optimal decoding algorithm; Channel coding; Error analysis; Error correction codes; Error probability; Lakes; Maximum likelihood decoding; Rate-distortion; Sampling methods; Source coding; Vectors;
Conference_Titel :
Information Theory Workshop, 2007. ITW '07. IEEE
Conference_Location :
Tahoe City, CA
Print_ISBN :
1-4244-1564-0
Electronic_ISBN :
1-4244-1564-0
DOI :
10.1109/ITW.2007.4313132
