Title :
On the solution of source matching problems
Author_Institution :
US Gen. Services Adm., Hood Coll., USA
Abstract :
This paper suggests a simple method for solving source matching problems wherein the problem is to find the minimax codes which minimize the maximum redundancies over classes of sources where relative entropy (cross entropy; discrimination information) is adopted as a criterion to measure the redundancy. An efficient algorithm is presented which relies upon and extends a vector quantization result first presented at IEEE PacRim ´97 at Victoria BC Canada. A computational complexity analysis and a numerical study demonstrate the efficiency of the suggested method
Keywords :
channel capacity; computational complexity; convergence of numerical methods; entropy; memoryless systems; minimax techniques; redundancy; source coding; vector quantisation; channel capacity; computational complexity; cross entropy; discrete memoryless source; discrimination information; efficient algorithm; minimax codes; redundancy; relative entropy; source matching problems; vector quantization; Channel capacity; Closed-form solution; Computational complexity; Educational institutions; Entropy; Equations; Minimax techniques; Vector quantization;
Conference_Titel :
Communications, Computers and Signal Processing, 1999 IEEE Pacific Rim Conference on
Conference_Location :
Victoria, BC
Print_ISBN :
0-7803-5582-2
DOI :
10.1109/PACRIM.1999.799544
