DocumentCode :
1204650
Title :
Sinkhorn Solves Sudoku
Author :
Moon, Todd K. ; Gunther, Jacob H. ; Kupin, Joseph J.
Author_Institution :
Electr. & Comput. Eng. Dept., Utah State Univ., Logan, UT
Volume :
55
Issue :
4
fYear :
2009
fDate :
4/1/2009 12:00:00 AM
Firstpage :
1741
Lastpage :
1746
Abstract :
The Sudoku puzzle is a discrete constraint satisfaction problem, as is the error correction decoding problem. We propose here an algorithm for solution to the Sinkhorn puzzle based on Sinkhorn balancing. Sinkhorn balancing is an algorithm for projecting a matrix onto the space of doubly stochastic matrices. The Sinkhorn balancing solver is capable of solving all but the most difficult puzzles. A proof of convergence is presented, with some information theoretic connections. A random generalization of the Sudoku puzzle is presented, for which the Sinkhorn-based solver is also very effective.
Keywords :
decoding; error correction codes; parity check codes; Sinkhorn solves Sudoku; Sudoku puzzle; belief propagation; discrete constraint satisfaction problem; error correction decoding problem; low-density parity-check decoding; stochastic matrices; Belief propagation; Constraint theory; Decoding; Error correction; Error correction codes; Helium; Jacobian matrices; Moon; Parity check codes; Stochastic processes; Belief propagation (BP); Sinkhorn; Sudoku; constraint satisfaction; low-density parity-check (LDPC) decoding;
fLanguage :
English
Journal_Title :
Information Theory, IEEE Transactions on
Publisher :
ieee
ISSN :
0018-9448
Type :
jour
DOI :
10.1109/TIT.2009.2013004
Filename :
4804943
Link To Document :
بازگشت