Title :
Symmetry Propagation: Improved Dynamic Symmetry Breaking in SAT
Author :
Devriendt, J. ; Bogaerts, B. ; De Cat, B. ; Denecker, M. ; Mears, C.
Author_Institution :
Dept. of Comput. Sci., KU Leuven, Leuven, Belgium
Abstract :
For constraint programming, many well performing dynamic symmetry breaking techniques have been devised. For propositional satisfiability solving, dynamic symmetry breaking is still either slower or less general than static symmetry breaking. This paper presents Symmetry Propagation, which is an improvement to Lightweight Dynamic Symmetry Breaking, a dynamic symmetry breaking approach from CP. Symmetry Propagation uses any given symmetry as a propagator, and as a result is a general symmetry breaking technique. Experiments with an implementation in the SAT solver Minisat show that on many benchmarks, Symmetry Propagation outperforms the state-of-the-art static symmetry breaking method Shatter.
Keywords :
computability; constraint handling; Minisat; SAT solver; constraint programming; lightweight dynamic symmetry breaking; propositional satisfiability solving; static symmetry breaking; symmetry popagation; Benchmark testing; Dynamic programming; Generators; Optimization; Orbits; Radiation detectors; Search problems; CP; Dynamic Symmetry Breaking; SAT;
Conference_Titel :
Tools with Artificial Intelligence (ICTAI), 2012 IEEE 24th International Conference on
Conference_Location :
Athens
Print_ISBN :
978-1-4799-0227-9
DOI :
10.1109/ICTAI.2012.16