Dealing with 4-variables by resolution: An improved MaxSAT algorithm

We study techniques for solving the Maximum Satisfiability problem (MaxSAT). Our focus is on variables of degree 4. We identify cases for degree-4 variables and show how the resolution principle and the kernelization techniques can be nicely integrated to achieve more efficient algorithms for the Ma...

Full description

Saved in:
Bibliographic Details
Published in:Theoretical computer science Vol. 670; pp. 33 - 44
Main Authors: Chen, Jianer, Xu, Chao, Wang, Jianxin
Format: Journal Article
Language:English
Published: Elsevier B.V 29.03.2017
Subjects:
Branch and bound
Maximum satisfiability
Parameterized algorithm
The resolution principle
Maximum satisfiability
Parameterized algorithm
Branch and bound
The resolution principle
ISSN:0304-3975, 1879-2294
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We study techniques for solving the Maximum Satisfiability problem (MaxSAT). Our focus is on variables of degree 4. We identify cases for degree-4 variables and show how the resolution principle and the kernelization techniques can be nicely integrated to achieve more efficient algorithms for the MaxSAT problem. As a result, we present an algorithm of time O⁎(1.3248k) for the MaxSAT problem, improving the previous best upper bound O⁎(1.358k) by Ivan Bliznets and Alexander Golovnev. •A faster algorithm is developed for the important MaxSAT problem.•Two new resolution-based reduction rules are introduced that produce instance structures for efficient branching processes.•The resolution principle and kernelization technique are nicely integrated in handling MaxSAT instances.•The algorithm achieves the most significant improvement over the previous best upper bound, compared to the works since 1999.
ISSN:0304-3975
1879-2294
DOI:10.1016/j.tcs.2017.01.020