Home

About

Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network
FAQHELPSITE MAPCONTACT US


  Advanced Search  

 
The MINCE heuristic for variable-ordering [1] can successfully reduce the size of BDDs and accelerate SAT-solving. Applications
 

Summary: 116
ABSTRACT
The MINCE heuristic for variable-ordering [1] can successfully
reduce the size of BDDs and accelerate SAT-solving. Applications
to reachability analysis have also been successful [12]. The main
drawback of MINCE is its implementation complexity - the authors
used a pre-existing min-cut placer [6] that is several times larger
than any existing SAT solver. Tweaking MINCE is difficult.
In this work we propose a replacement heuristic, FORCE which is
easy to implement from scratch and tweak. It is dramatically faster
than MINCE in practice. While FORCE may produce seemingly
inferior variable orderings, the difference with MINCE orderings
does not affect subsequent SAT-solving.
Categories and Subject Descriptors
I.1 [Symbolic and Algebraic Manipulation]: Algorithms.
General Terms
Algorithms, Performance, Experimentation, Verification.
Keywords
BDDs, SAT, CNF, backtrack search, variable order, pre-processing,
hypergraph, partitioning, placement.

  

Source: Aloul, Fadi - Department of Computer Engineering, American University of Sharjah
Sakallah, Karem A. - Department of Electrical Engineering and Computer Science, University of Michigan

 

Collections: Computer Technologies and Information Sciences; Engineering