Computational Complexity of Constraint Satisfaction

Abstract. The input to a constraint satisfaction problem (CSP) consists of a set of variables, each with a domain, and constraints between these variables formulated by relations over the appropriate domains; the question is if there is an assignment of values to the variables that satisfies all constraints. Different algorithmic tasks for CSPs (checking satisfiability, counting the number of solutions, enumerating all solutions) can be used to model many problems in areas such as computational logic, artificial intelligence, circuit design, etc. We will survey results on the complexity of these computational tasks as a function of properties of the allowed constraint relations. Particular attention is paid to the special case of Boolean constraint relations. Key words: satisfiability problems, constraint satisfaction, computational complexity, polymorphism, clone, Galois connection 1 Satisfiability Problems The propositional satisfiability problem SAT, i.e., the problem to deci...

• Type: Conference
• Year: 2007
• Where: CIE
• Authors: Heribert Vollmer
• Keywords: CIE 2007, Constraint Relation, Constraint Satisfaction, Satisfiability Problem