Formal Verification Techniques Based on Boolean Satisfiability Problem

This paper exploits Boolean satisfiability problem in equivalence checking and model checking respectively. A combinational equivalence checking method based on incremental satisfiability is presented. This method chooses the can didate equivalent pairs with some new techniques, and uses incremental satisfiability algorithm to improve its performance. By substituting the internal equivalent pairs and converting the equivalence relations into conjunctive normal form (CNF) formulas, this approach can avoid the false negatives, and reduce the search space of SAT procedure. Experimental results on ISCAS'85 benchmark circuits show that, the presented approach is faster and more robust than those existed in literature.This paper also presents an algorithm for extracting of unsatisfiable core, which has an important application in abstraction and refinement for model checking to alleviate the state space explosion bottleneck. The error of approximate extraction is analyzed by means of simulation. An analysis reveals that an interesting phenomenon occurs, with the increasing density of the formula, the average error of the extraction is decreasing. An exact extraction approach for MU subformula, referred to as pre-assignment algorithm, is proposed. Both theoretical analysis and experimental results show that it is more efficient.