Replication-Guided Enumeration of Minimal Unsatisfiable Subsets

In many areas of computer science, we are given an unsatisfiable Boolean formula F in CNF, i.e. a set of clauses, with the goal to identify minimal unsatisfiable subsets (MUSes) of F. The more MUSes are identified, the better insight into F’s unsatisfiability is obtained. Unfortunately, finding even a single MUS can be very time consuming since it naturally subsumes repeatedly solving the satisfiability problem, and thus a complete MUS enumeration is often practically intractable. Therefore, contemporary MUS enumeration algorithms tend to identify as many MUSes as possible within a given time limit. In this work, we present a novel MUS enumeration algorithm. Compared to existing algorithms, our algorithm is much more frugal in the number of performed satisfiability checks. Consequently, our algorithm is often able to find substantially more MUSes than contemporary algorithms.