ZBDD-Based Backtrack Search SAT Solver

We introduce a new approach to Boolean satisfiability that combines backtrack search techniques and zero-suppressed binary decision diagrams (ZBDDs). This approach implicitly represents satisfiability instances using ZBDDs, and performs search using an efficient implementation of unit propagation on the ZBDD structure. We describe how to perform backtrack search using ZBDDs as the underlying structure for clause representation. This methodology, which adapts backtrack search algorithms to such implicit representations, allows for a potential exponential increase in the size of the problems that can be handled. Our experimental results show consistent speedups over conventional approaches.