Accelerating External Search with Bitstate Hashing

In this paper we refine external exploration for explicit state model checking by a fusion with internal bitstate hashing. External A* provides a method to cope up with large state spaces by efficiently utilizing secondary storage devices like harddisk to maintain the open and closed lists. Duplicates are removed by a two-level refinement scheme that involves sorting a subset of the open list externally and subtracting a small subset of closed list from the open list. The bottleneck in External A* [7] is the duplicates removal phase that dominates the I/O complexity of External A*. Bitstate hashing provides a solution to faster duplicates removal by utilizing only few bits for each state. But bitstate hashing is faced with the problem of having no support for large open list and for solution reconstruction. We present a strategy to accelerate external search by using bitstate hashing for duplicates removal. Case studies with our experimental external explicit state model checker IO-HSF-SPIN illustrate the effectiveness of the proposed algorithms.