|
Abstract : |
This paper contributes to the solution of the minimization problem of Ordered Binary Decision Diagrams by means of variable reordering. We suggest a new heuristic that is based on sampling. A small OBDD sample is chosen from the OBDDs that are considered for minimization. Solving the problem for this small sample, we obtain a variable order that is adapt for the entire OBDDs. We present first experimental results with the Sample Reordering targeted at combinatorial verification. The suggested heuristic is substantially faster than Sifting. 1, |
