RCM-MVI: a recursive consensus MVL minimization algorithm
Gerhard W. Dueck, D. Michael Miller · 2002
A heuristic minimization algorithm based on the recursive application of the familiar consensus operation is presented. Unlike traditional consensus-based algorithms, it allows for the early detection of essential and pseudoessential prime implicants. This means that a solution is usually determined without all prime implicants being generated. The only heuristic in the algorithm deals with the breaking of prime implicant cycles. A very simple method presented here is surprisingly effective. The algorithm is applicable to the minimization of binary functions and can be applied to multiple-output problems as well. A primary motivation for this work has been to develop a minimization algorithm much simpler than existing algorithms so that it can serve as a basis for the future work on minimization with respect to constraints and to testability.>