Fast factorization method for implicit cube set representation

This paper presents a fast weak-division method for implicit cube set representation using Zero-Suppressed Binary Decision Diagrams, which are a new type of Binary Decision Diagram adapted for representing sets of combinations. Our new weak-division algorithm can be executed in a time almost proportional to the size of the graph, regardless of the number of cubes and literals. Based on this technique, we implemented a simple program for optimizing multilevel logic circuits. Experimental results indicate that we can quickly flatten and factorize multilevel logics even for parity functions and full adders, which have never been flattened in other methods. Our method greatly accelerates multilevel logic synthesis systems and enlarges the scale of applicable circuits.