A genetic algorithm for decomposition type choice in OKFDDs

Abstract

A genetic algorithm (GA) is applied to find decomposition type lists (DTLs) that minimize the size of ordered Kronecker functional decision diagrams (OKFDDs). OKFDDs are a data structure for representation and manipulation of Boolean functions. The choice of the DTL largely influences the size of the OKFDD, i.e. its size may vary from polynomial to exponential. In Dreschsler, Becker, and Jahnke (1995) heuristic methods have been presented. In this paper the authors show by experiments that better results can be obtained by using GAs.<>