A genetic algorithm for the construction of small and highly testable OKFDD-circuits
Rolf Drechsler, Bernd Becker, Nicole Göckel · 1996
A Genetic Algorithm (GA) is applied to derive circuits that combine area efficiency and testability. These circuits are obtained from Ordered Kronecker Functional Decision Diagrams (OKFDDs). In "Becker and Drechsler (1995)" a heuristic approach has been presented. In this paper we show how these results can further be improved by GAs. Finally, we apply our minimization algorithm to technology mapping for FPGAs. We present experimental results to show the efficiency of our approach. 1 Introduction Genetic Algorithms (GAs) are often used in optimization and machine learning (see "Goldberg (1989)" and "Davis (1991)"). In many applications they are superior to the classical optimization techniques, e.g. gradientdescent. Recently, GAs have succesfully been applied to several hard problems in Computer Aided Design (CAD) (see e.g. "Rahmani and Ono (1993)", "Esbensen and Mazumder (1994)" and "Drechsler, Becker, and Gockel (1995)"). In the following we briefly describe some of the problems oc...