Improved variable ordering of BDDs with novel genetic algorithm
N. Zhuang, MUHAMMAD S. T. BENTEN, Peter Y. K. Cheung · 2002
A new algorithm for variable ordering of binary decision diagram (BDD) is presented. The algorithm is based on a novel formulation of the Genetic Algorithm (GA) employing three dynamic GA parameters: population size, mutation rate and stop criteria. Test results using LGSynth93 benchmark circuits show that the new algorithm offers considerable improvements on large circuits when compared with previously published results.