A Power Effective Algorithm for State Encoding

Abstract

Reducing the area and power dissipation of FSM circuit is of significant importance for EDA technology. Many methods are adopted to achieve an effective and fast transformation of FSMs to binary codes, including Genetic algorithm (GA) and others. In this paper, we propose a GA based state assignment of a FSM circuit to gain the minimization of power consumption and area. We modify the traditional mutation to be an ordered operation, which is also a substitution of the crossover that guarantees every new individual owns better fitness than the old one. We test the proposed algorithm with benchmarks, as well as do the comparison with the published; our method saves both power and area dissipation in reasonable computation time.

DOI: http://dx.doi.org/10.11591/telkomnika.v12i6.5482