Reversible circuit synthesis with particle swarm optimization using crossover operator

Abstract

Reversible circuit synthesis is an important branch of low power consumption circuit design. The idea of a logic circuit with no losses of information during computation has impact on power consumption and on the other hand makes the use of classical circuit synthesis algorithms impossible. In the area of reversible circuit design there is still lack of good algorithms. During last 15 years many heuristic algorithms have been developed, however, they construct circuit implementations which are far from optimal. In the paper the metaheuristic evolutionary algorithm is used. The algorithm combines two of known metaheuristic approaches: particle swarm optimization (PSO) and genetic algorithms (GA). In the frame of PSO the new crossover genetic operator is used. The proposed Best Circuit Cost Crossover (BCCC) operator is designed especially for reversible circuit synthesis. In the result the hybrid PSO-GA algorithm is used for reversible circuit synthesis. Many of the existing approaches use Toffoli gates with positive controls only, while the presented algorithm operates on Toffoli gates with both positive and negative controls. The presented, approach applied to known benchmark functions, gives satisfactory results when compared with the approaches known in the literature. The resulting circuits are less redundant than those obtained via heuristic algorithms and for some of benchmarks are optimal or near optimal.