Tuning of parameters of a soft computing system for the synthesis of reversible circuits

Reversible circuit design is a promising area of study in the context of expected technological advances towards quantum computing. RIMEP (Reversible Improved Multi-Expression Programming) is a Linear Genetic Programming tool devoted to such a design. In spite of the success of RIMEP to deliver competitive solutions for reversible circuit design problems, finding an appropriate parameter setting has remained a persisting challenge as in most evolutionary approaches to problem solving. The parameter setting studied in this paper includes the probability of crossover and the probability of mutation, chromosome length, population size and maximum number of generations. Moreover, following the philosophy of soft computing, that is, searching for synergy among different methods to jointly solve a problem, the Design of Experiments method (DOE) is introduced to find the best combination of these parameters values. The analysis of variance (ANOVA) is used to determine the main and interaction effects of the considered parameters. Results verified the efficacy of the proposed systematic tuning approach for RIMEP to solve reversible design problems, including one in the context of Multiple-valued Logic.