Abstract:
Based on the dynamic logic form of circuits, a mathematic mode for low power dissipation is established for polarity optimization of fixed-polarity XNOR/OR circuits. By analyzing the standard genetic algorithm (TGA) and quantum algorithm (TQA), this paper proposes a new algorithm, population co-evolution algorithm(PCEA), which utilizes the idea of co-evolution. This proposed algorithm includes a master population and a small-scaled q-bit population as a slave one, and adopts the evolving strategies of parallel evolution, uniform evaluation, and regrouping the master population via the most excellent ones. Conventional strategies including selection, crossover and mutation are employed for the master population evolution. The evolvement of q-bit population is made by even evolution and multiple measurements so as to provide a group of individuals evenly covering the solution space for the master population, and thus to avoid the “premature” phenomena of the algorithm. Experiment results on eight MCNC Benchmark circuits verify the PCEA’s stability and efficiency.