Multi-objective Optimization Study Of Fixed Polarity XNOR/OR Circuits

At present,with the rapid development of large-scale integrated circuits,power consumption issues are increasingly attracting people's attention.The current circuit is mainly in the form of Boolean logic circuit,and the low power optimization of the circuit is mainly carried out for the Boolean circuit.However,compared with Boolean logic,fixed polarity Reed-Muller(RM)logic not only has good testability,but also has significant advantages in power consumption,area and speed,such as parity check,arithmetic logic and communication circuit in structure and function.Under different polarities,the corresponding logical expansion is different,and the corresponding circuit is not the same.For an n-variable Boolean circuit,it can eventually correspond to 2~n fixed-polarity XNOR/OR expansions,so we need to find the optimal polarity in these 2~n polarities,so that the area,power consumption,etc.of the corresponding circuit can be achieved.optimize.Based on this,this paper will combine the intelligent algorithm to carry out multi-objective optimization research on fixed polarity XNOR/OR circuits.This article has carried out a series of work on this,and the contents of the work are summarized as follows:1.This paper systematically summarizes the basic concepts and theoretical knowledge of fixed polarity XNOR/OR circuit,understands the current research status in the field of fixed polarity XNOR/OR circuit at home and abroad,and carries out a detailed analysis of the synthesis and optimization of the circuit;at the same time,it learns some basic concepts and theoretical knowledge of multi-objective optimization algorithm to prepare for the later algorithm research.2.This paper summarizes the conversion algorithm from Boolean logic to fixed-polarity XNOR/OR circuits.There are mainly folding methods,list methods,and quick list conversion methods.For large-scale circuits,fast list algorithm is generally used because of its fast speed and high efficiency.At the same time,the fixed polarity XNOR/OR circuit is analyzed theoretically,and the area optimization model and power optimization model are constructed respectively,which provide theoretical basis for multi-objective optimization.3.By improving the particle swarm optimization algorithm,the Discrete Multi-Objective Particle Swarm Optimization(DMOPSO)algorithm and the Chaos Discrete Multi-Objective Particle Swarm Optimization(CDMOPSO)algorithm are used respectively.Based on the above-mentioned area and power optimization model,combined with the theory of multi-objective optimization algorithm,a multi-objective decision-making optimization study is constructed.Finally,the verification test is performed on the MCNC Benchmark circuit in PLA format.