Approximate logic optimization for ISFPRM function

Abstract

There were many studies on the optimization of traditional Boolean (TB) logic functions. In contrast, research on the Reed-Muller (RM) logic is scarce. This paper proposes a logic optimization method for FPRM (Fixed Polarity RM, FPRM), which includes: design evaluation model of approximate circuit, and approximation logic based on genetic algorithm for incompletely specified FPRM. The performance of the algorithm uses the MCNC benchmark circuit as a test. The experimental results show that every 1% of errors can be obtained from 11.11% to 12.18% of area optimization.