Comparing the Performance of Novel Swarm Intelligence Optimization Methods for Optimal Design of the Sense Amplifier-based Flip-Flops

The effectiveness of multi-objective optimization methods, especially the methods based on Swarm Intelligence, has led the researchers to utilize them significantly to solve complex engineering problems with multiple conflicting objectives. This paper aimed at evaluating the performance of new and powerful multi-objective heuristic methods based on Swarm Intelligence (including multi-objective versions of MOPSO, MOGWO, NSGSA, MOGOA, MOIPO, MOMIPO, and MOALO algorithms), and used them for optimal design of the Sense Amplifier-based Flip-Flop (SAFF) using 0.18-µm CMOS technology. In this paper, the channel's width values of the transistors as designing variables, and total average power and delay as the fitness values of the two objective functions were assessed and optimized in terms of multi-objective optimization problem using intelligent optimization algorithm based on Swarm Intelligence assumption in order to achieve the desired values of power-delay product (PDP). Comparing the results obtained for all of the above multi-objective optimization methods, the Multi-Objective Grasshopper Optimization Algorithm (MOGOA) performed better. This method was able to perform very well in the statistical indices of fitness and multi-objective optimization criteria in comparison with other methods. It creates an appropriate trade-off between conflicting objective functions with average power of 24 µW, delay of 95.4 ps and PDP of 2.29 fJ.