Two-Archive Evolutionary Algorithm (TAEA)-Based Multi&Many Objective Analog IC Optimization

Abstract

Conventionally, metaheuristic algorithms have been employed as the core of the optimization engine, where Evolutionary Algorithms (EAs) are commonly preferred due to their high accuracy and efficiency for not only single-but also multi- and many-objective design problems. Although numerous successful derivations of the EAs have been proposed in the literature, the search for a better algorithm that fits the analog/RF IC sizing problem still pursuits. In this paper, one of the most promising EA-based approach, namely Two-Archive Evolutionary Algorithm (TAEA), is applied to the analog IC sizing problem at multi and many objective domains. To evaluate the performance of the TAEA, two analog circuits have been synthesized and the results have been compared with the results obtained by Non-Sorting Genetic Algorithm-II (NSGAII) and Multi-Objective Evolutionary Algorithm based Decomposition (MOEA/D) through the well-known Pareto-optimal front quality metrics. The study shows that TAEA performs head to head with two other common algorithms employed in analog IC synthesis. © 2024 Elsevier B.V., All rights reserved.