Application of swarm intelligence techniques to the design of analog circuits: evaluation and comparison

Swarm intelligence (SI) techniques are more and more used by analog designers in order to optimally size their circuits/systems’ performances. A particular interest is accorded to the multi-objective algorithms due to the fact that in most cases analog, mixed signal and radio-frequency sizing problems encompass at least two non-commensurable conflicting objectives. In most of the published papers, Pareto fronts are provided, and in the best cases they are compared to the tradeoff fronts obtained using other classical metaheuristics, but in a very subjective way. In this paper we present a comparison that deals with the multi-objective optimal design of analog circuits via the SI technique and other famous metaheuristics. Performance metrics are used to compare the obtained results. The paper argues and shows that SI techniques and particularly the particle swarm optimization technique is a priori the most adequate metaheuristic to use in the field of analog circuit sizing.