Multiple-optimization based-design of rf integrated inductors

In this paper, a multiple-objective Metaheuristics study is discussed. Initially, three mono-objective metaheuristics will be explored in order to design and optimize Radio-Frequency integrated inductors. These metaheuristics are: An evolutionary algorithm called The Differential Evolution (DE), An algorithm supported on Newton's laws of gravity and motion called the Gravitational Search Algorithm (GSA) and, finally, A swarm intelligence algorithm called the Particle Swarm Optimization (PSO). The performances of these three mono-objective metaheuristics are evaluated and compared over three benchmark functions and one application to optimize the layout of a RF silicon-based planar spiral inductor, the double-model is adopted. Secondly, three references multi-objective metaheuristics using Pareto front are used respectively the multi-objective PSO (MOPSO), the Pareto envelope-based selection algorithm-II (PESAII) and the multi-objective evolutionary algorithm based on decomposition (MOEA/D). The performances of these multi-objective optimization algorithms are evaluated and compared over two bi-objective benchmark functions and the same application used in the first section. Two conflicting performances were optimized, namely the quality factor 'Q (to be maximized) and the device area 'dout (to be minimized) for the RF inductor. It was concluded that the multiple-objective PSO are significantly more efficient and robust for difficult problems than the other metaheuristics.