Small signal parameter extraction of MESFET using quantum particle swarm optimization

Abstract

This paper discusses a novel technique to extract small signal equivalent circuit model parameters of GaAs MESFET device based on particle swarm optimization (PSO) technique. Three different variants of PSO namely basic PSO, Delta well quantum PSO (DQPSO) and Harmonic well quantum PSO (HQPSO) are implemented and compared. We find that these techniques extract the 16-element small signal model parameters of MESFET accurately. The simulations show that these algorithms are robust and are able to extract physically meaningful values for all circuit elements. The efficiency of this approach is demonstrated by the results that provide a good fit between measured and modeled S-parameter data over a frequency range of 0.5-25 GHz. Comparative results indicate that both DQPSO and HQPSO give good quality of solutions. We also find that basic PSO algorithm is better than DQPSO and HQPSO for all the performance evaluation parameters, i.e. mean, standard deviation, amplitude and phase relative error and computational time. © 2009 Elsevier Ltd. All rights reserved.