Modeling and optimization of microwave devices and circuits using genetic algorithms

This paper presents a new approach for the simulation and optimization of microwave devices using a genetic algorithm (GA). The proposed technique solves the equations that describe the semiconductor transport physics in conjunction with Poissonʹs equation, employing an adaptive real-coded GA. An objective function is formulated, and most of the GA parameters are recommended to change during the simulation. In addition, different methods for describing the way the GA parameters change are developed and studied. The effect of GA parameters including the mutation value, number of crossover points, selection criteria, size of population, and probability of mutation is analyzed. The technique is validated by simulating a submicrometer field-effect transistor, and then compared to successive over relaxation, showing the same degree of accuracy along with a moderate speed of convergence. The purpose of this paper is to introduce a new vision for a GA capable of optimizing real value functions with a considerably large number of variables. This paper also represents a fundamental step toward applying GAs to Maxwellʹs equations in conjunction with the hydrodynamic model, aiming to develop an optimized and unconditionally stable global-modeling simulator.