A global genetic algorithm based optimization technique for spiral inductor on silicon design

Spiral inductor on silicon, as the essential component of radio frequency integrated circuits, has attracted enormous academic and commercial interest. However, its quality factor is severely degraded by skin effect of metal traces and substrate coupling effect in radio frequency band, and a key problem when designing SIOS is to find optimal parameters to enhance the quality factor subject to some constraints. The optimization problem is formulated, which belongs to the class of nonlinear combinational optimization. Though some optimization algorithms have been proposed, they cannot perform global search to find more general parameters. This paper proposes a global optimization technique based on genetic algorithm, where all SIOS parameters are represented as chromosomes on which genetic operators are repeatedly applied to find better solutions. Experimental results show that genetic algorithm can effectively find out the optimal parameters to enhance the quality factor.