Optimization of large log-periodic dual-dipole antenna by using Genetic Algorithm on embedded element in small log-periodic array

The paper introduces a new method that can predict the reflection coefficient of a complete large log-periodic array (with large number of elements) by calculating the embedded S-parameters between all input and output ports of similar elements that are embedded in a much smaller log-periodic array. By this method we can apply global optimization scheme such as Genetic Algorithm on embedded element in the smaller log-periodic array to improve the reflection coefficient of the larger log-periodic array very efficiently. As an example, an Eleven feed for 2 - 13 GHz has been optimized by using Genetic Algorithm with the present method. The result is a reflection coefficient below -10 dB from 2 GHz to 13 GHz.