Radiation Pattern Optimization of Aperiodic Arrays Consisting of Sub-arrays

This paper proposes the use of a steepest descent algorithm (SDA) for optimizing the side-lobe characteristics of an aperiodic array consisting of 16 sub-arrays. While SDA is very attractive because it can give an optimum solution with relatively short convergence time, the solutions obtained are local optima and they depend much on initial conditions. To find the best solution that satisfies all the requirements including off-axis patterns, a genetic algorithm combined with SDA is proposed. The paper presents a unique method that gives initial positions by using binary codes that represent sub-array positions. The initial positions thus obtained have been used to find an optimum solution by SDA. An example of calculated radiation patterns is shown