On the optimal design of low sidelobe level linear antenna arrays using a class of evolutionary algorithms.

Antenna array synthesis problems are known to be nonlinear and non-convex optimization problems which require more robust optimization techniques than gradient-based techniques. This paper provides a comprehensive study of a class of evolutionary algorithms that consists of ten algorithms on various linear antenna array design problems. These linear antenna array design problems are concerned mostly with the design of low sidelobe level antenna arrays which is one of the most important design metrics for any system that integrates an antenna/array of antennas in its structure. In the past, many global optimization techniques as well as their variants were used in the antenna array design to overcome the weaknesses in gradient-based techniques. Most of the work introduced in the literature lacks the consistency while comparing the performance of more than one optimization technique over a certain set of optimization problems. This paper provides a fair comparison and re-assessment of the performance of a set of evolutionary algorithms that were applied in the past to solve various antenna array design problems. The performance of the contestant algorithms will be assessed, and a statistical analysis will be performed to compare these algorithms and test their robustness. [ABSTRACT FROM AUTHOR]