Optimizing the Array Factor of a Phased Array Antenna (PAA) using a genetic algorithm

Phased Array Antennas are among the most important 5G enabling technologies. Optimizing a Phased Array Antenna (PAA) is an elaborate task, especially for PAAs with increased number of elements. Traditional optimization methods that are either based on greedy, local optimization such as gradient methods or consist of random walk solution space searches, are far from ideal for solving the problem of arbitrary pattern synthesis in 1D and 2D antenna arrays, mainly due to the high dimensional, multimodal functional domains involved but also due to the inherent limitations of these approaches e.g., the majority of these methods assume the object function is continuous or even differentiable. On the other hand, natural optimization methods, e.g., genetic algorithms, simulated annealing and particle swarm optimization can be used to tackle this problem efficiently. In this study, a genetic algorithm was developed for optimizing a Planar (2D) Phased Array Antenna, using a custom objective function so that the Array Factor has specific characteristics, in terms of directivity, beamwidth and side-lobe level (SLL). The experimental results showed that the genetic algorithm can be used for optimizing a 2D PAA with very satisfactory results.