PARTICLE SWARM OPTIMIZATION OF ANTENNA ARRAYS WITH EFFICIENCY CONSTRAINTS

Abstract

Phased array antennas are a viable solution to a number of problems related to radio communications applications. In this work, the multi-objective stochastic MOPSO algorithm is used to optimize the spatial conflguration of a symmetric phased linear array. The deflned optimization goals were the suppression of the radiation pattern sidelobes at a specifled maximum scan angle as well as the minimization of the induced voltages correlation at the receiver front- end in order to maximize diversity performance. Non-linear constraints were enforced on the solution set, related to the multi-antenna system aperture e-ciency and related to the mismatching when the array is scanned. The obtained optimized conflgurations for an array composed of 16 dipoles resulted in reducing the sidelobes up to 2.5dB, when scanned 60 - away from broadside, compared to a linear array with elements spaced ‚=2 apart. Furthermore, the optimized dipole arrays were characterized by a maximum element correlation of 0.12 to 0.43. The performance of obtained conflgurations was shown to be tolerant to feed phase variations that appear in realistic implementations. The arrays were analyzed employing the Method of Moments (MoM).