Reducing Elements in Linear Antenna Array Using Levenberg-Marquardt Algorithm

Design of a non-uniform antenna array with reduced number of elements may be required for some applications. This paper introduces an iterative procedure, which is based on the Levenberg-Marquardt Algorithm (LMA), for reducing the number of elements in a linear array. By applying small perturbations on a given configuration, the algorithm finds the gradient that makes the total square error to be minimal, and then tests the new point in this direction until it converges to an optimal positions, amplitudes and phase, of array elements, regarding array factor requirements. Sensitivity of the resulted design to amplitude and frequency errors is examined. Simulation CST results of the theoretical design is also carried out, demonstrating the accuracy of the algorithm.