Non-uniform 3D antenna array optimal configuration via extremum seeking control

Date of Award


Degree Name

Ph.D. in Electrical and Computer Engineering


Department of Electrical and Computer Engineering


Advisor: Raúl Ordóñez


In this work, designing a system that controls a non-uniform antenna array will be attempted. This system will reconfigure the antenna array to find the position and orientation that produce the maximum objective (directivity). To create such a system, we will assume that we have N vehicles carrying antennas. These vehicles will be in 1D, 2D, and 3D spaces (e.g., quad-copters). Each vehicle with its antenna has five degrees of freedom. The five degrees are as follows, three degrees for the position of the vehicle (x, y, z), and two for the orientation of the antenna (θ, φ). The system should be able to test every possible value for these five parameters to dynamically choose the position and orientation that maximize the directivity. By measuring the performance function value online, a sequence, generated by three Extremum Seeking Control (ESC) algorithms that are compared in this work, guides the regulator that drives the state of system to a set point that optimizes the performance function. This approach develops a novel method to achieve reconfigurable antenna arrays, by defining an objective function to be optimized on-line using three different ESC methods: (a) Perturbation-based Extremum Seeking Control (PESC), (b) Numerical Optimization-based Extremum Seeking Control Using Direct Search (NOESC-DS), and (c) Numerical Optimization-based Extremum Seeking Control Using Particle Swarm Optimization (NOESC-PSO). Simulation results are provided for confirming the validity and the effectiveness of the proposed methods. We also studied the sensitivity of PESC and NOESC-PSO to changes in the initial conditions of the antenna array. We have found ESC to be a viable and effective method to achieve reconfigurable antenna arrays, capable of satisfying on the fly a given performance objective, even when the array elements can be moved in 3D space and the orientations of the antenna elements are allowed to vary. This is a challenging scenario, and our results indicate that it could be feasibly solved in practice using the proposed approach.


Antenna radiation patterns Simulation methods, Antenna arrays, Electrical Engineering

Rights Statement

Copyright 2016, author