Abstract
This study introduces a novel multi-objective optimization algorithm integrating Customized Mutated Particle Swarm Optimization (CM-PSO) and an innovative modified Genetic Algorithm (GA) using an unexplored merged chaotic map. The hybrid algorithm converges to desired results faster than CM-PSO and modified GA without trapping in local minima. Validation is conducted by designing a single-element and simple-structure dipole antenna so that its optimized $S_{11}$ is better than −30 dB at the resonance frequency and covers the 3.3 to 3.8 GHz frequency band with $S_{11} < -10$ dB. Certainly, the −30 dB and covering frequency band criteria can be modified in the proposed algorithm. In the algorithm, the isolation between elements of a quad-Multiple-Input/Multiple-Output antenna, constructed using optimized dipole antennas, is set to be less than −20 dB (changeable criteria) so that the smallest size can be achieved. Computer Simulation Technology (CST) Studio Suite carries out electromagnetic and high-frequency simulations, and the novel developed optimization algorithm in MATLAB determines what and how much parameter values need to be changed by CM-PSO or an innovative modified GA in order to enhance the antenna’s $S_{11}$ result and its Impedance Bandwidth (IBW). The input parameters of the algorithm are the dimensions of the proposed antenna’s elements, which significantly influence its performance.

This publication has 23 references indexed in Scilit: