Design and Optimization of Antenna for MIMO Systems via Heuristic Algorithms

Abstract

The radiation pattern and other electrical characteristics of an antenna primarily depend on its size and geometry. In the current age of communications, it is highly desirable to have compact and light weight communication devices. Therefore the most challenging issue of today’s wireless communication systems is to design antennas which are cost effective and of compact size without compromising the radiating efficiency and effectiveness. Since microstrip patch antennas are light weight, compact, easy to design and install; they are the best choice for today’s communication devices. That’s why patch antennas are widely used in numerous wireless applications as a single radiating element and as an array of multiple elements. However, designing of different shaped antennas to suit the electrical and/or physical requirements of the communication system is not an easy task. In this dissertation we have successfully designed microstrip patch antennas to achive desired electrical charcteristics such as frequency, bandwidth and quality factor alongwith the targeted physical characteristics such as length, radius or area etc. Keeping in mind the utilization of various shapes of antennas in communication systems; the dissertation includes designing of different geometrical shaped patch antennas such as square, circular and hexagonal. In addition to design patch antennas, we have succefully designed array of patch elements to be used in MIMO systems. The array has been designed for the desired electrical characteristics such as directivity, frequency, and return loss etc. Moreover, using this array design model, we estimated the channel capacity for ferquency selective Rayleigh fading channel. Since ultra wide band frequency systems are suitable for high data rates with low power consumption, we have particularly focused on this range of frequncy for antenna designing. Among all the heuristic approaches, literature shows that particle swarm optimization (PSO) and genetic algorithm (GA) are the most efficient and effective heuristic approaches in solving the problems of antenna and electromagnetics. Therefore we have employed these two approaches for designing and optimizing these antennas. We have also developed a new heuristic algorithm namely Wildebeest Herd Optimization (WHO) algorithm. It is based on the living wildebeest in their herd and migrating from one point to another for food and water. All the multi-geometrical antennas have been designed and optimized on WHO in addition to GA and PSO. The comparative analysis verifies that the proposed algorithm is more effective and efficient than the existing heuristic approaches. Moreover, artificial neural network (ANN) is employed to model the electrical/magnetic behavior of single antenna and arrays. The outcome of ANN models are compared with the results of numerical computational software Ansoft HFSS. In the light of this work, various future recommendations are also given at the end.