BBO Algorithms with Graded Emigration for Yagi-Uda Antenna Design Optimization for Maximal Gain

Biogeography Based Optimization (BBO) is a swarm based optimization algorithm that has shown impressive performance over other Evolutionary Algorithms (EAs). Immigration Refusal Biogeography Based Optimization (IRBBO), Enhanced Biogeography Based Optimization (EBBO), Blended Migration are the most improved version of BBO and are known as migration variants of BBO. In this paper, a new concept of graded emigration for EBBO is proposed for furthur improved convergence performance. This graded emigration is also experimented on other BBO variants and found to be a competitive option. To validate the performance of Graded Emigration (GE-EBBO), experiments have been conducted on a testbed of unimodal, multimodal and deceptive benchmark test functions. Besides validation, GE-EBBO is also subjected to evolve solution to a real world problem of designing a Yagi-Uda antenna for maximal gain. Designing a Yagi-Uda antenna involves determination of wire-element lengths and their spacings in between them those bear highly complex and non-linear relationships with antenna gain, impedance, and Side Lobe Level (SLL), etc. at a particular frequency of operation. In this paper, a comparative study among BBO, EBBO, IRBBO, PSO (Particle Swarm Optimization) and GE-EBBO is conducted to analyze convergence performance while evolving the antenna designs for maximum gain, multiple times. The average of 10 monte-carlo simulations are plotted for fair quantitative comparative study of convergence performance of these stochastic algorithms.