Novel ABC- and BBO-Based Evolutionary Algorithms and Their Illustrations to Wireless Communications

In this thesis we discuss some new Evolutionary Algorithms (EAs) as potential low-complex solvers to some optimization problems in wireless communications. Delivering high performance results while maintaining low computational complexity is extremely important in solving complex optimization problems or problems with a large search space. We propose our enhancements to Biogeography-Based Optimization (BBO) and Artificial Bee Colony (ABC) algorithms. We further present a novel high performance low-complex EA for optimization problems in both continuous and discrete domains, that combines the advantages of both BBO and ABC algorithms, which is referred to as the Hybrid ABC/BBO algorithm. This algorithm has shown higher performance in comparison to other EAs when applied to some optimization problems. We applied these algorithms to a single-objective unconstrained optimization problem (Multi Device STBC-MIMO), a single-objective constrained optimization problem (relay assignment in cognitive radio systems), and a multi-objective constrained optimization problem (Green Resource Allocation in cognitive radio systems). We provide the formulation of these problems and compared the hybrid algorithm results with exhaustive search (where applicable), and further demonstrate the superiority of the hybrid algorithm in terms of complexity and performance over ABC, BBO, other mainstream EAs and optimization solvers through simulations.