Several evolutionary algorithms (EAs) have emerged in recent decades that mimic the behaviour and evolution of biological entities. EAs are widely used to solve single and multi-objective optimization engineering problems. EAs have also been applied to a variety of microwave components, antenna design, radar design, and wireless communications problems. These techniques, among others, include genetic algorithms (GAs), evolution strategies (ES), particle swarm optimization (PSO), differential evolution (DE), and ant colony optimization (ACO). In addition, new innovative algorithms that are not only biology-based but also physics-based or music-based are also emerging, as are hybrid combinations of EAs. The use of evolutionary algorithms is having an increasing impact on antenna design and wireless communications problems. EAs combined with numerical methods in electromagnetics have obtained significant and successful results. This book aims to present some of the emerging Eas and their variants. Chapter 1 introduces the optimization methods in general and the evolutionary algorithms. Chapter 2 presents briefly some of the most popular evolutionary algorithms, such as particle swarm optimization (PSO), differential evolution (DE), and ant colony optimization (ACO) as well as some emerging ones. Chapter 3 focuses on antenna array synthesis, which constitutes a wide range of antenna design problems. Chapter 4 gives an overview of patch antenna design using evolutionary algorithms. Chapter 5 presents design cases from different microwave structure cases. Chapter 6 discusses on various representative design problems in wireless communications. Chapter 7 deals with design cases for 5G and beyond.
Inspec keywords: evolutionary computation; microstrip antenna arrays; microwave antenna arrays; 5G mobile communication
Other keywords: microwave structures design; wireless communications; 5G design problems; evolutionary algorithms; antenna array design; microstrip patch antenna design
Subjects: Antenna arrays; Optimisation techniques; Mobile radio systems
This chapter briefly introduces basic optimization concepts. Optimization problems exist in our everyday life. We often wonder about the route with less traffic to work, the less money to spend in various activities, or the best work schedule. Optimization can be thought as the art of making good decisions. In more precise manner, we may define optimization as the process of changing a system's state (or values of some unknown variables) toward a state of a minimum or maximum property. Several optimization problems exist in the whole engineering domain. A common problem in antennas design is to find the best geometry for a given antenna, while in wireless communications the association between a base station and user in an optimal manner is a very popular optimization problem.
Swarm intelligence (Si) algorithms are special category of EAs. The essential concept of SI is the collective behavior of decentralized and self-organized swarms. There are several popular SI algorithm families that among others include particle swarm optimization (PSO), ant colony optimization (ACO), and artificial bee colony (ABC). The popularity of SI algorithms is due to the fact that they in general can handle efficiently arbitrary optimization problems. Additionally, SI algorithms as it can be found from the literature have been widely utilized to solve several problems in antennas and in wireless communications. The swarm behavior of bird flocking and fish schooling is modeled mathemati-cally by the PSO algorithm. One may find several PSO variants in the literature. The most frequently applied PSO variants include the inertia weight PSO (IWPSO) and the constriction factor PSO (CFPSO). Moreover, comprehensive learning particle swarm optimizer (CLPSO) is a PSO algorithm that has been applied to antenna design problems. The PSO algorithm is intrinsically suitable for applica-tion to real-valued problems. Thus, binary PSO (BPSO) versions should be used for solving discrete-valued problems. BPSO is one of the most popular discrete PSO algorithms. BPSO maps real values to the discrete set by using a sigmoid trans-fer function. Additionally, several new transfer functions that perform better than the original algorithm have been introduced by the authors. Furthermore, Boolean PSO is another BPSO version with main characteristic the usage of binary oper-ators for velocity and position update. Several authors have applied Boolean PSO to antenna design problems.
This chapter presents several antenna array design cases by using different evolution-ary algorithms (EAs) and comparing results. The synthesis of antenna arrays plays a very important role in communication systems. Array synthesis is a classic and challenging optimization problem, which has been extensively studied using several analytical or stochastic methods. The increased use of such arrays creates more challenges upon the antenna engineers. More requirements, such as pattern shaping, low profile, wideband/narrowband, and interference cancellation; and more limitations such as power dissipation and antenna size, lead to the urgent need for simple, time saving, and efficient optimization tech-niques. Common optimization goals in array synthesis are the sidelobe level (SLL) suppression and the matching of the mainlobe to a desired pattern. Thus, the opti-mization problem is usually to find a set of element excitations and/or positions that closely match a desired pattern. The desired pattern shape can vary widely depending on the application. Several new synthesis and optimization techniques have emerged in the last two decades that mimic biological evolution, brain function, or the way biological entities communicate in nature. Several of these methods have been applied to the array design problem
This chapter describes patch design cases for different applications. The use of microstrip patch antennas in wireless communication systems provides several advantages like low profile, low cost, and ease of fabrication. Moreover, microstrip patch antennas can provide a possible solution for fifth generation (5G) antenna design. Different antenna shapes can be fabricated using the rectangular patch as an initial step. This type of antenna design requires the simultaneous optimization of several different geometrical parameters. An optimization algorithm or techniques is a suitable approach for solving this problem. In the literature, there are several examples of patch antenna design and optimization using different evolutionary algorithms (BAs) [1-5]. These approaches include genetic algorithms [6], particle swarm optimization (PSO) [7-9], differential evolution (DB) [10-13], teaching-learning-based optimization (TLBO)[14], Jaya [15], and a hybrid Jaya-GWO algorithm [16].
This chapter presents design cases from different microwave structure cases. The problem of a planar microwave absorber design lies in the minimization of the reflection coefficient of an incident plane wave in a multilayer structure for a desired range of angles and frequencies. The reflection coefficient depends on the thickness and the electric and magnetic properties of each layer. Several studies, which address this problem, exist in the literature. Evolutionary algorithms (EAs) like genetic algorithms (GAs) have been in several occasions applied in absorber design. In, a microGA algorithm with a predefined materials database was applied. The major drawback of a GA approach is the difficulty in implementation due to the algorithm-inherited complexity and the required long computational time. Moreover, the application of swarm intelligence optimizers in electromagnetic (EM) design problems has attracted several researchers. Particle swarm optimization (PSO) has also been used successfully in absorber design problems. Recently, artificial bee colony (ABC)has been applied in several cases to microwave absorber design. Differential evolution (DE) has also been applied to absorber design problem. A comparative study between PSO and DE for the microwave absorber design problems is reported. Moreover, other emerging nature-inspired algorithms have been deployed in the literature for solving the abovementioned problem.
This chapter presents representative examples of design problems in wireless communications using evolutionary algorithms (EAs). The following design problems are dealt with: peak-to-average power ratio reduction in OFDM systems, antenna selection in MIMO systems, cognitive radio engine design, spectrum allocation in cognitive radio networks and the optimization of wireless sensor networks.
This chapter presents design cases for 5G and beyond using EAs. The first design case is multi-objective optimization in 5G massive MIMO wireless networks. The current fifth generation offers extremely wide spectrum and multi-gigabit-per-second data rates for mobile users. The massive multiple input-multiple output (MIMO) concept is one of critical technologies of 5G networks, where each base station (BS) is equipped with a large number of antennas and can provide service to multiple users, over the same time and frequency band. The network designer should take into account different requirements and objectives for the proper operation of 5G wireless networks with massive MIMO. The second design case is joint power allocation and user association in non-orthogonal multiple access networks. Non-orthogonal multiple access (NOMA) techniques are going to have a critical role in 5G mobile networks. If the network resources are assigned to users with poor channel conditions then the current orthogonal multiple access (OMA) networks suffer from low spectrum efficiency. However, the above condition is entirely different in the case of NOMA systems in the power domain. The NOMA scheme allows the users to use the same frequency, time, and code simultaneously, while it allocates different levels of power.