Genetic algorithm assisted adaptive beamforming

Abstract

Genetic Algorithm (GA) assisted beamforming techniques are proposed as an alternative to conventional beamforming algorithms. The design of the corresponding GAs is highlighted and the achievable performance is characterised in terms of the Signal-to-Interference Ratio (SIR) and the Signal-to-Interference plus Noise Ratio (SINR). It is demonstrated that an attractive SINR versus complexity trade-off is achievable by the proposed GA-assisted beamformers, provided that the GA parameters are appropriately chosen.

Implementation of a novel Fibonacci branch search optimizer for the design of the low sidelobe and deep nulling adaptive beamformer

“…Using the given endpoints X A and X B , we can determine the segmentation points X S1 in the first global search stage by equation (11).…”

“…According to the above, an enormous amount of evolutionary algorithms has been dedicated to applying several optimization approaches for ABF problems in the past decades. Many literatures have shown that these algorithms are capable of finding global or strong local optima of non-linear multimodal functions with multidimensional solutions [10,11]; therefore, weights of the beamformer extracted by the optimization techniques according to the fitness function defined by specific criterion can be used to place a maximum beam and null in an array pattern in specified locations. Compared to other evolutionary algorithms, the PSO is much easier to implement and outperform better; thus, many examples have been successfully demonstrated and validated the design flexibility of PSO in the framework of ABF arrays [12][13][14].…”

Application of genetic algorithm in satellite communication task scheduling

“…Genetic algorithms 5 , inspired by the genetic process in the biological world, continuously evolve and iterate to find the optimal solution to a problem. They are primarily applicable to planning problems such as beamforming and path planning [6][7] , as well as scheduling problems such as resource allocation, plan generation, and task scheduling [8][9][10] . Genetic algorithms have demonstrated strong optimization capabilities for complex and challenging problems in combinatorial optimization domain and have been widely applied 11 .…”

A Beamformer Design Based on Fibonacci Branch Search

“…The main advantage of the evolutionary heuristics algorithms over the classical derived approaches in antenna systems is that they have no requirements for extra iterative derivations or computationally extensive routines in objective functions of ABF model, and some of the evolutionary algorithms have been dedicated to beamformer implementations for their ability to search the global optimum. Approaches such as genetic algorithms (GA), particle swarm optimization (PSO), and other modified techniques are a set of optimization algorithms that have been suggested in the past decades to solve a variety of ABF problems [13,14]. Many researches have shown that the excitation weights extracted by these optimization techniques defined by specific criteria can be used to place a maximum beam and null in an array pattern in specified locations, and they also require relatively lower mathematical complexity than derivative-based or iterative-based ABF methods.…”