ABSTRACT
The demand for mobile communication is exponentially increasing in the 5/6G era with many novel technologies such as the Internet of Things (IoT), Industry 4.0, autonomous vehicles, smart cities, and smart healthcare. The need for better coverage, improved capacity, and higher transmission speeds is on the rise. The efficient usage of the radio frequency spectrum is one of the solutions to handle the increasing demand and technical constraints. Adaptive beamforming array antenna is a vital component of the above solution and it plays a key role in the 5G (2020) implementation and in the future 6G (2030) wireless system. This chapter presents both linear and non-linear antennas using perceptron ANN learning algorithms and its coding for adaptive beamforming. The perceptron ANN algorithm calculates the optimum weights of both linear and non-linear antenna arrays to steer the radiation pattern by directing multiple narrow beams toward the desired users and creating nulls toward interferers. The single-layer perceptron ANN is shown to give accurate beamforming with both minimum computational time and electronic memory. Four major types of activation functions are commonly used to steer the beam toward the desired direction. These are implemented in the algorithm for both linear and non-linear smart antenna 362arrays with different physical configurations. The MATLABTM codes for perceptron ANN algorithm to drive the smart antenna and that of the traditional Least Mean Square (LMS) algorithm driven antenna are given. The perceptron ANN driven smart antenna can perform adaptive beamforming at low computational cost while demonstrating good accuracy and fast convergence time.
