Design Optimization of LNAs and Reflectarray Antennas Using the Full-Wave Simulation-Based Artificial Intelligence Models with the Novel Metaheuristic Algorithms

Abstract

In this chapter, the two primarily important highly nonlinear design problems of the contemporary microwave engineering which are Low Noise Amplifier (LNA)s and Reflect-array Antenna (RA)s are solved as Design Optimization problems. For this purpose, firstly the design problem is defined in terms of the feasible design variables (FDVs), the feasible design target space (FDTS), both of which are built up by integrating the artificial intelligence black-box models based upon the measurements or full-wave simulations and a suitable metaheuristic search algorithm. In the second stage, feasible design target (FDT) or objective function of the optimization procedure is determined as a sub-space of the FDTS. Thirdly, the cost function evaluating the objective is minimized employing a suitable metaheuristic search algorithm with respect to the FDVs. Finally the completed designs are verified by the professional Microwave Circuitor3-D EM simulators.