Abstract
In this study, the reflection phase characterization of Reflectarray Antenna is established as a highly nonlinear function within the continuous domain of the element geometry and substrate parameters within the defined bandwidth centered the resonant frequency employing the 3-D Computer Simulation Technology Microwave Studio (CST MWS) -based Multi-Layer Perceptron Neural Network (MLP NN). Thus, the 4- dimensional Minkowski space is mapped into the one- dimensional reflection phase space by this MLP NN Black-box analysis model. Furthermore, this analysis model will be used in the Particle Swarm Optimization (PSO) process to determine the optimum substrate thickness and geometrical parameters of the Minkowski reflectarray. Finally, performance of the illustrated analysis model has been established.