Microstrip reflectarray antennas

Printed board reflector antennas are the antennas which combine advantages of both the parabolic reflectors and the planar phased arrays [1- 6]. These antennas consist of the microstrip patch or aperture element antennas placed on a flat dielectric substrate and are illuminated by a feed antenna. These types of the antennas act as reactive surfaces, and reflect approximately all the electromagnetic waves with a proper phase shift incident upon their surfaces, thus the radiation is maximized in a required (?0,?0) direction. Since many elements are involved in the phasing process, the reflectarrays offer a lot of flexibilities in the pattern formation such as beam shaping, single/ multiple beam, beam scanning, reconfigurable beams. On the other hand, parabolic reflectors are difficult to manufacture due to their curved structure, at the same time the feed networks of the phased arrays are complicated and expensive because of the beam forming circuits and amplifier modules. In the microstrip reflectarrays, the beam shaping facilities are low cost, since in the microstrip reflectarrays using the printed dipoles or patches, phase shifting can be achieved by varying geometric dimensions of the elements on a proper dielectric substrate around their resonant dimensions. However the advanced novel element configurations with several degrees of freedom are also needed to satisfy pattern shaping requirements and to enhance the frequency behaviour and bandwidth. Management of the different parameters and the need of satisfying requirements that could also be in opposite each other could however make the design of a reflectarray quite complex.In design of a reflectarray, there are the two main problems: Pattern shaping and bandwidth. Fundamental tool to meet these requirements is a fast and accurate model of the unit cell. Particularly, microstrip reflectarrays employe patches with the novel and complicated geometries. Besides one must have an accurate phase versus element geometry calibration characteristic so that the element is able to provide the necessary phase shift to focus the main beam to a desired (?0, ?0) direction. In order to obtain the calibration phase characteristic and the corresponding optimum parameters of the element, an optimization is performed using the artificial intelligence model of the element. Here the optimum element is meant to have the phase characteristic with at least the 3600 phase range and a small gradient with respect to the geometry parameters. Thus sensitivity with respect to the fabrication errors will be minimized and bandwidth will be enhanced. Multilayer Perceptron Neural Network (MLPNN) will be employed in the artificial intelligence modeling of the unit cell [13]. The training and validation data for the element modellng is obtained using the ?infinite array? approach that takes into account the mutual coupling effects and specular reflections from the ground plane itself. The infinite array approach assumes all the elements to be identical and it employes ?the mathematical waveguide? simulator terminated by the unit cell consisting of a single element where the HFSS or CST commercial full-wave simulators are employed to obtain the reflection phase information of the element. Here the side and upper and lower walls of mathematical waveguide are the perfect magnetic and electric, respectively and EM analysis is done for the incidence TEM wave that can be considered as a good approximation for the TM and TE waves up to 400 tool [6]As a brief summary, artificial intelligence such as MLPNN will be used for the design and analysis of the microstrip and waveguide reflectarrays, respectively, then furthermore the full wave electromagnetic simulations will be obtained using Computer Simulation Technology Microwave Studio (CST MWS).