NEURAL NETWORKS MODELING OF VARIOUS SHAPED MICROSTRIP ANTENNAS: A STATE OF THE ART REVIEW

Abstract

In the last decade, because of learning and generalization functionality, neural network-based modelling has been used to compute various output parameters of microstrip antennas. Most of the neural models generated are software simulation based. Because the neural networks naturally display significant parallelism, a parallel hardware must be generated by taking advantage of the parallelism of the neural networks to construct a faster computer machine. This paper demonstrates a generalized model of neural networks built on a reconfigurable hardware framework based on field programmable gate array (FPGA) to compute various performance parameters of microstrip antennas. Thus, the proposed approach offers a forum for microwave applications to build low-cost neural network-based FPGA simulators. The findings obtained by this method are also in very good agreement with the available calculated results in the literature.