Linear Antenna Array Parameter Estimation using a Neuro-computational Approach

The power of estimating after learning as provided by artificial neural network has helped a lot of researchers to determine solution to many a problem in antenna research. The solutions to most of the problems involved unwieldy and lengthy computations. In this paper, the relationship of beamwidth and gain with respect to the number of antenna elements in a linear antenna system is modelled using a neuro-computational model. The model uses a multi-layer perceptron network. It was found that the network accurately modeled the relationship. The mean square error (MSE) was found to be of the order of 10−9.