Neural Network Approach to Diagnose Faults of Antenna Array

In this paper a neural network based technique to diagnose the type of fault occurring in antenna array is presented. The faults that can arise in an antenna array are classified as on off fault, current magnitude and phase fault, positional fault and frequency fault. These faults change the radiation pattern features like main lobe level, side lobe level etc, which may be highly unacceptable in many scenarios. The antenna array is diagnosed for proper functioning by observing changes in radiation pattern. A uniform linear array of 101 elements is considered for training the neural network. The input to the neural network is amplitude of deviation pattern and output of neural network is the type of fault. The performance of neural network is compared with probabilistic neural network and radial basis function neural network. Both networks showed high success rate. (5 pages)