Identification of Very Low Frequency (VLF) Whistlers transients using Feed Forward Neural Network (FFNN)

Abstract

The automatic identification of VLF whistler transients is an important practical goal for ionospheric and magnetospheric science because they give useful information regarding propagating medium particularly of the plasmasphere. We have developed a neural network based system to identify four types of whistlers (i.e diffuse, dispersive, multipath and spicky) recorded by DEMETER (Detection of electromagnetic emission form earthquake region) satellite. Wavelet transform is applied to extract the characteristics features of whistlers which are used to train the Feed Forward Neural Network (FFNN). The data required to train the network were collected from two year (2008-2010) observations of DEMETER satellite. The results show that the proposed FFNN can accurately identify the whistler transients.