Prediction of total electron content of the ionosphere using neural network

This paper presents the prediction of hourly vertical total electron content (vtec) using a neural network by utilizing the data from a gps ionospheric scintillation and tec monitor (gistm) receiver for six years (from 2005 to 2010) during low to medium solar activity (sunspot number (ssn) between 0.0 and 42.6). several network configurations were investigated to observe the effect of the number of neurons,and hidden layers. overall testing process for several network set-up yielded root mean square error (rmse) value of 3 to 7 tecu,absolute error of 2 to 6 tecu and relative error of 8% to 28%. testing using april 2010 to november 2010 data (ssn from 8.0 to 25.2) produced rmse value of 2.95 to 3.88 tecu,absolute error of 2.39 to 3.09 tecu and relative error of 8.11% to 16.18%,which are within the acceptable range. © 2016 penerbit utm press. all rights reserved.