An Ionospheric TEC Prediction Method Based on Convolution Operation and LSTM Neural Network

The total electron content (TEC) is an important physical quantity of ionosphere, which has a great influence on navigation, positioning, communication and radar. The short-term prediction of TEC has been a hot research in the field, and TEC prediction research based on time series prediction method, such as Long Short-term Memory (LSTM) neural network, has made great progress. In this paper, the characteristic data of ionospheric TEC after convolution operation, together with ionospheric TEC, day of the year and local time, are used as inputs to construct a LSTM neural network to predict ionospheric TEC. Simulation results showed that compared with other simple LSTM neural networks, the improved ionospheric TEC prediction method based on convolution operation and LSTM neural network in this paper can improve the prediction accuracy of ionospheric TEC to a certain extent, can effectively improve the accuracy of TEC prediction.