Implementation of Hybrid Deep Learning Model (LSTM-CNN) for Ionospheric TEC Forecasting Using GPS Data

Ruwali, Adarsha; Kumar, Arun; Prakash, Kolla Bhanu; Sivavaraprasad, G.; Ratnam, D. Venkata

doi:10.1109/lgrs.2020.2992633

Cited by 81 publications

(60 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…By the addition of connections with the past data, the deep learning method uses the result at the previous step as an input at the current step to learn time-series data. Based on deep learning, many forecasting methods are developed for ionospheric parameters prediction, such as a recurrent neural network (RNN) method [23,24], a long-short-term memory LSTM method [25,26], and an improved gated recurrent unit (GRU) method [27]. Unlike traditional neural networks models [28], deep learning models can infer the relationships between the previous time node and the latter time node of time-series data.…”

Section: Introductionmentioning

confidence: 99%

Forecasting Ionospheric foF2 Based on Deep Learning Method

Zhou

Tang

et al. 2021

Remote Sensing

View full text Add to dashboard Cite

In this paper, a deep learning long-short-term memory (LSTM) method is applied to the forecasting of the critical frequency of the ionosphere F2 layer (foF2). Hourly values of foF2 from 10 ionospheric stations in China and Australia (based on availability) from 2006 to 2019 are used for training and verifying. While 2015 and 2019 are exclusive for verifying the forecasting accuracy. The inputs of the LSTM model are sequential data for the previous values, which include local time (LT), day number, solar zenith angle, the sunspot number (SSN), the daily F10.7 solar flux, geomagnetic the Ap and Kp indices, geographic coordinates, neutral winds, and the observed value of foF2 at the previous moment. To evaluate the forecasting ability of the deep learning LSTM model, two different neural network forecasting models: a back-propagation neural network (BPNN) and a genetic algorithm optimized backpropagation neural network (GABP) were established for comparative analysis. The foF2 parameters were forecasted under geomagnetic quiet and geomagnetic disturbed conditions during solar activity maximum (2015) and minimum (2019), respectively. The forecasting results of these models are compared with those of the international reference ionosphere model (IRI2016) and the measurements. The diurnal and seasonal variations of foF2 for the 4 models were compared and analyzed from 8 selected verification stations. The forecasting results reveal that the deep learning LSTM model presents the optimal performance of all models in forecasting the time series of foF2, while the IRI2016 model has the poorest forecasting performance, and the BPNN model and GABP model are between two of them.

show abstract

Section: Introductionmentioning

confidence: 99%

Forecasting Ionospheric foF2 Based on Deep Learning Method

Zhou

Tang

et al. 2021

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Firstly, the precise orbit of GRACE-FO without high-order ionospheric delay correction is obtained by using the 15-day unprocessed observations of DOY 71-85 in 2019. Then the high-order ionospheric delay correction is calculated by Equations (8)- (12). The satellite-borne GPS pseudo-range and phase observations are corrected according to Equation (6) and Equation 7, and the corrected data are used for POD by way of simplified dynamics.…”

Section: Results and Analysis Of Podmentioning

confidence: 99%

“…Ionospheric delay is one of the main error terms in satellite-borne GPS POD technology. The variation of electron density distribution with time and space results in the complexity of ionospheric delay [10][11][12] . How to deal with ionospheric delay is an important factor to improve POD precision.…”

Section: Introductionmentioning

confidence: 99%

Effect of Higher-Order Ionospheric Delay on Precise Orbit Determination of GRACE-FO Based on Satellite-Borne GPS Technique

Guo

Xia

et al. 2021

IEEE Access

View full text Add to dashboard Cite

Ionospheric delay is one of the main errors in precise orbit determination (POD) of Gravity Recovery and Climate Experiment Follow-On (GRACE-FO) with satellite-borne GPS technique. Effective elimination or reduction of ionospheric delay can improve POD precision. For satellite-borne GPS dualfrequency data, the ionospheric-free linear combination is generally used to eliminate the first-order term, while igoring the influence of higher-order terms. To improve the POD precision of GRACE-FO, this paper presents an effective method to calculate higher-order ionospheric delay of GRACE-FO observation. Dualfrequency GPS receiver observation is used to calculate the total electron content on the signal propagation path. The magnetic field strength is calculated according to international geomagnetic reference field model, and the angle between the GPS signal propagation path and the geomagnetic field is calculated. Finally, the delays of second-order and third-order terms are calculated and introduced into GRACE-FO reduceddynamic POD. The effect of higher-order ionospheric delay on POD of GRACE-FO is analyzed in detail. The results show that the influence of ionospheric higher-order term on satellite-borne GPS observation is in centimeter level. The orbit precision of GRACE-FO can be improved by adding higher-order ionospheric delay, with improvement order of sub-millimeter level.

show abstract

“…To address these limitations and simultaneously to retain the advantages of deep learning in approximating temporal dependencies, long short-term memory (LSTM) architectures have been recently proposed [29]. In the context of TEC modeling, LSTM networks memorize temporal correlations of the TEC signals, therefore providing better modeling capabilities [30]- [32].…”

Section: A Related Workmentioning

confidence: 99%

Deep Recurrent Neural Networks for Ionospheric Variations Estimation Using GNSS Measurements

Kaselimi

Voulodimos

Doulamis

et al. 2022

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

Modeling ionospheric variability throughout a proper total electron content (TEC) parameter estimation is a demanding, however, crucial, process for achieving better accuracy and rapid convergence in precise point positioning (PPP). In particular, the single-frequency PPP (SF-PPP) method lacks accuracy due to the difficulty of dealing adequately with the ionospheric error sources. In order to apply ionosphere corrections in techniques, such as SF-PPP, external information of global ionosphere maps (GIMs) is crucial. In this article, we propose a deep learning model to efficiently predict TEC values and to replace the GIM-derived data that inherently have a global character, with equal or better in accuracy regional ones. The proposed model is suitable for predicting the ionosphere delay at different locations of receiver stations. The model is tested during different periods of time, under different solar and geomagnetic conditions and for stations in various latitudes, providing robust estimations of the ionospheric activity at the regional level. Our proposed model is a hybrid model comprising of a 1-D convolutional layer used for the optimal feature extraction and stacked recurrent layers used for temporal time series modeling. Thus, the model achieves good performance in TEC modeling compared to other state-of-the-art methods.

show abstract

Implementation of Hybrid Deep Learning Model (LSTM-CNN) for Ionospheric TEC Forecasting Using GPS Data

Cited by 81 publications

References 12 publications

Forecasting Ionospheric foF2 Based on Deep Learning Method

Forecasting Ionospheric foF2 Based on Deep Learning Method

Effect of Higher-Order Ionospheric Delay on Precise Orbit Determination of GRACE-FO Based on Satellite-Borne GPS Technique

Deep Recurrent Neural Networks for Ionospheric Variations Estimation Using GNSS Measurements

Contact Info

Product

Resources

About