Application Terminal GNSS and LEO Multimode Multifrequency Ionospheric Disturbance Monitoring Method

Abstract

This application relates to a method for monitoring ionospheric disturbances using GNSS and LEO multi-mode multi-frequency applications. The method includes: collecting observation data from the entire GNSS and LEO satellite systems at all frequencies; calculating the ionospheric amplitude scintillation index, phase index, and absolute / relative TEC index; outputting high-frequency adjustable raw data; integrating the four indices with space weather data to construct a training set; extracting feature factors after removing outliers using the isolated forest algorithm; inputting these features into an LSTM model for training and hyperparameter optimization; and finally inputting real-time index data into the trained model to achieve real-time monitoring of ionospheric disturbances. This method enables real-time monitoring of ionospheric scintillation phenomena.

Description

Technical Field

[0001] This application relates to the field of ionospheric disturbance monitoring technology, and in particular to a method for monitoring ionospheric disturbances using GNSS and LEO multi-mode multi-frequency applications. Background Technology

[0002] Ionospheric disturbances are accompanied by abrupt changes in the electron content within the regional ionospheric atmosphere, causing variations in the propagation delay of GNSS radio signals within the ionosphere, thereby affecting the positioning, timing, and reliability of satellite navigation systems. Existing ionospheric disturbance monitoring methods have several shortcomings: they do not support the processing of low Earth orbit (LEO) satellite signals, resulting in low spatial resolution of observational data; they lack the capability to monitor the ionospheric phase scintillation index and amplitude scintillation index across the entire system and all frequencies; they lack the capability to monitor ionospheric electron density (TEC); and they lack the capability to monitor and provide early warning of ionospheric disturbance phenomena. Summary of the Invention

[0003] Therefore, it is necessary to provide an application terminal GNSS and LEO multi-mode multi-frequency ionospheric disturbance monitoring method that can realize real-time monitoring of ionospheric scintillation phenomena, addressing the aforementioned technical problems.

[0004] A method for monitoring ionospheric disturbances using GNSS and LEO multimode multifrequency applications, the method comprising:

[0005] Step 1: Collect observation data of the entire GNSS and LEO satellite system and all frequency points. The observation data includes channel correlation accumulated values, carrier phase observation values, and pseudorange observation values.

[0006] Step 2: Calculate the ionospheric amplitude scintillation index based on the channel correlation cumulative value, calculate the ionospheric phase scintillation index based on the carrier phase observation value, construct the dual-frequency observation equation, calculate the absolute TEC index based on the dual-frequency pseudorange observation value in the dual-frequency observation equation, calculate the relative TEC index based on the dual-frequency carrier phase observation value, and output the four types of indices and high-frequency adjustable raw data;

[0007] Step 3: Collect raw data of four types of indices and high-frequency adjustable data, and simultaneously collect space weather data. Integrate the two types of data to build a model training set;

[0008] Step 4: Use the Isolation Forest algorithm to identify and remove outliers from the training set to obtain the processed training set;

[0009] Step 5: Extract features from the processed training set to obtain ionospheric perturbation-related feature factors;

[0010] Step 6: Input the ionospheric disturbance-related feature factors into the LSTM prediction model for model training and hyperparameter optimization;

[0011] Step 7: Input the high-frequency adjustable raw data output in real time in Step 2 into the LSTM prediction model trained and optimized in Step 6 to realize real-time monitoring and early warning of ionospheric disturbance phenomena.

[0012] The aforementioned application-based GNSS and LEO multi-mode multi-frequency ionospheric disturbance monitoring method overcomes the limitations of traditional methods that only support GNSS signals by simultaneously acquiring channel correlation accumulation values, pseudorange observations, and carrier phase observations across the entire GNSS and LEO satellite system and all frequency points. This significantly improves the spatial resolution and coverage of the observation data, providing more comprehensive and high-density raw data support for ionospheric scintillation and disturbance monitoring. Based on the aforementioned multi-source observation data, the ionospheric amplitude scintillation index, phase index, absolute TEC index, and relative TEC index are calculated respectively, and high-frequency adjustable raw data are output. This is the first time that synchronous monitoring of ionospheric scintillation characteristics and electron density content across the entire system and all frequency points has been achieved, making up for the shortcomings of traditional methods that cannot calculate the scintillation index and TEC. It can comprehensively characterize the amplitude, phase, and electron content variations of ionospheric scintillation; it integrates four types of ionospheric indices with space weather data to construct a training set, uses the isolated forest algorithm to remove anomalous noise data, and extracts key feature factors to input into the LSTM prediction model to complete training and hyperparameter optimization, effectively eliminating the influence of interference data on monitoring results, while enabling the model to learn the temporal evolution of ionospheric disturbances; it inputs the high-frequency index data output in real time into the trained and optimized LSTM model to achieve real-time monitoring and early warning of ionospheric disturbances, upgrading traditional passive observation to active prediction and early warning, significantly improving the accuracy and reliability of ionospheric disturbance monitoring, providing high-resolution support for navigation and communication systems, and ultimately achieving stable, accurate, and real-time monitoring of ionospheric scintillation phenomena. Attached Figure Description

[0013] Figure 1 This is a flowchart illustrating a method for monitoring GNSS and LEO multimode multifrequency ionospheric disturbances using an application terminal, as described in one embodiment.

[0014] Figure 2 This is a schematic diagram of the process architecture of a method for monitoring ionospheric disturbances using terminal GNSS and LEO multimode multifrequency in one embodiment;

[0015] Figure 3 This is a schematic diagram of the LSTM prediction model process in one embodiment. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0017] In one embodiment, such as Figure 1 and Figure 2 As shown, a method for monitoring ionospheric disturbances using GNSS and LEO multimode multifrequency applications is provided, comprising the following steps:

[0018] Step 1: Collect observation data of the entire GNSS and LEO satellite system and all frequency points. The observation data includes channel correlation accumulated values, carrier phase observation values, and pseudorange observation values.

[0019] The channel-related cumulative value is the receiver's sum of satellite signals for each 1 ms The relevant integration results; the pseudorange observations and carrier phase observations at all frequencies are the raw observation data of pseudorange and carrier phase under the full-frequency signals of GNSS and LEO satellites output by the receiver. This step covers the acquisition of signals from the entire system and all frequencies, providing a complete data source for subsequent index calculations.

[0020] Step 2: Calculate the ionospheric amplitude scintillation index based on the channel correlation cumulative value, calculate the ionospheric phase scintillation index based on the carrier phase observation value, construct the dual-frequency observation equation, calculate the absolute TEC index based on the dual-frequency pseudorange observation value in the dual-frequency observation equation, calculate the relative TEC index based on the dual-frequency carrier phase observation value, and output the four types of indices and high-frequency adjustable raw data.

[0021] The ionospheric amplitude scintillation index is the S4 index, used to characterize the intensity of ionospheric amplitude scintillation; the ionospheric phase scintillation index is used to characterize the degree of ionospheric phase scintillation disturbance; the absolute TEC index is the absolute total electron content of the ionosphere, reflecting the absolute level of ionospheric electron density; the relative TEC index is the relative total electron content of the ionosphere, reflecting the trend of ionospheric electron density variation; the high-frequency adjustable raw data is the observation and calculation result with configurable output frequency. In this application, the basic output frequency is 50Hz, which can meet the requirements of real-time monitoring and high-precision analysis; this step supports both GNSS and LEO satellite signal processing, breaking through the limitation of traditional methods that only support GNSS, and improving the resolution of space monitoring.

[0022] Step 3: Collect raw data of four types of indices and high-frequency adjustable data, and simultaneously collect space weather data. Integrate the two types of data to build a model training set.

[0023] Space weather data includes external environmental data such as solar activity, geomagnetic activity, and interplanetary disturbances related to ionospheric disturbances; the model training set mainly consists of historical time-series data, which integrates internal ionospheric feature data with external space weather data to provide complete and multi-dimensional training samples for the LSTM prediction model, ensuring the comprehensiveness and accuracy of the model's predictions.

[0024] Step 4: Use the Isolation Forest algorithm to identify and remove outliers from the training set to obtain the processed training set.

[0025] The Isolation Forest algorithm is an unsupervised anomaly detection algorithm suitable for identifying outliers in time-series monitoring data. It can quickly locate abnormal data caused by equipment failure, signal interference, and outliers. Outlier removal can eliminate the interference of noisy data on model training, improve the fitting accuracy and generalization ability of the LSTM prediction model, and avoid distortion of prediction results caused by outlier data.

[0026] Step 5: Extract features from the processed training set to obtain ionospheric perturbation-related feature factors.

[0027] Feature extraction involves dimensional reduction and key information extraction from the fused data. Ionospheric disturbance-related feature factors include ionospheric amplitude scintillation index, ionospheric phase scintillation index, absolute TEC index, relative TEC index, and space weather-related features. These feature factors can comprehensively characterize the state of ionospheric disturbance and provide effective input for subsequent prediction models.

[0028] Step 6: Input the ionospheric perturbation-related feature factors into the LSTM prediction model for model training and hyperparameter optimization.

[0029] The LSTM prediction model is a long short-term memory neural network, suitable for prediction and trend analysis of time series data, capable of capturing the temporal evolution patterns of ionospheric disturbances. Model training involves learning ionospheric disturbance change patterns based on historical feature factor data. Hyperparameter optimization involves tuning parameters such as the number of network layers, neurons, learning rate, and number of iterations to improve the model's prediction accuracy and convergence speed. Figure 3 The complete flow of the LSTM prediction model is shown below.

[0030] Step 7: Input the high-frequency adjustable raw data output in real time in Step 2 into the LSTM prediction model trained and optimized in Step 6 to realize real-time monitoring and early warning of ionospheric disturbance phenomena.

[0031] Real-time monitoring determines the current state of the ionosphere based on four types of indices calculated in real time; early warning predicts the trend of ionospheric disturbances in the future based on the LSTM model, and outputs early warning information when the predicted value exceeds the preset threshold, realizing the transformation from post-analysis to pre-warning mode, and meeting the needs of navigation, communication and other systems to respond quickly to ionospheric anomalies.

[0032] The aforementioned application-based GNSS and LEO multi-mode multi-frequency ionospheric disturbance monitoring method overcomes the limitation of traditional methods that only support GNSS signals by simultaneously acquiring channel correlation accumulation values, pseudorange observation values, and carrier phase observation values ​​across the entire GNSS and LEO satellite system and frequency points. Utilizing a massive number of LEO satellites, it constructs a GNSS+LEO multi-source satellite observation system, significantly improving the spatial resolution and coverage of the observation data. This provides more comprehensive and high-density raw data support for ionospheric scintillation and disturbance monitoring, facilitating refined modeling of ionospheric disturbances and anomalous regions. Based on the aforementioned multi-source observation data, the ionospheric amplitude scintillation index, phase index, absolute TEC index, and relative TEC index are calculated respectively, and high-frequency adjustable raw data is output. This achieves, for the first time, simultaneous monitoring of ionospheric scintillation characteristics and electron density content across the entire system and frequency points, addressing the shortcomings of traditional methods in calculating scintillation indices and TEC. It can comprehensively characterize the amplitude, phase, and electron content variation patterns of ionospheric scintillation. The four types of ionospheric indices are... The training set is constructed by fusing data from spatial weather data. Anomaly and noise data are removed using the isolated forest algorithm, and key feature factors are extracted and input into the LSTM prediction model to complete training and hyperparameter optimization. This effectively eliminates the influence of interference data on monitoring results and enables the model to learn the temporal evolution of ionospheric disturbances. The high-frequency exponential data output in real time is input into the trained and optimized LSTM model to achieve real-time monitoring and early warning of ionospheric disturbances. This upgrades traditional passive observation to active prediction and early warning, significantly improving the accuracy and reliability of ionospheric disturbance monitoring, providing high-resolution support for navigation and communication systems, and ultimately achieving stable, accurate, and real-time monitoring of ionospheric scintillation.

[0033] In one embodiment, the ionospheric amplitude scintillation index is calculated based on the channel-related cumulative value, including:

[0034] According to each channel ms Related cumulative value calculation every 20 ms The broadband energy and narrowband energy are calculated; then the signal strength at 50Hz frequency is calculated based on the broadband energy and narrowband energy, and then the signal strength is low-pass filtered to calculate the signal strength after de-trending.

[0035] The ionospheric amplitude scintillation index is calculated based on the signal strength data after detrending.

[0036] Specifically, this step involves 1 ms The relevant cumulative value is 20. ms Broadband / narrowband energy ensures the stability of signal strength calculation; low-pass filtering eliminates multipath effects and high-frequency noise; detrend processing removes slowly changing signal terms, highlighting rapid fluctuations caused by ionospheric scintillation, thus improving the accuracy of S4 index calculation and accurately reflecting the amplitude scintillation characteristics of the ionosphere.

[0037] In one embodiment, according to each channel ms Related cumulative value calculation every 20 ms Broadband and narrowband energy, including:

[0038] According to each channel ms Related cumulative value calculation every 20 ms The broadband and narrowband energies are respectively:

[0039] ;

[0040] ;

[0041] in, Indicates broadband energy. Indicates narrowband energy. Indicates the first i 1 ms In-phase components of the relevant accumulated values, Indicates the first i 1 ms Orthogonal components of the relevant cumulative values.

[0042] Specifically, every 20 ms A set of broadband and narrowband energy calculations were completed, with an output frequency of 50Hz, consistent with the subsequent signal strength and S4 exponent calculation frequencies to ensure data timing synchronization. The energy was calculated by the sum of the squares of the in-phase and quadrature components, which conforms to the satellite signal correlation processing principle and provides reliable basic data for signal strength calculation.

[0043] In one embodiment, calculating the signal strength at a frequency of 50 Hz based on broadband energy and narrowband energy includes:

[0044] The signal strength at 50Hz is calculated based on broadband and narrowband energy:

[0045] ;

[0046] in, Indicates broadband energy. This indicates narrowband energy.

[0047] Specifically, using the ratio of broadband energy to narrowband energy to calculate signal strength can effectively suppress the influence of receiver hardware gain changes on signal strength, so that the signal strength only reflects the signal fluctuations caused by ionospheric disturbances, thereby improving the anti-interference capability of subsequent scintillation index calculation.

[0048] In one embodiment, the signal strength is low-pass filtered, and the signal strength after descaling is calculated, including:

[0049] The signal strength is obtained by low-pass filtering. The signal strength after detrending is calculated as follows:

[0050] ;

[0051] in, Indicates signal strength.

[0052] Specifically, the low-pass filter is implemented using three cascaded second-order IIR filters, which can effectively filter out high-frequency interference such as multipath propagation and receiver thermal noise, while preserving the slow trend of signal change. The trend is eliminated by subtracting the filtered signal strength from the original signal strength, highlighting the rapid fluctuation component caused by ionospheric scintillation, so that the S4 exponent can accurately characterize the intensity of ionospheric disturbance. The ordinary differential equation of the second-order IIR filter is shown below:

[0053] .

[0054] In one embodiment, the ionospheric amplitude scintillation index is calculated based on the detrended signal strength data, including:

[0055] The ionospheric amplitude scintillation index, calculated based on the detrended signal strength data, is:

[0056] ;

[0057] in, This represents the signal strength data after detrending. To calculate the mean.

[0058] In one embodiment, the ionospheric phase scintillation index is calculated based on carrier phase observations, including:

[0059] Carrier phase observations are divided into 20 ms The carrier phase observation at a frequency of 50 Hz was extracted in one step;

[0060] High-pass filtering of the carrier phase observations yields the de-trending carrier phase.

[0061] A sixth-order IIR filter is used to process the detrended carrier phase, and the mean value of the processed carrier phase data is calculated to obtain the ionospheric phase scintillation index.

[0062] Specifically, 50Hz carrier phase observations can capture rapid changes in ionospheric phase; high-pass filtering can eliminate slow phase changes caused by relative motion between the satellite and receiver, satellite clock bias, and receiver clock bias, retaining only phase fluctuations caused by ionospheric disturbances; the 6th-order IIR filter has good phase-preserving characteristics, avoiding phase distortion introduced by filtering and improving the accuracy of phase scintillation index calculation. The ordinary differential equation of the 6th-order IIR filter is shown below:

[0063] .

[0064] In one embodiment, the ionospheric phase scintillation index is calculated by averaging the processed carrier phase data, including:

[0065] The ionospheric phase scintillation index is calculated by averaging the processed carrier phase data as follows:

[0066] ;

[0067] in, To calculate the mean.

[0068] In one embodiment, the pseudorange observations include dual-frequency pseudorange observations and dual-frequency carrier phase observations; the absolute TEC index is calculated based on the dual-frequency pseudorange observations, including:

[0069] The absolute TEC index is calculated based on dual-frequency pseudorange observations as follows:

[0070] ;

[0071] in, , The center frequencies of the two signals are... The difference between the two-frequency pseudorange observations. , These are pseudorange observations at two frequency points. The pseudorange hardware delay deviation between the two frequency points. , The pseudorange hardware delay deviation of the receiver at the two frequency points.

[0072] Specifically, the absolute TEC index is calculated based on dual-frequency pseudorange observations. By correcting the pseudorange hardware delay deviation between the satellite and the receiver, the influence of equipment hardware errors on the TEC calculation is eliminated, and the absolute total electron content of the ionosphere is obtained. It can directly reflect the absolute level of ionospheric electron density and provide a quantitative indicator for monitoring the background state of the ionosphere.

[0073] In one embodiment, the relative TEC index is calculated based on dual-frequency carrier phase observations, including:

[0074] The relative TEC index is calculated based on dual-frequency carrier phase observations:

[0075] ;

[0076] in, , The center frequencies of the two signals are... , These are carrier phase observations at two frequency points. , The carrier phase wavelengths at two frequency points, , For the carrier phase integer ambiguity at two frequency points, The hardware delay deviation of the satellite carrier phase between two frequency points. The receiver carrier phase hardware delay deviation between the two frequency points. This represents the combined path length difference of the dual-frequency carrier phase observations.

[0077] Specifically, the relative TEC index is calculated based on dual-frequency carrier phase observations, correcting integer ambiguity and hardware delay deviations in carrier phase between the satellite and receiver. It can accurately reflect the relative changes in the total electron content of the ionosphere, with a higher sensitivity than the absolute TEC index. It can capture weak ionospheric disturbances and improve the sensitivity of disturbance monitoring.

[0078] In a specific embodiment, the phase scintillation index, amplitude scintillation index, relative TEC index, and absolute TEC index of the ionosphere at all frequencies of GNSS and LEO satellites are calculated in real time, and high-frequency adjustable raw data are output. At the same time, the monitoring results of ionospheric disturbances can be displayed in real time through host computer software.

[0079] It should be understood that, although Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0080] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0081] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for monitoring GNSS and LEO multi-mode multi-frequency ionospheric disturbances, characterized in that, The method includes: Step 1: Collect observation data of the entire GNSS and LEO satellite system and all frequency points. The observation data includes channel correlation accumulated values, carrier phase observation values, and pseudorange observation values. Step 2: Calculate the ionospheric amplitude scintillation index based on the channel correlation accumulated value, calculate the ionospheric phase scintillation index based on the carrier phase observation value, construct the dual-frequency observation equation, calculate the absolute TEC index based on the dual-frequency pseudorange observation value in the dual-frequency observation equation, calculate the relative TEC index based on the dual-frequency carrier phase observation value, and output the four types of indices and high-frequency adjustable raw data; Step 3: Collect raw data of four types of indices and high-frequency adjustable data, and simultaneously collect space weather data. Integrate the two types of data to build a model training set; Step 4: Use the Isolation Forest algorithm to identify and remove outliers from the training set of the model to obtain the processed training set; Step 5: Extract features from the processed training set to obtain ionospheric perturbation-related feature factors; Step 6: Input the ionospheric disturbance-related feature factors into the LSTM prediction model for model training and hyperparameter optimization; Step 7: Input the high-frequency adjustable raw data output in real time in Step 2 into the LSTM prediction model trained and optimized in Step 6 to realize real-time monitoring and early warning of ionospheric disturbance phenomena.

2. The method of claim 1, wherein, The ionospheric amplitude scintillation index is calculated based on the channel-related cumulative values, including: According to the channel every ms The relevant cumulative value calculates every 20 ms The wideband energy and narrowband energy; again based on the wideband energy and narrowband energy, the signal strength of 50Hz frequency is calculated, and then the signal strength is low-pass filtered, and the signal strength after the trend is calculated. The ionospheric amplitude scintillation index is calculated based on the signal strength data after the trend is eliminated.

3. The method according to claim 2, characterized in that, According to the channel every ms The relevant cumulative value is calculated every 20 ms The wideband energy and the narrowband energy include: According to each channel ms Related cumulative value calculation every 20 ms The broadband and narrowband energies are respectively: in, Indicates broadband energy. Indicates narrowband energy. Indicates the first i 1 ms In-phase components of the relevant accumulated values, Indicates the first i 1 ms Orthogonal components of the relevant cumulative values.

4. The method according to claim 2, characterized in that, Calculating the signal strength at 50Hz based on broadband and narrowband energy includes: The signal strength at 50Hz is calculated based on broadband and narrowband energy: in, Indicates broadband energy. This indicates narrowband energy.

5. The method according to claim 2, characterized in that, The signal strength is low-pass filtered, and the signal strength after descaling is calculated, including: The signal strength is obtained by low-pass filtering. The signal strength after detrending is calculated as follows: in, Indicates signal strength.

6. The method according to claim 2, characterized in that, The ionospheric amplitude scintillation index is calculated based on the detrended signal strength data, including: Based on the detrended signal strength data, the ionospheric amplitude scintillation index is calculated as follows: in, This represents the signal strength data after detrending. To calculate the mean.

7. The method according to claim 1, characterized in that, The ionospheric phase scintillation index is calculated based on the carrier phase observations, including: Carrier phase observations are divided into 20 ms The carrier phase observations at a frequency of 50 Hz were extracted in one step. High-pass filtering is applied to the carrier phase observations to obtain the de-trending carrier phase; A sixth-order IIR filter is used to process the detrended carrier phase, and the ionospheric phase scintillation index is calculated based on the mean of the processed carrier phase data.

8. The method according to claim 7, characterized in that, The ionospheric phase scintillation index is calculated by averaging the processed carrier phase data, including: The ionospheric phase scintillation index is calculated by averaging the processed carrier phase data as follows: in, To calculate the mean.

9. The method according to claim 1, characterized in that, The pseudorange observations include dual-frequency pseudorange observations and dual-frequency carrier phase observations; the absolute TEC index is calculated based on the dual-frequency pseudorange observations, including: The absolute TEC index is calculated based on dual-frequency pseudorange observations as follows: in, , The center frequencies of the two signals are... The difference between the two-frequency pseudorange observations. , These are pseudorange observations at two frequency points. The pseudorange hardware delay deviation between the two frequency points. , The pseudorange hardware delay deviation of the receiver at the two frequency points.

10. The method according to claim 1, characterized in that, The relative TEC index is calculated based on dual-frequency carrier phase observations, including: The relative TEC index is calculated based on dual-frequency carrier phase observations: in, , The center frequencies of the two signals are... , These are carrier phase observations at two frequency points. , The carrier phase wavelengths at two frequency points, , For the carrier phase integer ambiguity at two frequency points, The hardware delay deviation of the satellite carrier phase between two frequency points. The receiver carrier phase hardware delay deviation between the two frequency points. This represents the combined path length difference of the dual-frequency carrier phase observations.

Images