A correction method for X-band tropospheric scatter prediction model based on measured data

Abstract

The present invention provides a method for correcting an X-band tropospheric scatter prediction model based on measured data. The method comprises: obtaining measured tropospheric scatter path loss data and dividing it into a training set and a test set; calculating the median loss using the model to be corrected; constructing a characteristic matrix of a weighted linear regression model based on the number of communication links and the median radio wave propagation loss; constructing a target variable vector based on the training set; constructing a weight matrix; establishing a weighted linear least squares model; calculating regression parameters; and correcting the model to be corrected based on the regression parameters. This embodiment applies the weighted linear least squares method and fully incorporates a large amount of measured path loss data at different over-the-horizon communication distances at sea in the X-band to further correct the tropospheric scatter empirical model to be corrected. The corrected tropospheric scatter radio wave propagation prediction model can more accurately predict tropospheric scatter radio wave propagation path loss.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a method for correcting an X-band tropospheric scatter prediction model based on measured data. Background Art

[0002] Numerous scatterers in the troposphere (such as various vortexes, clouds, warm and cold fronts, and horizontal stratification) cause refraction and re-radiation of frequencies above very high frequencies, particularly microwaves and millimeter waves, enabling radio waves to propagate beyond line of sight. This propagation mechanism is known as tropospheric scatter. Using tropospheric scatter for beyond-line-of-sight communications offers advantages such as high speed, beyond-line-of-sight, nuclear explosion resistance, security, confidentiality, and maneuverability. Therefore, tropospheric scatter is an important communication method. In particular, in offshore beyond-line-of-sight communications, tropospheric scatter can achieve a maximum communication range of 600 to 700 kilometers.

[0003] The use of tropospheric scatter for maritime beyond-horizon communications and system performance evaluation requires accurate prediction of radio propagation loss using tropospheric scatter models. Currently, commonly used models for calculating tropospheric scatter radio propagation path loss are empirical models, such as those developed by the ITU (International Telecommunication Union) and the tropospheric scatter model proposed by Academician Zhang Minggao of the 22nd Institute of the China Electronics Technology Group Corporation. These empirical models calculate tropospheric scatter propagation loss in different climate zones based on varying link lengths and frequencies. However, these empirical models have not been fully validated with measured path loss data for different beyond-horizon communication distances in the maritime X-band. This results in a certain discrepancy between the model predictions and the measured path loss, making it difficult to accurately evaluate the performance of maritime beyond-horizon communication systems.

[0004] Therefore, it is necessary to improve one or more problems existing in the above-mentioned related technical solutions.

[0005] It should be noted that this section is intended to provide background or context for the technical solutions of the present disclosure stated in the claims. The description herein is not admitted to be prior art by virtue of being included in this section. Summary of the Invention

[0006] The object of the present invention is to provide a method for correcting an X-band tropospheric scatter prediction model based on measured data, thereby overcoming one or more problems caused by the limitations and defects of related technologies, at least to a certain extent.

[0007] The present invention provides a method for correcting an X-band tropospheric scatter prediction model based on measured data, comprising:

[0008] Obtain X-band tropospheric scatter path loss data at different beyond-horizon communication distances at sea, preprocess the data, and divide the preprocessed data into training and test sets;

[0009] The median value of radio wave propagation loss at different beyond-horizon communication distances at sea is calculated using the model to be corrected.

[0010] constructing a characteristic matrix of a weighted linear regression model according to the number of communication links and the median value of the radio wave propagation loss;

[0011] Constructing a target variable vector of a weighted linear regression model based on the training set;

[0012] Construct the weight matrix of the weighted linear regression model;

[0013] A weighted linear least squares model is established based on the characteristic matrix and target variable vector of the weighted linear regression model;

[0014] Calculating the regression parameters of the weighted linear least squares model according to the characteristic matrix, the weight matrix and the weighted linear least squares model of the weighted linear regression model;

[0015] The model to be corrected is corrected according to the regression parameters to obtain a corrected X-band tropospheric scatter prediction model.

[0016] In the present invention, the correction method further includes:

[0017] The revised X-band tropospheric scatter prediction model is used to calculate the revised median of the X-band tropospheric scatter radio wave propagation loss, and the calculated results are compared with the data of the test set to evaluate the accuracy and applicability of the revised X-band tropospheric scatter prediction model.

[0018] In the present invention, the process of preprocessing data includes:

[0019] The measured path loss data is segmented at preset time intervals, and the average value of the measured path loss data in each time period is calculated, and the average value is used as the data of the training set and the test set.

[0020] In the present invention, the average value is randomly divided into a training set and a test set in a ratio of 7:3.

[0021] In the present invention, the weight value of each communication link in the weight matrix is ​​the number of time periods divided by the communication link.

[0022] In the present invention, the step of constructing a characteristic matrix of a weighted linear regression model according to the number of communication links and the median of the radio wave propagation loss includes:

[0023] The beyond-line-of-sight communication distances are arranged from short to long, the corresponding calculated median of the radio wave propagation loss is used as the first column of the feature matrix, the second column of the feature matrix is ​​assigned to all 1 columns, and the number of the communication links is used as the number of rows of the feature matrix.

[0024] In the present invention, the expression for calculating the median value of radio wave propagation loss using the modified X-band troposcatter prediction model is as follows:

[0025]

[0026] in, is the median value of radio wave propagation loss calculated using the modified X-band troposcatter prediction model; and are the slope correction coefficient and the horizontal offset correction term, respectively, and are regression parameters in the weighted linear least squares model; is the median value of the radio wave propagation loss calculated using the model to be corrected.

[0027] The technical solution provided by the present invention can have the following beneficial effects:

[0028] This method, based on measured data, refines an existing empirical tropospheric scatter model using a weighted linear least squares method. This method incorporates extensive measured path loss data from X-band maritime over-the-horizon communication at various distances. The model further refines the existing empirical tropospheric scatter model. Compared to existing empirical models, the revised X-band tropospheric scatter prediction model more accurately predicts the propagation path loss of tropospheric scatter waves, providing technical guidance for selecting power amplifiers and transceiver antenna modules for maritime over-the-horizon communication systems. This allows for a margin within the maximum allowable link path loss. Furthermore, the revised model accurately predicts the maximum communication range of the communication system by combining maritime over-the-horizon communication system parameters (such as operating frequency, transmit power, and transmit and receive antenna gain). BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The accompanying drawings are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the specification, are used to explain the principles of the present disclosure. Obviously, the drawings described below are only some embodiments of the present disclosure, and those skilled in the art can derive other drawings based on these drawings without inventive effort.

[0030] Figure 1 A flowchart showing a method for correcting an X-band tropospheric scatter prediction model based on measured data in an exemplary embodiment of the present disclosure is shown;

[0031] Figure 2A schematic diagram showing the prediction of path loss at different communication distances by the modified scattering model in an exemplary embodiment of the present disclosure;

[0032] Figure 3 A comparison chart showing the predicted results of the corrected model of the present invention and the model to be corrected and the measured path loss at 150 km on June 25, 2024 is shown;

[0033] Figure 4 A comparison chart showing the predicted results of the corrected model of the present invention and the model to be corrected and the measured path loss at 150 km on July 7, 2024 is shown;

[0034] Figure 5 A comparison chart showing the predicted results of the corrected model of the present invention and the model to be corrected and the measured path loss at 188 km on June 25, 2024 is shown;

[0035] Figure 6 A comparison chart showing the predicted results of the corrected model of the present invention and the model to be corrected and the measured path loss at 188 km on June 27, 2024 is shown;

[0036] Figure 7 A comparison chart showing the predicted results of the corrected model of the present invention and the model to be corrected and the measured path loss at 188 km on July 3, 2024 is shown;

[0037] Figure 8 A comparison chart showing the predicted results of the corrected model of the present invention and the model to be corrected and the measured path loss at 188 km on July 4, 2024 is shown;

[0038] Figure 9 A comparison chart showing the predicted results of the corrected model of the present invention and the model to be corrected and the measured path loss at 248 km on July 11, 2024 is shown;

[0039] Figure 10 A comparison chart showing the predicted results of the corrected model of the present invention and the model to be corrected and the measured path loss at 248 km on July 15, 2024 is shown;

[0040] Figure 11 The absolute error distribution diagram of the predicted results of the corrected model of the present invention and the model to be corrected and the measured path loss at 150 km is shown;

[0041] Figure 12 The absolute error distribution diagram of the predicted results of the corrected model of the present invention and the model to be corrected and the measured path loss at 188 km is shown;

[0042] Figure 13 The absolute error distribution diagram of the predicted results of the corrected model of the present invention and the model to be corrected and the measured path loss at 248 km is shown. DETAILED DESCRIPTION

[0043] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be embodied in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete and will fully convey the concepts of the example embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0044] In addition, the accompanying drawings are merely schematic illustrations of embodiments of the present disclosure and are not necessarily drawn to scale. Like reference numerals in the figures represent like or similar parts, and thus repeated descriptions thereof will be omitted. Some of the blocks shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically separate entities.

[0045] This example implementation provides a method for correcting an X-band troposcatter prediction model based on measured data. Figure 1 The method may include: S101-S108. Specifically as follows:

[0046] S101, obtaining X-band tropospheric scattering measured path loss data at different beyond-horizon communication distances at sea, preprocessing the data, and dividing the preprocessed data into a training set and a test set;

[0047] S102, calculating the median value of radio wave propagation loss at different beyond-horizon communication distances at sea using the model to be corrected;

[0048] S103, constructing a characteristic matrix of a weighted linear regression model according to the number of communication links and the median value of the radio wave propagation loss;

[0049] S104, constructing a target variable vector of a weighted linear regression model according to the training set;

[0050] S105, constructing a weight matrix of a weighted linear regression model;

[0051] S106, establishing a weighted linear least squares model based on the characteristic matrix of the weighted linear regression model and the target variable vector;

[0052] S107, calculating the regression parameters of the weighted linear least squares model according to the characteristic matrix, the weight matrix and the weighted linear least squares model of the weighted linear regression model;

[0053] S108 , correcting the model to be corrected according to the regression parameters to obtain a corrected X-band tropospheric scatter prediction model.

[0054] In this embodiment, a modified empirical tropospheric scatter model was further refined using a weighted linear least squares method, incorporating extensive measured path loss data for X-band maritime beyond-horizon communication distances. Compared to existing empirical models, the modified X-band tropospheric scatter prediction model more accurately predicts the propagation path loss of tropospheric scatter waves. This provides technical guidance for the selection of power amplifiers and transceiver antenna modules during the development of maritime beyond-horizon communication systems, ensuring a margin for the maximum link path loss allowed by the communication system. Furthermore, the modified model can accurately predict the maximum communication range of the communication system by combining maritime beyond-horizon communication system parameters (such as operating frequency, transmit power, and transmit and receive antenna gain).

[0055] It should be noted that the corrected model in this invention refers to the corrected X-band tropospheric scatter prediction model. The scattering model before correction and the original model both refer to the models to be corrected. The path loss estimate is the path propagation loss value calculated using the scattering model, which is also the median propagation loss.

[0056] The specific process of each step in the above embodiment is described below.

[0057] In S101, after obtaining the measured X-band tropospheric scatter path loss data at different over-the-horizon communication distances at sea, the data is preprocessed as follows:

[0058] The measured path loss data is segmented at preset time intervals, and the average value of the measured path loss data within each time interval is calculated. This average value is used as the data for the training and test sets. Specifically, the measured path loss data can be segmented at 10-minute intervals, and the average value of the measured path loss data within each time interval is calculated. These average values ​​are then randomly divided into the training and test sets in a ratio of 7:3. The training set data is used to construct a tropospheric scatter prediction model based on the weighted linear least squares method, and the test set data is used to verify the accuracy of the revised tropospheric scatter prediction model.

[0059] In S102, the model to be corrected in this application refers to the empirical tropospheric scattering model proposed by Academician Zhang Minggao (i.e., formula (1)). The model to be corrected is used to calculate the median radio wave propagation loss at different over-the-horizon communication distances at sea. , the calculation formula is:

[0060] (1)

[0061] in, is the meteorological factor (dB), is the communication frequency (MHz), is the minimum scattering angle (mrad), is the communication distance (km), is the exponential decay coefficient of tropospheric inhomogeneity intensity with height ( is the height from the lowest scattering point to the line connecting the transmitting and receiving points (km), is the height of the lowest scattering point above the ground (km), is the interface dielectric coupling loss (dB), and Represents the transmit and receive antenna gains (dB) respectively.

[0062] S103~S105 are the steps of constructing a weighted linear regression model.

[0063] In S103, a characteristic matrix of a weighted linear regression model is constructed based on the number of communication links and the median of the radio wave propagation loss. Represented by, where N represents the number of communication links, is the number of rows of the characteristic matrix, arranges the communication distances from short to long, takes the median of the radio wave propagation loss calculated according to formula (1) in S102 as the first column of the characteristic matrix, assigns the second column of the characteristic matrix to all 1 columns, and takes the number of communication links as the number of rows of the characteristic matrix, thereby obtaining the characteristic matrix .

[0064] It should be noted that when parameters such as the operating frequency and the transmit and receive antenna gains are fixed, the median values ​​of the radio wave propagation losses at different distances calculated by N and S102 are the same.

[0065] In S104, a target variable vector of a weighted linear regression model is constructed based on the training set. The training set data is used as the target variable vector , as the target of weighted linear regression model fitting.

[0066] In S105, the weight matrix of the weighted linear regression model is constructed , , is a diagonal matrix, (i=1,...,N) represents the weight of the i-th BLOSCO link, i.e., the number of preprocessed data segments arranged in ascending order of BLOSCO distance. To improve the fitting accuracy of the weighted linear regression model, this step assigns the BLOSCO link weights equal to the number of time periods into which the link is divided. Therefore, a greater number of time periods (i.e., measured data segments) results in a greater weight for the BLOSCO link in the fitting; a smaller number of time periods results in a smaller weight for the BLOSCO link in the fitting.

[0067] In S106, according to the characteristic matrix of the weighted linear regression model and the target variable vector A weighted linear least squares model is built as follows:

[0068] (2)

[0069] in, and They represent the slope correction coefficient and the horizontal offset correction term respectively, and are the regression parameters to be solved. Represents the feature matrix The first column of .

[0070] In S107, the regression parameters in the weighted linear least squares model are solved, that is, and .

[0071] (3)

[0072] in, , To include regression parameters and Column vector of , , , for The transposed matrix of .

[0073] In S108, the regression parameters obtained are used to correct the model to be corrected, and the following correction formula is obtained:

[0074] (4)

[0075] in, is the median value of the corrected radio wave propagation loss at different beyond-horizon communication distances at sea, that is, the corrected X-band tropospheric scatter prediction model.

[0076] After obtaining the revised X-band tropospheric scatter prediction model, the median radio wave propagation loss is calculated using the revised X-band tropospheric scatter prediction model, and the calculated results are compared with the data of the test set to evaluate the accuracy and applicability of the revised X-band tropospheric scatter prediction model. The median radio wave propagation loss obtained using formula (4) and formula (1) is compared with the measured path loss data to verify the accuracy and applicability of the revised model. Please refer to Figure 2 , Figure 2 Schematic diagram of using the modified backscattering model of this application to predict path loss at different communication distances. Figure 2 It can be seen that the path loss predicted by the modified scattering model of this application is closer to the measured value.

[0077] The accuracy and applicability of the median value of the radio wave propagation loss calculated by the revised model obtained in this application are described below through specific test examples.

[0078] The X-band scattering measured path loss data obtained from June 16 to July 11, 2024 on three cross-sea beyond-line-of-sight communication test links, namely, the 150km link between Jizhao Bay, Zhanjiang, Guangdong and Wenchang, Hainan, the 188km link between Yangxi, Guangdong and Wenchang, Hainan, and the 248km link between Yangjiang, Guangdong and Wenchang, Hainan, were selected to verify the accuracy and applicability of the revised tropospheric scattering prediction model.

[0079] The 150km cross-sea BLOSCO communication test link generated 946,278 valid scattering data points, the 188km cross-sea BLOSCO communication test link generated 945,164 valid scattering data points, and the 248km cross-sea BLOSCO communication test link generated 407,307 valid scattering data points. The total valid data points for these three test links totaled approximately 2.3 million. The training and test sets were split 7:3, resulting in approximately 1.61 million valid data points for the training set and 690,000 for the test set.

[0080] The operating frequency and communication distance of the maritime over-the-horizon communication system are input into the corrected and uncorrected tropospheric scatter prediction models respectively, and the tropospheric scatter path loss prediction results at different distances are calculated. Figure 3-Figure 10 Comparisons of the predicted and uncorrected troposcatter model results with the measured path loss for 150 km, 188 km, and 248 km BVR communication links are shown. It can be seen that at different X-band distances, the predicted results of the corrected troposcatter model are closer to the measured path loss than those of the uncorrected model.

[0081] Calculate the absolute error between the predicted results of the corrected tropospheric scatter model and the average of the measured path losses. Figure 11-13 It can be seen that on 150km, 188km, and 248km beyond-horizon communication links, the absolute error between the predicted results of the corrected troposcatter model and the measured path loss average is less than 5dB in 53.46%, 96%, and 100% of the cases, respectively. The absolute error between the predicted results of the uncorrected troposcatter model and the measured path loss average is less than 5dB in 29.16%, 82.86%, and 96.84% of the cases, respectively. Therefore, the absolute error between the predicted results of the corrected troposcatter model and the measured path loss average is smaller, verifying the accuracy and effectiveness of the proposed correction method.

[0082] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which are protected by the present invention.

Claims

1. A method for correcting an X-band tropospheric scatter prediction model based on measured data, characterized in that: include: Obtain X-band tropospheric scatter path loss data at different beyond-horizon communication distances at sea, preprocess the data, and divide the preprocessed data into training and test sets; The median value of radio wave propagation loss at different beyond-horizon communication distances at sea is calculated using the model to be corrected. constructing a characteristic matrix of a weighted linear regression model according to the number of communication links and the median value of the radio wave propagation loss; Constructing a target variable vector of a weighted linear regression model based on the training set; Construct the weight matrix of the weighted linear regression model; A weighted linear least squares model is established based on the characteristic matrix and target variable vector of the weighted linear regression model; Calculating the regression parameters of the weighted linear least squares model according to the characteristic matrix, the weight matrix and the weighted linear least squares model of the weighted linear regression model; The model to be corrected is corrected according to the regression parameters to obtain a corrected X-band tropospheric scatter prediction model.

2. The method for correcting the X-band tropospheric scatter prediction model based on measured data according to claim 1, characterized in that: The correction method further includes: The revised X-band tropospheric scatter prediction model is used to calculate the revised median of the X-band tropospheric scatter radio wave propagation loss, and the calculated results are compared with the data of the test set to evaluate the accuracy and applicability of the revised X-band tropospheric scatter prediction model.

3. The method for correcting the X-band tropospheric scatter prediction model based on measured data according to claim 1, characterized in that: The process of preprocessing data includes: The measured path loss data is segmented at preset time intervals, and the average value of the measured path loss data in each time period is calculated, and the average value is used as the data of the training set and the test set.

4. The method for correcting the X-band tropospheric scatter prediction model based on measured data according to claim 3, characterized in that: The average value is randomly divided into a training set and a test set in a ratio of 7:

3.

5. The method for correcting the X-band tropospheric scatter prediction model based on measured data according to claim 3, characterized in that: The weight value of each communication link in the weight matrix is ​​the number of time periods divided by the communication link.

6. The method for correcting the X-band tropospheric scatter prediction model based on measured data according to claim 1, characterized in that: The step of constructing a characteristic matrix of a weighted linear regression model according to the number of communication links and the median of the radio wave propagation loss includes: The beyond-line-of-sight communication distances are arranged from short to long, the corresponding calculated median of the radio wave propagation loss is used as the first column of the feature matrix, the second column of the feature matrix is ​​assigned to all 1 columns, and the number of the communication links is used as the number of rows of the feature matrix.

7. The method for correcting the X-band tropospheric scatter prediction model based on measured data according to claim 1, characterized in that: The expression for calculating the median radio wave propagation loss using the modified X-band troposcatter prediction model is as follows: in, is the median value of radio wave propagation loss calculated using the modified X-band troposcatter prediction model; and are the slope correction coefficient and the horizontal offset correction term, respectively, and are regression parameters in the weighted linear least squares model; is the median value of the radio wave propagation loss calculated using the model to be corrected.

Images