A GNSS spoofing source wide-area positioning method based on AIS data

Abstract

This invention relates to a wide-area localization method for GNSS deception sources based on AIS data, comprising: S1, setting a monitoring area and monitoring time window; S2, acquiring AIS data; S3, filtering abnormal data; S4, clustering the abnormal dataset to generate labeled data; S5, setting a minimum sample size, determining whether the number of data in the data group exceeds the minimum sample size, if so, proceeding to S6; otherwise, updating the monitoring area and monitoring time window, and returning to S2; S6, using a random sampling consensus method to obtain the weight distribution of the deception source interference situation; S7, analyzing the coordinates of deception sources corresponding to the same label, and analyzing the interference range and interference time; S8, outputting the coordinates, interference range, and interference time of the deception source, then updating the monitoring area and monitoring time window, and returning to S2. Compared with the prior art, this invention has advantages such as improving the accuracy of ship deception source localization.

Description

Technical Field

[0001] This invention relates to the technical field of GNSS deception source localization, and in particular to a wide-area localization method for GNSS deception sources based on AIS data. Background Technology

[0002] Global Navigation Satellite System (GNSS) plays a vital role in modern society. It provides high-precision positioning, navigation, and timing information, and is widely used in aviation, maritime, transportation, communication, and military fields, holding immense significance for socio-economic development and national security. However, due to the weak signal strength reaching the ground and the completely exposed signal structure, facilities and activities relying on GNSS are vulnerable to threats and challenges from GNSS interference sources. GNSS plays a crucial role in maritime navigation, but it also faces threats from deception sources. These illegal GNSS deception sources can easily manipulate a ship's position, speed, and time information by sending false GNSS signals, thereby guiding it onto predetermined routes, especially in vast ocean areas. Such malicious deception poses a serious threat to navigation in international waters and passage through territorial waters.

[0003] Current GNSS spoofing source localization methods fall into several categories, primarily ranging and non-ranging techniques. Rating-based localization includes methods based on interference signal strength (RJSS), time of arrival (TOA), and time difference of arrival (TDOA). However, these algorithms rely heavily on high-precision measurement equipment and have stringent time synchronization requirements, hindering the expansion of large-scale distributed monitoring networks. In contrast, non-ranging techniques employ various methods, including weighted centroid localization algorithms, virtual force iteration algorithms, geometric coverage algorithms, and algorithms based on neighbor point changes. Each method has its advantages and limitations, particularly in marine environments where the lack of synchronized, high-precision observation data restricts the performance of ranging-based localization techniques. Furthermore, when using non-ranging techniques, if a vessel enters the interference range of a spoofing source, the transmitted AIS data may be forged and invalid, leading to uneven data distribution, noise, and bias. Summary of the Invention

[0004] The purpose of this invention is to provide a wide-area positioning method for GNSS deception sources based on AIS data to improve the accuracy of ship deception source positioning.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] A wide-area localization method for GNSS spoofing sources based on AIS data, the method includes:

[0007] S1. Set the monitoring area and monitoring time window;

[0008] S2. Filter out AIS data under the monitoring area and monitoring time window set in S1, and preprocess the AIS data to obtain preprocessed AIS data;

[0009] S3. Set the update rules for AIS data, and filter out abnormal data from the preprocessed AIS data according to the update rules to form an abnormal dataset;

[0010] S4. Use DBSCAN to cluster the abnormal dataset, remove noisy data points, and generate multiple sets of label data, where each set of label data includes the coordinate points corresponding to one or more AIS data.

[0011] S5. Set the minimum sample size for deception source location, and determine whether the number of data in each data group with the same label exceeds the minimum sample size. If so, execute S6.

[0012] Conversely, update the monitoring area and monitoring time window, and return to S2;

[0013] S6. For data with the same label, a random sampling consensus method is used to obtain the weight distribution of the deception source interference situation corresponding to the data with the same label. The weight distribution represents the weight of each unit of the coordinate plane divided at a certain interval. Data with the same label are distributed in the coordinate plane.

[0014] S7. Based on the weight distribution, the weighted centroid estimation method is used to analyze the coordinates of the deception source corresponding to the same label, and the interference range and interference time of the deception source are analyzed.

[0015] S8. Output the coordinates of the deception source, the interference range, and the interference time. Then update the monitoring area and monitoring time window, and return to S2.

[0016] Furthermore, the specific steps of S3 are as follows:

[0017] Set the update rules for AIS data. Each AIS data entry for each ship is [time,lat,lon,speed], where time represents the timestamp, lat represents the latitude, lon represents the longitude, and speed represents the speed. The update rules include latitude rules, longitude rules, and speed rules. For each AIS data entry, AIS data that satisfies any one of the three rules is marked as true. Assuming that the i-th AIS data entry is marked as true, if among m consecutive data entries including the i-th AIS data entry, only the i-th AIS data entry is marked as true, then the i-th AIS data entry is considered abnormal data.

[0018] Furthermore, for the i-th AIS data entry marked as true, it satisfies any one of the latitude rule, longitude rule, and speed rule, where,

[0019] The latitude rule is: in time-series AIS data, the rate of change of the latitude of this data point compared to the previous or next data point exceeds the latitude rule boundary;

[0020] The longitude rule is: in time-series AIS data, the rate of change of longitude of this data point compared to the previous or next data point exceeds the boundary of the longitude rule;

[0021] The speed rule is: in time-series AIS data, the rate of change of speed of this data point compared to the previous or next data point exceeds the speed rule boundary.

[0022] Furthermore, the specific steps of S4 are as follows:

[0023] The longitude and latitude of each AIS data point in the abnormal dataset are converted into a set of eastward and northward coordinates. Each set of eastward and northward coordinates corresponds to a coordinate point. A neighborhood distance threshold and a minimum number of sample points in the neighborhood are set for the core object of DBSCAN clustering. DBSCAN clustering is then performed on each set of eastward and northward coordinates. Labels are assigned to each generated cluster, resulting in multiple sets of labeled data. Noisy clusters are assigned invalid labels, and noisy data points are removed.

[0024] Each set of label data includes coordinates corresponding to one or more AIS data. The coordinates corresponding to one AIS data represent a coordinate point corresponding to a set of eastward and northward coordinates obtained by converting the longitude and latitude of one AIS data.

[0025] Furthermore, the specific steps of S6 are as follows:

[0026] A1. For data with the same label, establish a two-dimensional northeast coordinate system plane containing all data with that label. The plane is divided into units at a certain interval. The initial weight of each unit is set to 0. Data with the same label, that is, the coordinate points of the same label, are distributed in the coordinate system plane.

[0027] A2. Extract any three coordinate points that are not on the same straight line, calculate the common center of the three coordinate points, use the common center as the interference candidate point, and update the weight of the unit where the interference candidate point is located.

[0028] A3. Repeat A2 until all cases of three coordinate points not on the same straight line are traversed, to obtain the weight distribution of the deception source interference situation. The weight distribution represents the weight of each unit obtained after traversing all cases of three coordinate points not on the same straight line.

[0029] Furthermore, the specific steps for calculating the common center of the three coordinate points are as follows:

[0030] Let the three coordinate points be i1 = (x1, y1), i2 = (x2, y2), and i3 = (x3, y3). Calculate the perpendicular bisectors of i1 and i2, as well as i1 and i3. Find the intersection of the two perpendicular bisectors, which is the center of the common circle.

[0031] Furthermore, the specific steps of S7 are as follows:

[0032] Obtain all interference candidate points for the same label in S6. Perform weighted centroid estimation based on the weights of the units where all interference candidate points are located in the weight distribution. Obtain the coordinates of the deception source under the label, and obtain the interference range and interference time of the deception source. Then update the monitoring area and monitoring time window, and return to S2.

[0033] Furthermore, the coordinates of the deception source are represented by a two-dimensional northeast coordinate system. The x-coordinate of the deception source in the two-dimensional northeast coordinate system plane is specifically: the sum of the products of the x-axis coordinates of all interference candidate points and the weight of the cell in which the interference candidate point is located, divided by the total weight, where the total weight represents the sum of the weights of the cells in which all interference candidate points are located.

[0034] The y-coordinate of the deception source in the two-dimensional northeast coordinate system plane is specifically: the sum of the products of the y-axis coordinates of all interference candidate points and the weight of the cell in which the interference candidate point is located, divided by the total weight.

[0035] Furthermore, the specific steps to obtain the interference range and interference time of the deception source are as follows:

[0036] Calculate the distance between the coordinate point corresponding to each AIS data in the abnormal dataset and the deception source. Sort the coordinate points corresponding to each AIS data in the abnormal dataset in ascending order according to the distance. The largest distance between the first q*100% of all coordinate points and the deception source is taken as the radius of the interference range. The interference range is determined based on the radius of the interference range. The overall time interval corresponding to the first q*100% of all coordinate points after sorting is the interference time, where q is the interference boundary threshold.

[0037] Furthermore, the overall time interval is specifically defined as the time interval between the minimum and maximum timestamps in the AIS data corresponding to all coordinate points in the first q*100% of the sorted data.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] This invention combines DBSCAN clustering, random sampling consensus, and weighted centroid estimation to locate deception sources. First, clustering is performed to identify anomalous data belonging to the same interference source (tag). Since anomalous data typically exhibits a circular or arc-shaped distribution, the possible interference sources for anomalous data belonging to the same tag are analyzed by finding the common center of these circles. Then, the cells containing the interference source are weighted; that is, the cell more likely to be a deception source has a higher weight. Finally, weighted centroid estimation is performed on the weighted candidate points to obtain the deception source information. This invention, by performing random sampling consensus analysis on clustered data belonging to the same deception source, can accurately identify the locations of multiple interference sources and correctly identify the interference source candidates. Furthermore, the method of analyzing the weights of the cells containing the interference source effectively analyzes the interference source and can effectively handle problems such as uneven distribution of anomalous points and data dispersion caused by independent GNSS receivers, thus improving the accuracy and stability of the positioning. Attached Figure Description

[0040] Figure 1 This is a flowchart of the present invention;

[0041] Figure 2 The location distribution of abnormal data and interference candidate points in this invention;

[0042] Figure 3 This is a three-dimensional diagram showing the weighted distribution of the deception source interference situation according to the present invention.

[0043] Figure 4 This is a two-dimensional diagram showing the weighted distribution of the deception source interference situation according to the present invention. Detailed Implementation

[0044] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. These embodiments are based on the technical solution of the present invention and provide detailed implementation methods and specific operating procedures. However, the scope of protection of the present invention is not limited to the following embodiments.

[0045] This invention provides a wide-area localization method for GNSS spoofing sources based on AIS (Automatic Identification System) data. Ship AIS data contains key information processed by the GNSS receiver, such as ship position (latitude and longitude), heading, and speed, among other dynamic time-series data. When a ship enters or leaves the interference range of a GNSS spoofing source, its AIS data undergoes abnormal changes. This invention utilizes AIS information from multiple ships to detect the timing and location of these abnormal changes. Considering that the GNSS spoofing source uses an omnidirectional antenna, these anomaly points are randomly distributed along a circular boundary centered on the spoofing source. First, a density-based spatial clustering of applications with noise (DBSCAN) method is used to group the anomaly points affected by different spoofing sources. Then, a random sample consensus algorithm (RANSAC) is used to estimate the interference field strength of the spoofing source. Finally, an improved centroid localization algorithm is used to estimate the location of the spoofing source and the interference range. This innovative method can provide strong support for the timely identification and handling of GNSS spoofing sources, thereby contributing to maintaining the safety and normal operation of maritime navigation. This invention provides a wide-area localization method for GNSS spoofing sources based on AIS data. The flowchart of the method is as follows: Figure 1 As shown. The method includes the following steps:

[0046] S1. Set the monitoring area and monitoring time window;

[0047] S2. Filter out AIS data under the monitoring area and monitoring time window set in S1, and preprocess the AIS data to obtain preprocessed AIS data;

[0048] S3. Set the update rules for AIS data, and filter out abnormal data from the preprocessed AIS data according to the update rules to form an abnormal dataset;

[0049] S4. Use DBSCAN to cluster the abnormal dataset, remove noisy data points, and generate multiple sets of label data, where each set of label data includes the coordinate points corresponding to one or more AIS data.

[0050] S5. Set the minimum sample size for deception source location, and determine whether the number of data in each data group with the same label exceeds the minimum sample size. If so, execute S6.

[0051] Conversely, update the monitoring area and monitoring time window, and return to S2;

[0052] S6. For data with the same label, a random sampling consensus method is used to obtain the weight distribution of the deception source interference situation corresponding to the data with the same label. The weight distribution represents the weight of each unit of the coordinate plane divided at a certain interval. Data with the same label are distributed in the coordinate plane.

[0053] S7. Based on the weight distribution, the weighted centroid estimation method is used to analyze the coordinates of the deception source corresponding to the same label, and the interference range and interference time of the deception source are analyzed.

[0054] S8. Output the coordinates of the deception source, the interference range, and the interference time. Then update the monitoring area and monitoring time window, and return to S2.

[0055] In S1, the monitoring area represents a region on the sea surface, and the monitoring time window represents a period of time.

[0056] In S2, the specific steps for preprocessing AIS data are cleaning, interpolation, and grouping.

[0057] In S3, the update rules are set as follows: latitude rule, longitude rule, and speed rule, which represent the upper limit of the update interval, the range of latitude and longitude changes per unit time, and the upper limit of speed-related acceleration. Abnormal data is filtered based on the update rules.

[0058] In S3, because GNSS spoofing sources can cause ship positioning within the interference area to jump from one area to another, the AIS time-series data will show dynamic information such as longitude, latitude, speed, and heading changing in a "step-like" pattern over time. This invention refers to the data at these jump points as anomalies. A schematic diagram of the location distribution of anomaly data and interference candidate points is shown below. Figure 2 As shown.

[0059] AIS data consists of time-series multivariate data for each ship. n Its composition and characteristics mainly consist of timestamps, latitude, longitude, and speed. i ,lat i ,lon i speed i The formula is calculated using four variables, where N is the total number of ships in the AIS data, and M is the number of dynamic data points for the nth ship. (RULE) lat ,RULE lon ,RULE speed For setting update rule boundaries, anomaly iFor data i, an anomaly indicator is set to TRUE if any one of the three rule conditions is met. If, among m consecutive data points including the i-th AIS data point, only the i-th AIS data point is marked as true, then the i-th AIS data point is considered an anomaly. m can be 20.

[0060] The AIS data is as follows:

[0061] AIS = {Ship1, Ship2, ..., Ship} N}

[0062] Ship n ={[time1,lat1,lon1,speed1],……,[time M ,lat M ,lon M speed M ]}

[0063] ={[time1,N1,E1,speed1],……,[time M N M E M speed M ]}

[0064] ={[time1,numN1,numE1,speed1],……,[time M ,numN M ,numE M speed M ]}

[0065] The abnormal indicators of i data are:

[0066]

[0067] In other words, for the i-th AIS data item marked as true, it satisfies the latitudinal rule (lat i -lat i-1 ) / (time i -time i-1 )>RULE lat Longitude rules (lon) i -lon i-1 ) / (time i -time i-1 )>RULE lon and speed rules (speed) i -speed i-1 ) / (time i -time i-1 )>RULEspeed Any one of the rules.

[0068] In S4, DBSCAN is used to complete the grouping labels of the abnormal dataset.

[0069] The monitoring center was set as the anchor point. GPS latitude and longitude coordinates were converted from WGS84 to ENU (East-West-North) coordinate system. The longitude and latitude of each AIS data point in the abnormal dataset were transformed into a set of east and north coordinates. Each set of east and north coordinates determined a corresponding coordinate point. A neighborhood distance threshold and a minimum number of samples in the neighborhood were set for the core object of DBSCAN clustering. DBSCAN clustering was performed on each set of east and north coordinates. Labels were assigned to each generated cluster, resulting in multiple sets of labeled data. Noisy clusters were assigned invalid labels, and noisy data points were removed.

[0070] Each set of label data includes coordinates corresponding to one or more AIS data. The coordinates corresponding to one AIS data represent a coordinate point corresponding to a set of eastward and northward coordinates obtained by converting the longitude and latitude of one AIS data.

[0071] The east and north coordinates of outlier points in the outlier data that need to be clustered may exhibit the same outlier characteristics due to factors such as AIS equipment status and GNSS signal quality, even if the data does not involve GNSS deception sources. Such noisy data needs to be filtered out. Furthermore, if outlier data points from different deception sources do not overlap, they also need to be grouped and labeled.

[0072] DBSCAN is a density-based clustering algorithm that determines whether samples belong to the same category based on the density of their distribution. DBSCAN is based on the concept of neighborhood, using parameters (∈, MinPts) to describe the density of the sample distribution. Here, ∈ represents the neighborhood distance threshold for the core object, and MinPts represents the minimum number of samples in the neighborhood at distance ∈. With a feature dimension of 2, and using the k-distance distribution to select ∈, MinPts is set to 4. Since the sample data uses east-north coordinates, Euclidean distance can be directly used for distance calculation. After clustering, the generated clusters are sequentially grouped and labeled, and noise points are assigned invalid labels.

[0073] In steps S5 and S6, candidate data storage and RANSAC-based candidate location weighting are performed. A threshold n is set for the number of grouped samples for deception source location. If the number is less than n, it may be due to insufficient monitoring time or too few vessels being interfered with by deception sources, requiring further extension of the monitoring time to increase the number of effective samples. Therefore, this invention stores this portion of data as cached data and loads it into step S4 for re-clustering in the next time period. If the number reaches n, this invention uses RANSAC to obtain a series of candidate deception source locations and iteratively accumulates the weights.

[0074] In step S6, the RANSAC algorithm is used to estimate the center position of the scattered points through random sampling and iteration. The two-dimensional coordinate system is then rasterized, and the weights of the raster cells with estimated centers are iteratively accumulated to obtain the weight distribution.

[0075] The specific steps of S6 are as follows:

[0076] A1. For data with the same label, establish a two-dimensional northeast coordinate system plane containing all data with that label. The plane is divided into units at a certain interval. The initial weight of each unit is set to 0. Data with the same label, that is, the coordinate points of the same label, are distributed in the coordinate system plane.

[0077] A2. Extract any three coordinate points that are not on the same straight line, calculate the common center of the three coordinate points, use the common center as the interference candidate point, and update the weight of the unit where the interference candidate point is located.

[0078] A3. Repeat A2 until all cases of three coordinate points not on the same straight line are traversed, to obtain the weight distribution of the deception source interference situation. The weight distribution represents the weight of each unit obtained after traversing all cases of three coordinate points not on the same straight line.

[0079] The basic idea of ​​the RANSAC method is to estimate the parameters of the mathematical model that the normal observation data follows by randomly sampling and iteratively based on a set of observation datasets containing outliers. The specific implementation of the RANSAC method of this invention is described below.

[0080] The rasterized data divides the two-dimensional northeast-oriented coordinate plane containing the target into a series of tiny planar resolution cells at intervals of Δl. Δl is a technical indicator of the coarse localization accuracy of the interference source, aiming to distribute outliers within different cells as much as possible. Each planar resolution cell has a weight accumulator with an initial value of 0.

[0081] Three data points that are not on the same straight line are randomly selected from the set of outlier data points, denoted as i1 = (x1, y1), i2 = (x2, y2), and i3 = (x3, y3). Assuming that all three outlier data points are affected by the same source of interference, the coordinates of i1, i2, and i3 can be used to determine the center of a circle where the three points are concyclic.

[0082] Calculate the slope of the perpendicular bisector:

[0083]

[0084]

[0085]

[0086]

[0087] Calculate the equation of the perpendicular bisector:

[0088] perpendicular bisectors of i1 & i2: yy m1 =k1·(xx) m1 )

[0089] perpendicular line between i1 and i3: yy m2 =k2·(xx) m2 )

[0090] Find the intersection of the perpendicular bisectors to obtain the coordinates of the center of the circle:

[0091]

[0092]

[0093] obtained All of these will be considered as candidate points for locating interference sources, and the results will be updated. The weight accumulator of the plane resolution cell is updated to the settlement result. The information of the anomalies i1, i2, and i3.

[0094]

[0095] Q means The weight of the unit.

[0096] The three-dimensional diagram of the weight distribution of the deception source interference situation is as follows: Figure 3 As shown, the two-dimensional weight distribution diagram of the deception source interference situation is as follows: Figure 4 As shown.

[0097] Because GNSS receivers on ships operate independently and their navigation routes are unrelated, they are affected differently by deception sources. Theoretically, all anomalous data are unlikely to lie on the same circular boundary, but are more likely to exhibit a ring or arc-shaped distribution. If anomalous data from a different deception source is incorrectly associated, the generated false candidate points will deviate from the true deception source location, while the correct candidate points will cause the weight of the deception source's location area to show a spectral peak, thus obtaining a weight distribution that approximates the deception source's interference situation.

[0098] In steps S7 and S8, the location of the deception source is estimated based on weighted centroid estimation, and information analysis such as the location of the deception source, interference range, and interference time is completed. The location of the deception source is obtained by weighted centroid estimation of the weighted candidate points. An interference boundary threshold q (q>90%) is set, the distance between the abnormal sample points and the estimated location of the deception source is calculated and sorted in ascending order, the maximum distance of the first q*100% of sample points is the interference range radius, and the overall time interval of the first q*100% of sample points is the interference time period. The coordinates of the deception source are represented by a two-dimensional northeast coordinate system. The x-coordinate of the deception source in the two-dimensional northeast coordinate system plane is specifically: the sum of the products of the x-axis coordinates of all interference candidate points and the weights of the units in which the interference candidate point is located, divided by the total weights, where the total weights represent the sum of the weights of the units in which all interference candidate points are located.

[0099] The y-coordinate of the deception source in the two-dimensional northeast coordinate system plane is specifically: the sum of the products of the y-axis coordinates of all interference candidate points and the weight of the cell in which the interference candidate point is located, divided by the total weight.

[0100] The sum of the products of the x-axis coordinates of all interference candidate points and the weight of the cell in which the interference candidate point is located represents the product of the coordinates of a candidate point i and the weight of i in the cell. Then, the products of all candidate points are added together.

[0101] The coordinates of the deception source are:

[0102]

[0103] Where, x i and y i W represents the coordinates of the i-th interference candidate point in the two-dimensional northeast coordinate system plane. i X represents the weight of the cell containing the i-th interference candidate point, N represents the total number of interference candidate points, and X represents the weight of the cell containing the i-th interference candidate point. spoof and Y spoof These represent the coordinates of the deception source in a two-dimensional northeast-oriented coordinate system plane.

[0104] The overall time interval is specifically defined as the time interval between the minimum and maximum timestamps in the AIS data corresponding to all coordinate points in the first q*100% of the sorted data.

[0105] This invention utilizes spatiotemporal big data from AIS data received via GNSS to detect GNSS spoofing sources. When interfered with by GNSS spoofing sources, the time-series information such as latitude, longitude, and speed in a ship's AIS data will exhibit unreasonable jumps. When multiple independent spoofing sources exist and their interference ranges do not overlap, multiple anomalous data clusters will appear on the two-dimensional coordinate system, along with outliers due to positioning failures. This invention first filters out anomalous datasets through statistical data analysis, then uses a density-based spatial clustering method (DBSCAN) to group anomalous points interfered with by different spoofing sources. Next, the Random Sample Consensus (ANSAC) algorithm is used to estimate the candidate locations of spoofing sources and assign probability weights. Finally, a weighted centroid positioning algorithm is used to estimate the location of the spoofing sources and the interference range. In summary, this innovative method provides a new approach to locating GNSS spoofing sources, overcoming some limitations and problems of existing technologies. Especially in marine environments, it can more effectively detect and locate GNSS spoofing sources, potentially improving the security and reliability of GNSS systems.

[0106] The advantages of this invention are:

[0107] 1. No longer dependent on high-precision measurement equipment and time synchronization: Traditional deception source location technology requires high-precision measurement equipment and time synchronization, while this invention overcomes this limitation and no longer requires these conditions, thus making it more flexible and easier to implement.

[0108] 2. Applicable to large-scale distributed monitoring networks: The method of this invention is applicable to large-scale distributed monitoring networks and has excellent scalability. This means that it can cover a wide geographical area and monitor multiple potential sources of deception, thereby enhancing the security of the system.

[0109] 3. DBSCAN Preprocessing of Data Grouping: By introducing DBSCAN as a preprocessing method, this invention effectively solves the problem of locating multiple independent, non-interfering deception sources. Thus, even in the presence of multiple interference sources, their locations can be accurately identified.

[0110] 4. RANSAC method for candidate spoofing source localization and weighting: Compared with traditional direct centroid estimation, this invention uses the RANSAC method, which can effectively handle problems such as uneven distribution of outliers and data dispersion caused by independence between GNSS receivers. This improves the accuracy and stability of localization.

[0111] 5. Provides higher accuracy positioning results and additional information: The weighted centroid estimation algorithm of this invention not only provides higher accuracy positioning results for deception sources, but also outputs additional information such as interference range and interference time, which is very helpful for monitoring and eliminating deception sources. This additional information can help operators better understand and respond to interference.

[0112] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.

Claims

1. A wide-area positioning method for GNSS spoofing sources based on AIS data, characterized in that, The methods include: S1. Set the monitoring area and monitoring time window; S2. Filter out AIS data under the monitoring area and monitoring time window set in S1, and preprocess the AIS data to obtain preprocessed AIS data; S3. Set the update rules for AIS data, and filter out abnormal data from the preprocessed AIS data according to the update rules to form an abnormal dataset; S4. Use DBSCAN to cluster the abnormal dataset, remove noisy data points, and generate multiple sets of label data, where each set of label data includes the coordinate points corresponding to one or more AIS data. S5. Set the minimum sample size for deception source location, and determine whether the number of data in each data group with the same label exceeds the minimum sample size. If so, execute S6. Conversely, update the monitoring area and monitoring time window, and return to S2; S6. For data with the same label, a random sampling consensus method is used to obtain the weight distribution of the deception source interference situation corresponding to the data with the same label. The weight distribution represents the weight of each unit of the coordinate plane divided at a certain interval. Data with the same label are distributed in the coordinate plane. S7. Based on the weight distribution, the weighted centroid estimation method is used to analyze the coordinates of the deception source corresponding to the same label, and the interference range and interference time of the deception source are analyzed. S8. Output the coordinates of the deception source, the interference range, and the interference time. Then update the monitoring area and monitoring time window, and return to S2.

2. The GNSS spoofing source wide-area positioning method based on AIS data according to claim 1, characterized in that, The specific steps for S3 are as follows: Set the update rules for AIS data. Each AIS data entry for each ship is [time,lat,lon,speed], where time represents the timestamp, lat represents the latitude, lon represents the longitude, and speed represents the speed. The update rules include latitude rules, longitude rules, and speed rules. For each AIS data entry, AIS data that satisfies any one of the three rules is marked as true. Assuming that the i-th AIS data entry is marked as true, if among m consecutive data entries including the i-th AIS data entry, only the i-th AIS data entry is marked as true, then the i-th AIS data entry is considered abnormal data.

3. The GNSS spoofing source wide-area positioning method based on AIS data according to claim 2, characterized in that, For the i-th AIS data entry marked as true, it satisfies any one of the following rules: latitude rule, longitude rule, and speed rule. The latitude rule is: in time-series AIS data, the rate of change of the latitude of this data point compared to the previous or next data point exceeds the latitude rule boundary; The longitude rule is: in time-series AIS data, the rate of change of longitude of this data point compared to the previous or next data point exceeds the boundary of the longitude rule; The speed rule is: in time-series AIS data, the rate of change of speed of this data point compared to the previous or next data point exceeds the speed rule boundary.

4. The GNSS spoofing source wide-area positioning method based on AIS data according to claim 2, characterized in that, The specific steps of S4 are as follows: The longitude and latitude of each AIS data point in the abnormal dataset are converted into a set of eastward and northward coordinates. Each set of eastward and northward coordinates corresponds to a coordinate point. A neighborhood distance threshold and a minimum number of sample points in the neighborhood are set for the core object of DBSCAN clustering. DBSCAN clustering is then performed on each set of eastward and northward coordinates. Labels are assigned to each generated cluster, resulting in multiple sets of labeled data. Noisy clusters are assigned invalid labels, and noisy data points are removed. Each set of label data includes coordinates corresponding to one or more AIS data. The coordinates corresponding to one AIS data represent a coordinate point corresponding to a set of eastward and northward coordinates obtained by converting the longitude and latitude of one AIS data.

5. A GNSS spoofing source wide-area positioning method based on AIS data according to claim 4, characterized in that, The specific steps of S6 are as follows: A1. For data with the same label, establish a two-dimensional northeast coordinate system plane containing all data with that label. The plane is divided into units at a certain interval. The initial weight of each unit is set to 0. Data with the same label, that is, the coordinate points of the same label, are distributed in the coordinate system plane. A2. Extract any three coordinate points that are not on the same straight line, calculate the common center of the three coordinate points, use the common center as the interference candidate point, and update the weight of the unit where the interference candidate point is located. A3. Repeat A2 until all cases of three coordinate points not on the same straight line are traversed, to obtain the weight distribution of the deception source interference situation. The weight distribution represents the weight of each unit obtained after traversing all cases of three coordinate points not on the same straight line.

6. A GNSS spoofing source wide-area positioning method based on AIS data according to claim 5, characterized in that, The specific steps for calculating the common center of the three coordinate points are as follows: Let the three coordinate points be i1 = (x1, y1), i2 = (x2, y2), and i3 = (x3, y3). Calculate the perpendicular bisectors of i1 and i2, as well as i1 and i3. Find the intersection of the two perpendicular bisectors, which is the center of the common circle.

7. A GNSS spoofing source wide-area positioning method based on AIS data according to claim 5, characterized in that, The specific steps for S7 are as follows: Obtain all interference candidate points for the same label in S6. Perform weighted centroid estimation based on the weights of the units where all interference candidate points are located in the weight distribution. Obtain the coordinates of the deception source under the label, and obtain the interference range and interference time of the deception source. Then update the monitoring area and monitoring time window, and return to S2.

8. A GNSS spoofing source wide-area positioning method based on AIS data according to claim 7, characterized in that, The coordinates of the deception source are represented by a two-dimensional northeast coordinate system. The x-coordinate of the deception source in the two-dimensional northeast coordinate system plane is specifically: the sum of the products of the x-axis coordinates of all interference candidate points and the weight of the cell in which the interference candidate point is located, divided by the total weight, where the total weight represents the sum of the weights of the cells in which all interference candidate points are located. The y-coordinate of the deception source in the two-dimensional northeast coordinate system plane is specifically: the sum of the products of the y-axis coordinates of all interference candidate points and the weight of the cell in which the interference candidate point is located, divided by the total weight.

9. A GNSS spoofing source wide-area positioning method based on AIS data according to claim 8, characterized in that, The specific steps to obtain the interference range and interference time of the deception source are as follows: Calculate the distance between the coordinate point corresponding to each AIS data in the abnormal dataset and the deception source. Sort the coordinate points corresponding to each AIS data in the abnormal dataset in ascending order according to the distance. The largest distance between the first q*100% of all coordinate points and the deception source is taken as the radius of the interference range. The interference range is determined based on the radius of the interference range. The overall time interval corresponding to the first q*100% of all coordinate points after sorting is the interference time, where q is the interference boundary threshold.

10. A GNSS spoofing source wide-area positioning method based on AIS data according to claim 9, characterized in that, The overall time interval is specifically defined as the time interval between the minimum and maximum timestamps in the AIS data corresponding to all coordinate points in the first q*100% of the sorted data.

Images