Navigation decoy method and device for unmanned aerial vehicle countering

Abstract

The invention discloses a navigation decoy method and device for unmanned aerial vehicle countering, and the method comprises the steps: S1, obtaining the position information of a target unmanned aerial vehicle at a plurality of moments, and constructing a position prediction model; s2, estimating navigation signal intensity when the target unmanned aerial vehicle is located at a selected position to obtain predicted navigation signal intensity; s3, generating a navigation decoy signal according to the predicted navigation signal strength; and S4, when the target unmanned aerial vehicle arrives at a selected position, transmitting a navigation decoy signal to the target unmanned aerial vehicle. The method can improve the success rate of navigation decoy.

Description

Technical Field

[0001] The present invention relates to the technical field related to drone control, and more specifically, to a navigation deception method and device for drone countermeasures. Background Art

[0002] Drones are widely used in logistics, aerial photography, reconnaissance and other fields. However, due to the lack of effective control measures, the "illegal flying" of drones has posed a threat to normal production and life, such as affecting flight takeoffs and landings, stealing confidential information, etc. There are already some drone deception technical solutions in the existing technology, but the anti-deception technology of drones is also being upgraded. Once it finds that the navigation signal is abnormal, it may also take corresponding response strategies. Therefore, it is necessary to design a technical solution that can overcome the above defects to a certain extent. Summary of the invention

[0003] An object of the present invention is to provide a navigation deception method and device for countering unmanned aerial vehicles, which can improve the success rate of navigation deception.

[0004] In order to achieve these objectives and other advantages of the present invention, according to one aspect of the present invention, the present invention provides a navigation decoy method for countering drones, including: S1: obtaining position information of a target drone at multiple times and constructing a position prediction model; S2: estimating the navigation signal strength when the target drone is located at a selected position to obtain a predicted navigation signal strength; S3: generating a navigation decoy signal based on the predicted navigation signal strength; S4: transmitting a navigation decoy signal to the target drone when the target drone arrives at the selected position.

[0005] Furthermore, in S2, the location information, terrain parameters and weather parameters of the selected location are input into a pre-established signal strength prediction model to obtain the predicted navigation signal strength; wherein, the signal strength prediction model is obtained by training a decision tree model.

[0006] Furthermore, the strength of the navigation decoy signal is at least consistent with the predicted navigation signal strength.

[0007] Furthermore, there are multiple selected positions, and the multiple selected positions are determined by gradually fine-tuning the position prediction model.

[0008] Furthermore, when the target UAV reaches the first selected position, a navigation decoy signal is transmitted to the target UAV, and it is determined whether the decoy is successful. If the decoy is not successful, S2-S4 are repeated.

[0009] Furthermore, the method for determining whether the deception is successful includes: obtaining multiple position information of the target UAV between the first selected position and the second selected position, and determining whether it conforms to the position prediction model; if so, determining that the deception is not successful.

[0010] According to another aspect of the present invention, a navigation decoy device for countering drones is also provided, including: a construction module, used to obtain position information of a target drone at multiple times and construct a position prediction model; a prediction module, used to estimate the navigation signal strength when the target drone is located at a selected position, and obtain a predicted navigation signal strength; a generation module, used to generate a navigation decoy signal according to the predicted navigation signal strength; a decoy module, used to transmit a navigation decoy signal to the target drone when the target drone reaches the selected position.

[0011] Furthermore, the prediction module inputs the location information, terrain parameters and weather parameters of the selected location into a pre-established signal strength prediction model to obtain the predicted navigation signal strength; wherein the signal strength prediction model is obtained by training a decision tree model.

[0012] Furthermore, there are multiple selected positions, and the multiple selected positions are determined by gradually fine-tuning the position prediction model. According to another aspect of the present invention, there is also provided a navigation deception device for countering drones, comprising: a processor; a memory for storing executable instructions; wherein the processor is configured to execute the executable instructions to implement the navigation deception method for countering drones.

[0013] The present invention has at least the following beneficial effects: The present invention obtains the position information of the target UAV at multiple times, constructs a position prediction model, and estimates the predicted navigation signal strength when the target UAV is located at a selected position, and generates a navigation decoy signal according to the predicted navigation signal strength, so as to transmit the navigation decoy signal to the target UAV when the target UAV reaches the selected position; the present invention predicts the current position of the target UAV and the navigation signal strength thereof, so that the generated navigation decoy signal is less different from the real navigation signal, thereby reducing the possibility of being identified and improving the success rate of decoy.

[0014] Other advantages, objectives and features of the present invention will be embodied in part through the following description, and in part will be understood by those skilled in the art through study and practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A flowchart of an embodiment of the present application; Figure 2 This is a framework diagram of an embodiment of the present application. DETAILED DESCRIPTION

[0016] The present invention is further described in detail below in conjunction with the accompanying drawings so that those skilled in the art can implement the invention with reference to the description.

[0017] It should be understood that the terms such as "having", "including" and "comprising" used in the embodiments of the present application do not exclude the existence or addition of one or more other elements or their combinations. All directional indications (such as up, down, left, right, front, back...) in the embodiments of the present application are only used to explain the relative position relationship, movement, etc. between the components in a certain specific posture. If the specific posture changes, the directional indication will also change accordingly. When an element is referred to as "fixed on" or "set on" another element, it can be directly on the other element or there may be a centering element at the same time. When an element is referred to as "connecting" another element, it can be directly connected to another element or it can be indirectly connected to another element through a centering element. The description of "first", "second", etc. in the embodiments of the present application is only for descriptive purposes, and cannot be understood as indicating or implying its relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" can explicitly or implicitly include at least one of the features.

[0018] It should be noted that the technical solutions between the various embodiments of the present application can be combined with each other, but it must be based on the fact that ordinary technicians in the field can implement it. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by this application.

[0019] like Figure 1-2 As shown, an embodiment of the present application provides a navigation deception method for countering a drone, including: S1: Obtain the location information of the target UAV at multiple times and build a location prediction model; For example, a variety of monitoring technologies and equipment are used to obtain the location information of the target UAV at multiple different times, and the location information includes latitude information, longitude information and altitude information; these monitoring technologies may include but are not limited to radar monitoring systems, optical tracking equipment, acoustic tracking equipment and technical means based on wireless signal positioning, etc.; the radar monitoring system can transmit electromagnetic waves and receive the echo signal reflected by the target UAV, and accurately calculate the position parameters such as the distance, azimuth and pitch angle of the target UAV relative to the radar station at different times according to the time delay, Doppler frequency shift and other information of the echo; the optical tracking equipment uses high-resolution cameras and image recognition The identification algorithm is used to capture and analyze the image of the target drone in the air in real time, and its spatial position at different times is determined by processing the coordinates, size, and relative position relationship of the drone with known reference objects in the image. Based on the technology of wireless signal positioning, the communication signal and navigation signal transmitted by the target drone are received, and the approximate position of the drone at each time is calculated by using a variety of positioning methods such as signal strength, arrival time difference, and arrival angle. The sound signal of the target drone is obtained by using three or more microphones, and the position of the target drone can be estimated according to the time when the sound signal reaches each microphone and the position of the microphone itself. For example, the position prediction model is constructed by using an exponential smoothing model, which can quickly establish a position prediction model with high accuracy after the target drone is detected. The specific formula is as follows: L xi , L yi and L zi Respectively refers to t i The smoothed values ​​of latitude, longitude, and altitude at the moment, L xi-1 , L yi-1 and L zi-1 Respectively refers to t i-1 The smoothed values ​​of latitude, longitude, and altitude at the time, x i ,y i and z i They refer to the actual values ​​of latitude, longitude, and altitude at time ti, respectively. x , α y and α z They are the smoothing factors of latitude, longitude and altitude, with values ​​between 0 and 1, and are obtained by training using the location information collected at several moments. When calculating specifically, the smoothing values ​​of latitude, longitude and altitude are initialized to L x0 , L y0 and L z0, can be estimated using the position information collected at several moments, taking the first position information or the average of the first few position information, and then iterating using the above formula.

[0020] S2: estimating the navigation signal strength when the target UAV is located at a selected position to obtain a predicted navigation signal strength; the selected position is predicted according to a position prediction model, and also needs to be specifically determined in combination with a set UAV countermeasure range and a position of a navigation decoy device; Exemplarily, the navigation signal strengths of multiple locations are collected, and a regression model, a machine learning model, or a neural network model is established to predict the navigation signal strength of the selected location; S3: generating a navigation decoy signal according to the predicted navigation signal strength; Specifically, according to the navigation information coding rules followed by the navigation system of the target UAV, a signal coding content containing false position, speed, time and other navigation-related information is generated and embedded into the decoy signal waveform, so that after receiving the decoy signal, the UAV parses out the false navigation information and mistakenly thinks it is valid information from the real navigation satellite, thereby changing its own navigation state; For example, the strength of the navigation decoy signal should be consistent with the predicted navigation signal strength, or slightly higher than the predicted navigation signal strength (determined in consideration of the distance between the navigation decoy device and the drone, and the distance factor is ignored when the location is close), so as to avoid being considered abnormal by the target drone due to excessively high signal strength, or being rejected by the target drone due to excessively low signal strength; S4: When the target UAV reaches the selected location, transmitting a navigation decoy signal to the target UAV; In this step, the target drone is continuously tracked and its location information is collected. When it reaches the selected location, a navigation decoy signal is generated; It can be seen that this embodiment uses the position prediction model to accurately predict the position of the target UAV, and then accurately estimates the predicted navigation signal strength, so that the generated navigation decoy signal is less different from the real navigation signal, reducing the possibility of being identified and improving the success rate of decoy.

[0021] In another embodiment, in S2, the location information, terrain parameters and weather parameters of the selected location are input into a pre-established signal strength prediction model to obtain the predicted navigation signal strength; wherein the signal strength prediction model is obtained by training a decision tree model; In this embodiment, compared with some complex deep learning models, the training process of the decision tree model does not require large-scale iterative optimization, complex gradient calculation, etc., and is relatively simple and easy to understand. When processing tasks such as navigation signal strength prediction, which usually do not have a super large data scale, the training process can be completed relatively quickly to obtain a usable prediction model; once the decision tree model training is completed, when predicting the navigation signal strength, it is only necessary to determine which child node to go to along the branch structure of the tree according to the feature value of the input sample, and finally reach the leaf node to obtain the corresponding prediction value; this process has a relatively small amount of calculation, and can quickly give a prediction result. For navigation deception scenarios with high real-time requirements, the required signal strength information can be provided in a timely manner; Specifically, location information refers to the latitude information, longitude information and altitude information mentioned above. Terrain parameter data can be obtained through a high-precision geographic information system (GIS), including information such as terrain undulation, surface roughness, terrain slope and landform type. The location, height, density and other conditions of artificial buildings such as buildings, large bridges and towers, and natural tall trees in space can also be obtained through electronic maps. Weather parameter data includes various weather-related parameters such as atmospheric temperature, humidity, air pressure, wind speed, wind direction and precipitation conditions (different precipitation types such as rain, snow, fog and their intensity). Considering that weather conditions have a significant impact on electromagnetic signal propagation, for example, rainfall will absorb and scatter signals, resulting in signal strength attenuation; it is also necessary to obtain actual navigation signal strength measurement data. Under the above-mentioned various locations, terrains and weather conditions, professional signal strength detection equipment is used to measure the navigation signal strength on the spot and use it as a label for model training; Perform comprehensive data cleaning operations on all types of collected data to remove abnormal values ​​(such as obviously unreasonable signal strength values ​​caused by measurement equipment failure, or erroneous data points whose coordinates are beyond the reasonable range, etc.) and duplicate data records to ensure data accuracy and consistency and avoid these abnormal data from interfering with subsequent model training; Encode the collected data to make it suitable as input for the decision tree model. For example, there are five different types of landforms. After one-hot encoding, five binary feature columns will be generated, one for each landform type. This will clearly indicate which landform the location belongs to, making it easier for the model to process and analyze during training. In view of the large differences in the coordinate values ​​in the location information, the various quantitative indicators in the terrain parameters, and the numerical ranges in the weather parameters, in order to enable the decision tree model to treat each feature fairly during the training process and avoid the dominant influence of some feature values ​​being too large or too small on the model, these numerical variables are normalized or standardized; The CART (Classification and Regression Tree) algorithm is used to build a decision tree model. This algorithm can be used for both classification and regression tasks, and is very suitable for predicting continuous numerical target variables such as navigation signal strength. In the model initialization stage, some key hyperparameters need to be set, such as the maximum depth of the tree (a moderate value is initially set, such as 10, which limits the growth of the decision tree to avoid over-fitting caused by excessive complexity of the tree, but the specific appropriate value needs to be further optimized and determined through cross-validation and other methods), the minimum number of sample splits (set to a smaller value, such as 2, which means that the node must contain at least 2 samples before the node can be split to prevent the generation of too fragmented leaf nodes. Similarly, this parameter will be adjusted according to the training situation later), the minimum number of sample leaves (set to 1 to ensure that the leaf node contains at least 1 sample to make it representative) and the splitting standard for feature selection (here the mean square error is selected as an indicator to measure the quality of node division. Because it is a regression task, the mean square error can well reflect the degree of difference between the predicted value and the actual value). The feature data after preprocessing, selection and engineering and the corresponding actual navigation signal strength data are divided into training set and test set in a certain ratio. The division ratio is 8:2, that is, 80% of the data is used for model training to let the model learn the relationship between features and signal strength; 20% of the data is used as a test set to evaluate the prediction performance of the model on unseen data after the model training is completed, so as to verify the generalization ability of the model and ensure that the model will not be overfitted and can accurately predict the navigation signal strength for new scenarios in actual applications; The training set data is input into the selected CART algorithm for training. The decision tree algorithm will continuously build the branch structure of the tree based on the relationship between the features and the target variable. At each node, the mean square error corresponding to different features is calculated, and the feature with the smallest mean square error is selected as the basis for dividing the node. The nodes are continuously split so that the data in each leaf node are as similar as possible in terms of navigation signal strength. This process is to learn the rules in the data for subsequent predictions. During the training process, it will be iterated many times. Each iteration will adjust the structure and parameters of the tree according to the current training data, and gradually optimize the model until the preset stop condition is reached (for example, the maximum number of iterations is reached, or the performance on the validation set is no longer improved, etc.); After completing the construction, evaluation and optimization of the model, the location information, terrain parameters and weather parameters of the selected location of the target UAV (these parameters undergo the same preprocessing, feature selection and engineering steps to keep them consistent with the data format when the model was trained) can be input into the trained decision tree model. The model will output the corresponding navigation signal strength prediction value based on the complex decision rules it has learned and the mapping relationship between features and signal strength. This is the predicted navigation signal strength.

[0022] In another embodiment, there are multiple selected positions, and the multiple selected positions are determined by gradually fine-tuning the position prediction model; Specifically, the smoothing factor α of the latitude, longitude, and altitude mentioned above can be x , α y and α z Perform gradual fine-tuning. For example, except for the first selected position, the subsequent selected positions are obtained by the fine-tuned position prediction model, such as multiplying by the adjustment coefficient, and each is gradually increased or decreased in proportion, so that the selected position obtained after adjustment changes slowly compared with the original trajectory, avoiding large changes that cause target drone identification and cause deception failure; it should be pointed out that the navigation signal strength prediction and navigation decoy signal generation are re-performed for each selected position.

[0023] In another embodiment, when the target UAV reaches the first selected position, a navigation decoy signal is transmitted to the target UAV, and it is determined whether the decoy is successful. If the decoy is not successful, S2-S4 are repeated, that is, the selected position is re-determined, the navigation signal strength is predicted, the navigation decoy signal is generated, and the navigation decoy signal is transmitted, until the decoy is successful; Exemplarily, the method for determining whether the deception is successful includes: obtaining multiple position information of the target UAV between the first selected position and the second selected position, and determining whether it conforms to the position prediction model; if so, determining that the deception is not successful, that is, determining whether the target UAV deviates from the original trajectory according to the position prediction model; if not, determining that the deception is not successful.

[0024] In another embodiment, when it is determined that the deception fails, deceptive sensitive information is transmitted to the target UAV in a non-sensitive area to mislead the target UAV to fly to the non-sensitive area.

[0025] Specifically, multiple high-precision radars, optical detectors, radio spectrum analyzers and other equipment can be used to work together to cover the target area of ​​interest to the target drone and a certain range around it, ensuring that the target drone's position, altitude, speed, model and other key information can be captured in real time and accurately. The data of these monitoring devices are aggregated to the central control unit through a high-speed data link. The central control unit has a built-in intelligent analysis algorithm, which quickly processes the collected data according to preset rules to determine whether the target drone is approaching a sensitive area. Once it is determined that the deception has failed, the subsequent deception process will be immediately initiated.

[0026] Next, multiple decoy signal transmitting base stations and sensitive ground buildings are set up in non-sensitive areas that are far away from sensitive areas and have been pre-surveyed and selected. These base stations are equipped with high-power transmitting devices specifically for common drone communication frequency bands, and can quickly adjust the transmission parameters according to the instructions of the central control unit. After receiving the decoy instruction, the transmitting base station carefully constructs content containing false sensitive information, such as some confidential information, based on the model and communication protocol characteristics of the target drone. When the drone captures this false sensitive information, it can make the drone operator believe that the place where the information is sent has important reconnaissance value, thereby changing the flight direction of the drone.

[0027] At the same time, in order to ensure the deception effect, during the launch of deceptive sensitive information, the flight attitude changes of the target drone are continuously observed through the monitoring system, and the intensity, frequency and content details of the deceptive sensitive information are dynamically adjusted according to the real-time response of the drone. If the flight trajectory of the drone is found to deviate from the original sensitive direction, the intensity of the deceptive sensitive information will be gradually weakened to guide it to land smoothly in the selected non-sensitive area; if the drone still does not change its course as expected, the deception strategy will be further strengthened to increase the credibility of the false information until it is successfully lured away from the sensitive area to ensure the airspace safety of the sensitive area.

[0028] The embodiment of the present application also provides a navigation decoy device for countering drones, comprising: a construction module, used to obtain the position information of the target drone at multiple times and construct a position prediction model; a prediction module, used to estimate the navigation signal strength when the target drone is located at a selected position, and obtain a predicted navigation signal strength; a generation module, used to generate a navigation decoy signal according to the predicted navigation signal strength; a decoy module, used to transmit a navigation decoy signal to the target drone when the target drone reaches the selected position; This embodiment uses a computer program to construct a construction module, a prediction module, a generation module and a decoy module. The construction module collects data to establish a position prediction model. The prediction module predicts the navigation signal strength according to the position information. The generation module calls the radio frequency signal generator to generate a navigation decoy signal. The decoy module calls the antenna to transmit the navigation decoy signal. This embodiment uses the position prediction model to accurately predict the position of the target UAV, and then accurately estimates the predicted navigation signal strength, so that the generated navigation decoy signal is less different from the real navigation signal, which reduces the possibility of being identified and improves the success rate of decoy.

[0029] In another embodiment, the prediction module inputs the location information, terrain parameters and weather parameters of the selected location into a pre-established signal strength prediction model to obtain the predicted navigation signal strength; wherein the signal strength prediction model is obtained by training a decision tree model; this embodiment uses a decision tree model to construct a signal strength prediction model, and its training process does not require large-scale iterative optimization, complex gradient calculations, etc. The training process can be completed relatively quickly, the amount of calculation is relatively small, and the prediction results can be given quickly. For navigation spoofing scenarios with high real-time requirements, the required signal strength information can be provided in a timely manner. For the specific establishment process, please refer to the previous text.

[0030] In another embodiment, there are multiple selected positions, and the multiple selected positions are determined by gradually fine-tuning the position prediction model; this embodiment can smooth the latitude, longitude, and altitude in the above text. x , α y and α z Gradually fine-tune the position so that the selected position obtained according to the adjusted position prediction model changes slowly compared to the original trajectory, avoiding large changes that may cause the target drone to be identified and cause the deception to fail.

[0031] An embodiment of the present application also provides a navigation deception device for countering drones, comprising: a processor; a memory for storing executable instructions; wherein the processor is configured to execute the executable instructions to implement the navigation deception method for countering drones; the device of this embodiment can be a mobile phone, a laptop computer, a tablet computer, a vehicle-mounted terminal, a drone, etc., which has a memory and a processor arranged therein to execute the target detection method of the above-mentioned embodiment.

[0032] Although the embodiments of the present invention have been disclosed as above, they are not limited to the applications listed in the specification and the implementation modes, and they can be fully applied to various fields suitable for the present invention. For those familiar with the art, additional modifications can be easily implemented. Therefore, without departing from the general concept defined by the claims and the scope of equivalents, the present invention is not limited to the specific details and the illustrations shown and described herein.

Claims

1. A navigation deception method for countering drones, applied to a navigation deception device, characterized in that: include: S1: Obtain the location information of the target UAV at multiple times and build a location prediction model; S2: estimating the navigation signal strength when the target UAV is located at a selected position to obtain a predicted navigation signal strength; S3: generating a navigation decoy signal according to the predicted navigation signal strength; S4: When the target UAV arrives at the selected location, a navigation decoy signal is transmitted to the target UAV.

2. The navigation deception method for countering drones according to claim 1, characterized in that: In S2, the location information, terrain parameters and weather parameters of the selected location are input into a pre-established signal strength prediction model to obtain the predicted navigation signal strength; Wherein, the signal strength prediction model is obtained by training a decision tree model.

3. The navigation deception method for countering drones according to claim 1, characterized in that: The strength of the navigation decoy signal is at least consistent with the predicted navigation signal strength.

4. The navigation deception method for countering drones according to claim 1, characterized in that: There are multiple selected positions, and the multiple selected positions are determined by gradually fine-tuning the position prediction model.

5. The navigation deception method for countering drones according to claim 4, characterized in that: When the target UAV reaches the first selected position, a navigation decoy signal is transmitted to the target UAV, and it is determined whether the decoy is successful. If the decoy is not successful, S2-S4 are repeated.

6. The navigation deception method for countering drones according to claim 5, characterized in that: Methods for determining whether the deception was successful include: A plurality of position information of the target UAV is obtained between the first selected position and the second selected position, and it is determined whether it conforms to the position prediction model. If so, it is determined that the deception is not successful.

7. A navigation deception device for countering drones, characterized in that: include: A construction module is used to obtain the location information of the target UAV at multiple times and build a location prediction model; A prediction module, used to estimate the navigation signal strength when the target UAV is located at a selected position to obtain a predicted navigation signal strength; A generating module, used for generating a navigation decoy signal according to the predicted navigation signal strength; The deception module is used to transmit a navigation deception signal to the target UAV when the target UAV reaches a selected position.

8. The navigation deception device for countering drones according to claim 7, characterized in that: The prediction module inputs the location information, terrain parameters and weather parameters of the selected location into a pre-established signal strength prediction model to obtain the predicted navigation signal strength; wherein the signal strength prediction model is obtained by training a decision tree model.

9. The navigation deception device for countering drones according to claim 7, characterized in that: There are multiple selected positions, and the multiple selected positions are determined by gradually fine-tuning the position prediction model.

10. A navigation decoy device for countering drones, characterized in that: include: processor; A memory for storing executable instructions; Wherein, the processor is configured to execute the executable instructions to implement the navigation spoofing method for drone countermeasure as described in any one of claims 1 to 6.

Images