An identity authentication method based on unmanned aerial vehicle air-ground safe communication

Abstract

The application discloses an identity authentication method based on unmanned aerial vehicle air-ground safe communication and relates to the technical field of unmanned aerial vehicle safe communication, which can solve the problem of how to realize fast and reliable identity authentication of unmanned aerial vehicle physical entities in a weak network environment with unstable and interruptible wireless channels to prevent simulator fraud and replay attacks, and comprises the following steps: obtaining historical flight data of the unmanned aerial vehicle and determining response delay statistics of the unmanned aerial vehicle according to the historical flight data; establishing and updating historical response criteria of the unmanned aerial vehicle according to the response delay statistics of the unmanned aerial vehicle; obtaining current authentication features sent by the unmanned aerial vehicle when the communication link between the unmanned aerial vehicle and the ground control station is reconnected after being interrupted; performing statistical consistency comparison on the current authentication features according to the historical response criteria; and determining whether the identity authentication of the unmanned aerial vehicle is passed according to the comparison result.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) secure communication technology, and specifically to an identity authentication method based on UAV air-to-ground secure communication. Background Technology

[0002] With the widespread application of industrial drones in fields such as power line inspection, border security, and logistics transportation, the security and reliability of their air-to-ground communication links have become increasingly critical issues. In actual operations, drones often need to operate under non-line-of-sight communication conditions such as complex terrain or strong electromagnetic interference, which may lead to frequent short-term interruptions of the wireless data link. When re-establishing a connection after a communication interruption, fast and reliable authentication of the drone to prevent unauthorized access is a prerequisite for ensuring the safety of the entire operation.

[0003] Currently, mainstream identity authentication technologies rely on digital certificates with pre-set keys, digital signatures, or transport layer security handshake protocols. These methods can effectively verify the legitimacy of digital identities at the network protocol level, but their effectiveness is limited when dealing with high-fidelity simulated software attacks or complex wireless transmission environments. Especially in weak network environments with poor signal quality and severe packet loss, these authentication mechanisms based on multiple handshakes or continuous signal analysis often face the dilemma of significantly increased authentication latency or failure of feature extraction, leading to a decrease in the efficiency and reliability of the overall authentication process. Summary of the Invention

[0004] To address the current technical challenge of achieving rapid and reliable authentication of UAV physical entities in weak network environments with unstable and easily interrupted wireless channels, thereby preventing simulator spoofing and replay attacks, this invention aims to provide an authentication method based on secure air-to-ground communication for UAVs. The specific technical solution adopted is as follows: In a first aspect, the present invention provides an identity authentication method based on UAV air-to-ground secure communication, comprising: acquiring historical flight data of the UAV and determining a response delay statistic of the UAV based on the historical flight data; wherein the response delay statistic is used to characterize the distribution characteristics of the time delay of the UAV's response to disturbances within a historical time period; establishing and updating a historical response benchmark of the UAV based on the response delay statistic of the UAV; wherein the historical response benchmark is used to characterize the individual physical response characteristics of the UAV; acquiring the current authentication feature sent by the UAV when the communication link between the UAV and the ground control station is reconnected after an interruption; wherein the current authentication feature is a physical response feature extracted by the UAV based on the current flight status data; performing a statistical consistency comparison of the current authentication feature based on the historical response benchmark, and determining whether the UAV's identity authentication is successful based on the comparison result.

[0005] In a second aspect, the present invention provides an identity authentication system based on UAV air-to-ground secure communication, comprising: a ground control station and a UAV; the ground control station being configured to perform the method described in the first aspect and any possible implementation thereof; and the UAV being configured to, during flight, extract physical response features based on current flight state data to generate current authentication features, and, upon reconnection after a communication link with the ground control station is interrupted, send a reconnection request containing the current authentication features to the ground control station.

[0006] Thirdly, the present invention provides an electronic device, comprising: a processor and a memory; wherein the memory is used to store one or more programs, the one or more programs including computer execution instructions, and when the electronic device is running, the processor executes the computer execution instructions stored in the memory to cause the electronic device to perform the identity authentication method for UAV air-to-ground secure communication as described in the first aspect and any possible implementation thereof.

[0007] The present invention has the following beneficial effects: by continuously learning and constructing the historical response statistical benchmark of individual UAVs during normal communication by the ground control station, and by comparing the physical response characteristics generated by the UAV in real time with the benchmark when communication is interrupted and reconnected, the invention achieves fast and reliable identity authentication of UAV physical entities under unstable wireless channel conditions, effectively defending against deception attacks based on software simulation or data replay. Attached Figure Description

[0008] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 This is a flowchart illustrating an identity authentication method based on UAV air-to-ground secure communication, provided as an embodiment of the present invention. Figure 2 This is a schematic diagram of the architecture of an identity authentication system based on UAV air-to-ground secure communication, provided in one embodiment of the present invention. Figure 3 This is a schematic diagram of the architecture of a ground control station provided in one embodiment of the present invention. Detailed Implementation

[0010] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the specific implementation methods, structures, features, and effects of the present invention will be described in detail below with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0011] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0012] The specific scheme of the identity authentication method based on UAV air-to-ground secure communication provided by the present invention will be described in detail below with reference to the accompanying drawings.

[0013] For example, such as Figure 1 The diagram shown is a flowchart illustrating an identity authentication method based on UAV air-to-ground secure communication according to an embodiment of the present invention, including the following steps: S101. Acquire historical flight data of the UAV and determine the response delay statistics of the UAV based on the historical flight data. The response delay statistics are used to characterize the distribution characteristics of the time delay of the UAV's response to disturbances within a historical time period.

[0014] Specifically, historical flight data is a sequence of multiple response delay features periodically calculated and reported by the UAV during the period of communication with the UAV. Each "response delay feature" is a specific numerical parameter that characterizes the physical response speed of the UAV from sensing environmental disturbances (manifested as track deviations) to executing corresponding attitude correction actions (manifested as control commands) within its corresponding historical sampling period (e.g., 2 seconds). Its value is calculated by time-series matching of multiple valid "disturbance-correction" event pairs extracted within the period and then calculating the average time difference.

[0015] It should be noted that the method for determining the response delay characteristics constituting historical flight data is the same as the method for determining the current response delay characteristics described in subsequent sections S301-S304. This is because, during normal communication between the UAV and the ground control station, the UAV will also follow the procedures outlined in S301. The method described in S304 calculates a response delay feature value based on the real-time track deviation sequence and control command sequence within each sampling period, and sends this value as periodic reporting data to the ground control station. Therefore, each response delay feature in the historical flight data is essentially the "current response delay feature" at each past sampling moment, i.e., a core parameter in the authentication features of the historical version. By statistically analyzing these historical response delay features, the ground control station obtains a historical response benchmark that characterizes the individual physical response properties of the UAV.

[0016] Subsequently, the ground control station performs statistical analysis on the historical flight data constituted by this time series to determine a quantitative indicator used to describe the central trend and dispersion of the series, namely the response delay statistic. The determination process is described in S301-S304 below.

[0017] Optionally, the response delay statistics include at least the historical response delay mean and the historical response delay standard deviation. The ground control station acquires the UAV's historical flight data and determines the UAV's response delay statistics based on the historical flight data, specifically including the following steps: (1) Based on historical flight data, determine the historical response delay characteristics corresponding to each of the multiple historical sampling periods. The historical response delay characteristics are used to characterize the average time difference between the UAV being disturbed and generating a corrective action within the corresponding historical sampling period.

[0018] In this step, the ground control station directly obtains the historical response delay characteristics for each period from the data periodically reported by the UAV. These characteristics constitute the original sample set for statistical analysis.

[0019] (2) Determine the mean and standard deviation of historical response delay based on all historical response delay characteristics.

[0020] Furthermore, the ground control station maintains a sliding time window with a preset length, and calculates the arithmetic mean of the latest multiple historical response delay feature samples within the window as the historical response delay mean, and calculates the standard deviation of these samples as the historical response delay standard deviation, thereby realizing the dynamic updating of response delay statistics.

[0021] Therefore, the ground control station can dynamically characterize and track the unique, statistically distributed response hysteresis characteristics of the UAV based on its recent actual flight response data.

[0022] S102. Based on the response delay statistics of the UAV, establish and update the historical response benchmark of the UAV. The historical response benchmark is used to characterize the individual physical response characteristics of the UAV.

[0023] Specifically, the historical response benchmark is a data structure stored in non-volatile memory, its core being a parameter pair consisting of the historical response delay mean and the historical response delay standard deviation. This parameter pair statistically characterizes the central location and fluctuation range of the physical response characteristics of an individual UAV. Therefore, the historical response benchmark serves as a unique "physical fingerprint" for the UAV, used in subsequent authentication processes to verify whether the response characteristics of entities to be authenticated originate from the same physical entity.

[0024] For example, the ground control station establishes and updates the UAV's historical response baseline based on the UAV's response delay statistics. Specifically, the ground control station stores the calculated historical response delay mean and historical response delay standard deviation as a set of core parameters, and repeats the statistical calculation process in S101 to update these parameters each time a new response delay characteristic report is received. It should be noted that the specific process of the aforementioned sub-steps can be found in S201-S203 below, and will not be repeated here.

[0025] Therefore, the ground control station continuously learns and refines the individual response model of the UAV during the communication connection, and immediately freezes the current historical response benchmark the moment the communication link is interrupted, using it as the basis for comparison in subsequent reconnection authentication.

[0026] S103. When the communication link between the UAV and the ground control station is reconnected after an interruption, the current authentication feature sent by the UAV is obtained. The current authentication feature is the physical response feature extracted by the UAV based on the current flight status data.

[0027] For example, the ground control station acquires the current authentication features sent by the drone, specifically including the following steps: (1) Receive the reconnection request message sent by the UAV during communication reconnection.

[0028] Specifically, the ground control station listens for a link re-establishment request from the interrupted UAV on the communication interface. This request is a data packet conforming to the communication protocol. The interaction process for this request is existing technology and will not be described in detail here.

[0029] (2) Determine the current authentication feature based on the reconnection request message.

[0030] This step can be performed in two ways, as follows: Scenario 1: The reconnection request message includes the current authentication characteristics.

[0031] In this scenario, the current authentication characteristic is the structured data uploaded by the UAV and encapsulated in the payload of the reconnection request data packet after the communication link between the UAV and the ground control station has been interrupted and reconnected. In other words, the current authentication characteristic can be determined by the UAV.

[0032] Case 2: The reconnection request message includes the track deviation sequence and control command sequence within the current sampling window.

[0033] In this case, the UAV adds the track deviation sequence and control command sequence within the current sampling window to the reconnection request message sent to the ground control station. Then, after the ground control station obtains the aforementioned two sequences, it determines the current authentication characteristics on its own.

[0034] It should be noted that the specific process for determining the current authentication features can be found in S301-S304 below, and will not be repeated here.

[0035] Therefore, the ground control station can obtain the credential data generated instantly by the drone to prove its physical identity the moment it reconnects.

[0036] S104. Perform a statistical consistency comparison of the current authentication features based on the historical response benchmark, and determine whether the drone's identity authentication is successful based on the comparison results.

[0037] For example, when the ground control station performs a statistical consistency comparison of the current authentication features based on historical response benchmarks and determines whether the UAV's identity authentication is successful based on the comparison results, the process specifically includes: the ground control station parses the current response delay feature from the current authentication features and reads the historical response delay mean and historical response delay standard deviation from the frozen historical response benchmarks. Next, the ground control station calculates the standardized statistical distance of the current response delay feature relative to the historical statistical parameter and compares this distance with a preset consistency threshold to obtain the statistical consistency comparison result. Finally, based on the successful statistical consistency comparison, the ground control station further performs a physical rationality check on the current authentication features (such as checking whether the response delay is within the physically possible range) and makes a final authentication decision by combining the results of the two checks. It should be noted that the specific procedures of the aforementioned sub-steps can be found in S401-S403 below, and will not be repeated here.

[0038] Therefore, the ground control station can efficiently and reliably identify legitimate physical drone entities and reject simulation software or replay attacks by rigorously comparing the "snapshot" of the drone's real-time response with its "physical fingerprint" learned over a long period of time and verifying it according to physical rules.

[0039] Based on the above technical solution, this invention enables the ground control station to continuously learn and construct historical response statistical benchmarks for individual UAVs during normal communication. When communication is interrupted and reconnected, the physical response characteristics generated by the UAV in real time are statistically compared with the benchmarks. This achieves fast and reliable identity authentication of UAV physical entities under unstable wireless channel conditions and effectively defends against deception attacks based on software simulation or data replay.

[0040] For example, in another identity authentication method based on UAV air-to-ground secure communication provided by one embodiment of the present invention, the historical response benchmark of the UAV is established and updated according to the response delay statistics of the UAV, specifically including the following steps: S201. During the communication connection with the UAV, continuously receive the response delay characteristics periodically reported by the UAV, and maintain a historical observation window based on the response delay characteristics reported most recently.

[0041] In this step, the historical observation window is a data structure with a fixed capacity stored in memory, used to save several recently received response delay features in chronological order. Its function is to ensure that the samples used for statistical calculations always come from the most recent and representative flight status of the UAV.

[0042] Specifically, when the ground control station maintains a communication connection with the UAV, it receives a response delay characteristic calculated and sent by the UAV at fixed intervals (e.g., every 2 seconds). The ground control station will receive each new message. Add it to the end of the historical observation window. If the window is full (i.e., the number of stored features reaches the preset window length N), remove the oldest feature from the window to maintain the N most recent response delay features within the window. This process is implemented using a sliding window algorithm with a finite-length First-In-First-Out (FIFO) queue. For example, the window length N can be 50, which means the baseline is obtained from the data of the most recent approximately 100 seconds (50 × 2 seconds). This value balances the stability of the statistics (requiring sufficient samples) with adaptability to the slow changes in UAV characteristics.

[0043] S202. Calculate the response delay statistics based on all response delay characteristics within the historical observation window.

[0044] Furthermore, the ground control station performs statistical analysis on all N response delay feature samples currently stored within the historical observation window. Specifically, the response delay statistic, i.e., the historical response delay mean, is calculated using the following two formulas. and historical response delay standard deviation :

[0045]

[0046] in, This represents the value of the i-th response delay feature in the historical observation window. This represents the historical average response time, i.e., the average response time across all values ​​in the window. The arithmetic mean is used to characterize the central trend of the drone's recent response delay. This represents the standard deviation of historical response delays, i.e., all responses within the window. Compared to The degree of dispersion is a measure used to characterize the range of fluctuations in its response delay. N represents the length of the historical observation window (i.e., the number of feature samples). The calculated... and These two parameters together constitute the "historical response baseline" of the drone at the current moment. This baseline is a tuple ( , This statistically completes the concentrated location and dispersion of the physical response characteristics of individual drones, which is equivalent to their unique "physical fingerprint".

[0047] S203. In response to the detection of a communication link interruption with the UAV, stop updating the historical observation window and use the response delay statistics at the time of the interruption as the historical response benchmark for subsequent verification.

[0048] Finally, the ground control station continuously monitors the communication link status with the UAV. Upon detecting a communication link interruption (e.g., the consecutive loss of multiple heartbeat signals), a baseline freeze procedure is immediately triggered. In this procedure, the ground control station first stops performing the sliding window update operation in S201, meaning it no longer receives new data. Or update the window content. Then, the pre-calculated average historical response delay corresponding to the historical observation window at the moment of interruption will be used. and historical response delay standard deviation Perform sealing. This applies to the sealed parameters ( , This serves as a "frozen historical response baseline," used to verify the consistency of the current authentication features sent by the drone when it subsequently attempts to reconnect. This freezing mechanism ensures that the comparison baseline used for authentication is stable and consistent, avoiding baseline uncertainty caused by the inability to obtain new data during the disconnection period.

[0049] Based on the above technical solution, this embodiment of the invention calculates and updates the response delay statistics of the UAV in real time by dynamically maintaining a finite-length historical observation window, thereby establishing a historical response benchmark that accurately characterizes its individual characteristics. This benchmark is automatically frozen in the event of a communication interruption, ensuring the stability of the authentication reference. This allows the authentication method to be based on the latest and most reliable "physical fingerprint" of the UAV, improving the accuracy of authentication and its robustness to slow changes in entity characteristics.

[0050] For example, in another identity authentication method based on UAV air-to-ground secure communication provided by an embodiment of the present invention, the current response delay characteristic is determined specifically according to the following steps: S301. Obtain the track deviation sequence and control command sequence within the current sampling window.

[0051] Specifically, the two sequences acquired in this step are data collected in real time within a preset sampling window (e.g., 2 seconds). The "track deviation sequence" records the magnitude of the UAV's deviation from its predetermined flight path due to environmental wind disturbances; its value reflects the excitation effect of the external environment on the aircraft. The "control command sequence" records the throttle or control surface control quantities output by the flight control system to the motors in real time to correct the above deviations; its value reflects the aircraft's response to the excitation. The two sequences are strictly synchronized in time, together forming the original data pair characterizing the UAV's "stimulus-response" dynamic characteristics.

[0052] Optionally, the specific method for obtaining the track deviation sequence is as follows: the airborne terminal uses a global navigation satellite system (such as Global Positioning System (GPS) or Real-time Kinematic (RTK)) and an inertial navigation system (INS) to output the current real-time three-dimensional coordinates, compares these coordinates with the route coordinates points pre-planned and sent by the ground control station in real time, obtains the track deviation value at each moment by calculating Euclidean distance or cross-track error algorithm, and records them in chronological order to form a sequence.

[0053] Optionally, the specific method for obtaining the control command sequence is as follows: directly obtain it from the output end of the flight control system, such as reading the control pulse width modulation (PWM) signal value, DShot protocol digital command, or control quantity on the controller area network (CAN) bus from the main control chip of the flight control system in real time, and recording it in chronological order.

[0054] S302. Extract the set of deviation peak points from the track deviation sequence and extract the set of command peak points from the control command sequence.

[0055] This step aims to extract discrete key points representing effective physical events (significant perturbations and their corrections) from continuous sequence data, thereby forming a set of deviation peak points and a set of command peak points.

[0056] First, the validity of the data needs to be determined: calculate the variance of the trajectory deviation sequence within the current sampling window. If the variance is less than a preset silence threshold, the current environment is considered stable, and no valid features exist, thus terminating the extraction process. If the variance is greater than or equal to the silence threshold, peak extraction is performed. The silence threshold is typically determined based on the noise variance of the sensor under windless hovering conditions; an example value could be 0.01 m / s². 2 ) 2 The purpose is to filter out data dominated by sensor noise during stable flight, thus avoiding invalid calculations when there are no effective features.

[0057] Next, for the track deviation sequence, all local maxima (i.e., data points whose values ​​are greater than their immediate neighbors) are identified, and points with amplitudes less than a preset noise threshold are filtered out to suppress background noise, resulting in preliminary candidate deviation peaks. For the control command sequence, the same local maxima identification and noise filtering operations are performed to obtain candidate command peaks. The noise threshold is set based on the sensor resolution and typical signal amplitude; exemplary values ​​could be 0.05m (for deviations) and 2% (for commands).

[0058] Finally, the identified deviation peak candidate points are sorted from largest to smallest amplitude, and only the top P points (e.g., P=10) are retained to form the deviation peak point set; the same sorting and truncation operation is performed on the instruction peak candidate points to form the instruction peak point set. This Top-P truncation strategy greatly reduces the subsequent computational complexity while retaining the most important features, making it suitable for operation in embedded systems.

[0059] S303. Perform timing matching between the deviation peak point set and the instruction peak point set to obtain at least one matching event pair.

[0060] Optionally, when a ground control station or drone performs this step, it specifically includes the following steps: (1) Construct a bipartite graph with the set of deviation peak points as the first node set and the set of instruction peak points as the second node.

[0061] In this step, each deviation peak point (including the time of occurrence) will be... and amplitude (This is considered a binary division) Figure 1 The nodes on the side, each instruction peak point (including the time of occurrence) and amplitude ( ) is considered as a node on the other side. Construct a graph structure containing all possible potential causal relationships to provide a basis for subsequent global optimization and screening.

[0062] (2) For nodes in the first node set and nodes in the second node set, if the time difference between the two nodes is within a preset reasonable time interval, a connection edge is established between the two nodes, and a weight is assigned to the connection edge according to the relationship between the corresponding amplitudes of the two nodes.

[0063] Furthermore, the preset reasonable time interval is determined by the minimum response delay. and maximum response latency The distinction between the two is determined by the mechanical inertia of the UAV body and is used to screen for non-physical correlations. For any deviation point i and command point j, the correlation is determined based on their occurrence time. , Calculate the time difference = - Only when ≤ ≤ Only when the deviation point is reached is a directed edge (from the deviation point to the instruction point) established between the two points. For each established edge, a weight is assigned to it. This weight is used to quantify the plausibility of this candidate causal relationship. The weight is calculated using the following formula:

[0064] in, This represents the normalized amplitude of the command peak point j, obtained by adjusting the amplitude of peak point j. The result is obtained after performing maximum-minimum normalization to eliminate dimensions. This represents the normalized amplitude of the deviation peak point i, obtained by adjusting the amplitude of the deviation peak point i. The result is obtained after normalizing the maximum and minimum values ​​to eliminate dimensions. This is a preset nominal gain coefficient, determined by the flight control system's control law design. It represents the ratio of the correction command to the deviation under ideal conditions and is a dimensionless constant. It should be noted that... The value will not be zero in the algorithm process because peak extraction is only performed when the variance of the track deviation sequence exceeds the silent threshold.

[0065] In the above formula, the actual observed gain ratio is calculated ( ) and ideal nominal gain The absolute difference. The smaller this difference, the more the magnitude of the correction aligns with the control system's expectations, and the higher the likelihood that it represents a true causal relationship. Therefore, the calculated weights... The larger the denominator "1 + difference" is, the smaller it will be (because the denominator "1 + difference" will be). The formula is... The format ensures that the weight is always positive and does not exceed 1.

[0066] (3) In the bipartite graph, determine the matching result with the largest total weight to obtain the matching event pair.

[0067] Finally, on the constructed weighted bipartite graph, a maximum weight perfect matching algorithm for bipartite graphs (such as the Kuhn-Munkres algorithm) is run. The algorithm's function is to select a matching scheme from the graph such that the selected edges have no common nodes (i.e., a deviation point is paired with only one instruction point, and vice versa), and the sum of the weights of all selected edges is maximized. The algorithm's output is the final set of matched event pairs. This step, through global optimization, robustly reconstructs the most physically consistent causal relationship sequence from discrete points that may contain noise, packet loss, or temporal misalignment.

[0068] S304. Calculate the current response delay feature based on the time difference of each event pair in the matched event pair.

[0069] For example, K matching event pairs are obtained via S303. For each matching pair k, their time differences are known. The current response delay characteristic is obtained by calculating its arithmetic mean. The current response delay characteristic Core quantitative features used to characterize the dynamic response characteristics of drones at the current moment.

[0070] Optionally, after S304, an effective correction percentage for auxiliary features can also be calculated. Where M represents the total number of points in the set of deviation peak points. This represents the proportion of perturbation events that were successfully matched and explained. A higher value indicates a more complete physical causal relationship in the current data. This effective correction proportion is used to determine the results of the subsequent second verification.

[0071] Based on the above technical solution, this embodiment of the invention extracts discrete peak points from continuous data and utilizes a bipartite graph maximum weight matching algorithm based on physical constraints. This enables robust and accurate reconstruction of the physical causal relationship of a UAV's "disturbance-correction" from potentially incomplete and noisy real-time data, and calculates quantitative characteristics characterizing its response speed. This method does not rely on waveform continuity and is insensitive to data loss, thus achieving reliable extraction of the dynamic characteristics of physical entities in weak network environments.

[0072] For example, in another identity authentication method based on UAV air-to-ground secure communication provided by one embodiment of the present invention, a statistical consistency comparison of the current authentication features is performed based on historical response benchmarks, and the identity authentication of the UAV is determined based on the comparison results. Specifically, this is determined according to the following steps: S401. Obtain the mean and standard deviation of historical response delay from the historical response benchmark.

[0073] Specifically, the historical response baseline in this step is a set of statistical parameters frozen and saved when the communication link was interrupted. When the ground control station needs to perform authentication comparison, it reads this set of parameters from non-volatile memory. The core of this baseline consists of two values: (Historical average response delay) and (Standard deviation of historical response delay). This represents the average response latency of the drone in the period before the connection was lost. This represents the range of fluctuation in its response delay. These two parameters together constitute the statistical reference used for comparison; the specific calculation process can be found in sections S201-S203 above.

[0074] S402. Calculate the standardized statistical distance based on the current response delay characteristics, the historical response delay mean, and the historical response delay standard deviation.

[0075] Furthermore, the ground control station obtains the current response delay feature from the parsed current authentication feature. In order to quantify Compared with historical benchmarks (by and To determine the degree of deviation (of the constituent elements), a normalized statistical distance, i.e., the standardized statistical distance, needs to be calculated. For example, the standardized statistical distance is calculated using the following formula:

[0076] in, This represents the standardized statistical distance. This represents the current response delay characteristics parsed from the drone reconnection request (or calculated based on the track deviation sequence and control command sequence contained in the reconnection request). This represents the average historical response delay obtained from the historical response benchmark. This represents the standard deviation of historical response delay obtained from the historical response benchmark. This represents an extremely small positive number and is called a smoothing factor. It is used to prevent the denominator from being zero. For example, it takes the value of 10 to the power of negative 5.

[0077] In the above formula, the current feature is calculated first. Compared with historical average The absolute value of the deviation. Then, divide this absolute deviation by the historical standard deviation. With smoothing factor The sum of the values. Dividing by the standard deviation standardizes the absolute deviation, making it a dimensionless value that is independent of the original data units and only represents how many standard deviations it deviates from the historical distribution.

[0078] S403. Compare the standardized statistical distance with the preset consistency threshold to obtain the comparison results.

[0079] In this step, the calculated standardized statistical distance Z is compared with a preset consistency threshold. Compare them. Consistency threshold. The value is based on the statistical significance test principle. An example value is 3.0. The rule for its selection is: under the assumption that the data follows a normal distribution, the probability of a Z-value greater than 3 is extremely small (approximately 0.3%). Therefore, setting the threshold to 3 means that if the current drone's response characteristics belong to the same statistical distribution as the historical baseline (i.e., the same physical entity), then its Z-value has a 99.7% probability of being less than or equal to 3. This threshold can be adjusted according to the actual application's requirements for security and false rejection rate.

[0080] Furthermore, if Z≤ If Z > 0, then the current characteristics are considered statistically consistent with the historical baseline, and the comparison result is considered passed; if Z > 0. If the deviation is statistically significant, the comparison result is considered unsuccessful.

[0081] S404. Determine whether the identity authentication of the drone has passed based on the comparison results.

[0082] Optionally, when the ground control station determines whether the UAV's identity authentication is successful based on the comparison results, it specifically includes the following steps: (1) Perform physical validity verification on the current authentication features and obtain the verification results.

[0083] First, it verifies whether the current response latency characteristic is within a preset physically reasonable time interval, obtaining the first verification result. Specifically, this "physically reasonable time interval" is determined by the drone's minimum response latency. and maximum response latency These two parameters are defined by the machine's mechanical structure and are preset constants. The inspection conditions are: ≤ ≤ If the conditions are met, the first verification result passes; if not (response is too fast or too slow), the verification is directly deemed to have failed.

[0084] Subsequently, when the current authentication feature contains a valid correction percentage, it is verified whether the valid correction percentage is greater than or equal to a preset valid threshold, resulting in a second verification result. The valid correction percentage is calculated based on the number of matching event pairs and the number of elements in the deviation peak point set. Specifically, the valid correction percentage is parsed from the current authentication feature. Effective threshold An example value is 0.6. The rule for its value is that this threshold sets a minimum standard for the effectiveness of closed-loop control. A value that is too low indicates that most detected disturbances could not be matched with a reasonable correction instruction, suggesting that the data may not have been generated in real time, but rather came from historical replays with misaligned timing. ≥ If so, the second verification result passes.

[0085] Finally, if both the first and second verification results are passed, the verification result is considered passed. Specifically, the final conclusion of the physical rationality verification depends on the simultaneous satisfaction of the above two checks (response delay range check and effective correction ratio check). Failure of either check will result in the entire verification failing.

[0086] (2) When the comparison result is passed and the verification result is passed, the identity authentication of the drone is determined to be successful.

[0087] Finally, a final logical AND decision is made. Only when the "statistical consistency comparison result" obtained in S403 is successful and the "physical rationality verification result" obtained in S404(1) is also successful, will the ground control station finally determine that the UAV's identity authentication is successful, restore its communication link, and lift the baseline freeze. If either result is unsuccessful, authentication fails, the reconnection request is rejected, and a security log is recorded.

[0088] Based on the above technical solution, this invention addresses the problem of simulators struggling to reproduce individual statistical characteristics by introducing "standardized statistical distance" to rigorously quantify and compare the real-time response characteristics of drones with their historical physical fingerprints. Simultaneously, "physical rationality verification" (including response delay boundary checks and closed-loop validity checks) constitutes a dual defense against replay attacks and the identification of control failures. This multi-layered verification mechanism ensures that only drone entities with consistent statistical characteristics and reasonable physical behavior can pass authentication, thus achieving fast, robust, and highly secure physical entity identification even in complex weak network environments.

[0089] For example, such as Figure 2 The diagram shown is an architectural schematic of an identity authentication system (hereinafter referred to as the identity authentication system) based on UAV air-to-ground secure communication according to an embodiment of the present invention. The identity authentication system 10 includes: a ground control station 11 and a UAV 12, which are described below: I. Ground Control Station 11.

[0090] Ground control station 11 is a command and control unit deployed on the ground. It is responsible for establishing a communication link with UAV 12 and, when reconnecting after a communication interruption, performing authentication decisions on the physical entity.

[0091] For example, such as Figure 3 As shown, the ground control station 11 may include a response delay assessment module 111, a response baseline maintenance module 112, a current authentication receiving module 113, and a comparison and verification module 114, which will be described in turn below: (1) Response delay assessment module 111 is used to acquire historical flight data of UAV 12 and determine the response delay statistics of UAV 12 based on the historical flight data. For details of the implementation, please refer to S101 above.

[0092] (2) The response baseline maintenance module 112 is used to establish and update the historical response baseline of the UAV 12 based on the response delay statistics, and freeze the historical response baseline when a communication link interruption is detected. For details of the implementation, please refer to S201-S203 above.

[0093] (3) The current authentication receiving module 113 is used to obtain the current authentication feature sent by the UAV 12 when the communication link between the UAV 12 and the ground control station 11 is interrupted and reconnected. For details of the implementation, please refer to S103 above.

[0094] (4) The comparison and verification module 114 is used to perform statistical consistency comparison and physical rationality verification on the current authentication features based on the historical response benchmark, and determine whether the identity authentication of the UAV 12 is successful based on the results. For details of the implementation, please refer to S401-S404 above.

[0095] Optionally, if the reconnection request message sent by the UAV includes the track deviation sequence and control command sequence within the current sampling window, the ground control station 11 can extract the aforementioned two sequences and determine the current authentication feature. The specific process for determining the current authentication feature can be found in S301-S304 above.

[0096] II. Unmanned Aerial Vehicles (UAVs) 12.

[0097] The UAV 12 is an aerial mobile platform that performs flight missions. It is mainly responsible for sensing its own status in real time during flight and generating and reporting current authentication features to prove its physical identity when reconnecting with the ground control station 11.

[0098] Optionally, the UAV 12 is equipped with an onboard computing unit for extracting physical response features based on current flight status data to generate current authentication features, and sending a reconnection request containing the current authentication features to the ground control station 11 when the communication link with the ground control station 11 is interrupted and then reconnected. The process of generating the current authentication features is detailed in S301-S304 above.

[0099] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0100] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. An identity authentication method based on UAV air-to-ground secure communication, characterized in that, The method includes: Historical flight data of the UAV is acquired, and the response delay statistics of the UAV are determined based on the historical flight data; wherein, the response delay statistics are used to characterize the distribution characteristics of the time delay of the UAV in responding to disturbances within a historical time period. Based on the response delay statistics of the UAV, a historical response benchmark for the UAV is established and updated; wherein, the historical response benchmark is used to characterize the individual physical response characteristics of the UAV. When the communication link between the UAV and the ground control station is reconnected after being interrupted, the current authentication feature sent by the UAV is obtained; wherein, the current authentication feature is the physical response feature extracted by the UAV based on the current flight status data; The current authentication features are statistically compared based on the historical response benchmark, and the identity authentication of the drone is determined based on the comparison results.

2. The identity authentication method based on UAV air-to-ground secure communication according to claim 1, characterized in that, Based on the response delay statistics of the UAV, establish and update the historical response benchmark of the UAV, specifically including: During the communication connection with the UAV, the system continuously receives the response delay characteristics periodically reported by the UAV and maintains a historical observation window based on the response delay characteristics reported most recently. The response delay statistics are calculated based on all response delay characteristics within the historical observation window; wherein, the response delay statistics include the historical response delay mean and the historical response delay standard deviation; In response to the detection of a communication link interruption with the UAV, the updating of the historical observation window is stopped, and the response delay statistics at the time of interruption are used as the historical response benchmark for subsequent verification.

3. The identity authentication method based on UAV air-to-ground secure communication according to claim 1, characterized in that, Obtaining the current authentication features sent by the drone specifically includes: Receive the reconnection request message sent by the UAV during communication reconnection; If the reconnection request message includes the current authentication features, extract the current authentication features from the reconnection request message; or, If the reconnection request message includes a track deviation sequence and a control command sequence within the current sampling window, the current authentication feature is determined. The current authentication features include the current response latency feature and the effective correction percentage.

4. The identity authentication method based on UAV air-to-ground secure communication according to claim 1, characterized in that, The current response latency characteristic is determined according to the following steps: Obtain the track deviation sequence and control command sequence within the current sampling window; Extract the set of deviation peak points from the track deviation sequence, and extract the set of command peak points from the control command sequence; Perform timing matching between the set of deviation peak points and the set of instruction peak points to obtain at least one matching event pair; The current response delay feature is calculated based on the time difference of each event pair in the matched event pair.

5. The identity authentication method based on UAV air-to-ground secure communication according to claim 4, characterized in that, Timing matching of the deviation peak point set and the instruction peak point set specifically includes: Construct a bipartite graph with the set of deviation peak points as the first node set and the set of instruction peak points as the second node set; For nodes in the first node set and nodes in the second node set, if the time difference between the two nodes is within a preset reasonable time interval, a connection edge is established between the two nodes, and a weight is assigned to the connection edge according to the relationship between the corresponding amplitudes of the two nodes. In the bipartite graph, the matching result with the largest total weight is determined to obtain the matching event pair.

6. The identity authentication method based on UAV air-to-ground secure communication according to claim 2, characterized in that, A statistical consistency comparison of the current authentication features is performed based on the historical response benchmark, specifically including: Obtain the historical response delay mean and the historical response delay standard deviation from the historical response benchmark; Calculate the standardized statistical distance based on the current response delay characteristics, the historical response delay mean, and the historical response delay standard deviation; The standardized statistical distance is compared with a preset consistency threshold to obtain the comparison result.

7. The identity authentication method based on UAV air-to-ground secure communication according to claim 6, characterized in that, The determination of whether the drone's identity authentication is successful based on the comparison results specifically includes: Perform a physical validity check on the current authentication feature to obtain the check result; When both the comparison result and the verification result are passed, the identity authentication of the drone is determined to be successful.

8. The identity authentication method based on UAV air-to-ground secure communication according to claim 7, characterized in that, The physical validity of the current authentication feature is verified to obtain the verification result, which specifically includes: Verify whether the current response delay characteristic is within a preset physically reasonable time interval to obtain the first verification result; When the current authentication feature includes a valid correction ratio, the valid correction ratio is checked to see if it is greater than or equal to a preset valid threshold, and a second verification result is obtained; wherein, the valid correction ratio is calculated based on the number of matching event pairs and the number of elements in the deviation peak point set; When both the first verification result and the second verification result are passed, the verification result is determined to be passed.

9. The identity authentication method based on UAV air-to-ground secure communication according to any one of claims 1-8, characterized in that, The response latency statistics include at least the historical response latency mean and the historical response latency standard deviation; acquiring historical flight data of the UAV and determining the UAV's response latency statistics based on the historical flight data specifically includes: Based on the historical flight data, historical response delay characteristics corresponding to each of the multiple historical sampling periods are determined; wherein, the historical response delay characteristics are used to characterize the average time difference between the UAV being disturbed and generating a corrective action within the corresponding historical sampling period; Based on all the historical response delay characteristics, determine the historical response delay mean and the historical response delay standard deviation.

10. An identity authentication system based on UAV air-to-ground secure communication, characterized in that, The system includes: a ground control station and unmanned aerial vehicles (UAVs); The ground control station is configured to perform the method as described in any one of claims 1 to 9; The UAV is configured to extract physical response features based on current flight status data during flight to generate current authentication features, and to send a reconnection request containing the current authentication features to the ground control station when the communication link with the ground control station is interrupted and reconnected.

Images