LoRa terminal positioning method assisted by single unmanned aerial vehicle

Abstract

The invention relates to the technical field of unmanned aerial vehicle positioning, and discloses a single unmanned aerial vehicle assisted LoRa terminal positioning method, which comprises the following steps: receiving a signal sent by a LoRa terminal by using an unmanned aerial vehicle as a mobile base station, and extracting a received signal strength indication to establish a distance measurement model; carrying out recursive estimation on the terminal position by adopting extended Kalman filtering; for a fixed terminal scene, a precision evaluation model is established based on distance-dependent CRLB, and the path of the unmanned aerial vehicle is optimized by taking the maximum determinant of a Fisher information matrix as a criterion; for a mobile terminal scene, FIM is reconstructed, CRLB is optimized by using a Minkowski inequality, and path planning is carried out by taking the maximum increment determinant of the FIM as a criterion. According to the method, a large number of fixed base stations do not need to be deployed, and low-power-consumption and large-range high-precision positioning of the Internet of Things terminal is achieved within limited energy consumption and time.

Description

Technical Field

[0001] This invention relates to the field of drone positioning technology, and in particular to a LoRa terminal positioning method assisted by a single drone. Background Technology

[0002] With the development of IoT technology, the demand for positioning of large-scale low-power terminals is constantly increasing. Applications such as environmental monitoring, smart agriculture, logistics tracking, and animal behavior monitoring often require long-term, low-power positioning of a large number of terminal devices. However, existing positioning technologies still have certain limitations in large-scale low-power scenarios.

[0003] Currently, common positioning technologies mainly include Global Positioning System (GPS), Radio Frequency Identification (RFID), and wireless signal positioning. GPS positioning has high accuracy, but the terminal power consumption is high, resulting in high equipment cost, making it difficult to meet the long-term operation requirements of low-power IoT terminals. RFID technology is generally suitable for short-range positioning, with a limited positioning range, making it difficult to apply in large-scale outdoor scenarios.

[0004] In recent years, long-range radio (LoRa) communication technology based on Low Power Wide Area Network (LPWAN) has been increasingly applied to IoT positioning scenarios. LoRa features long communication distance, low power consumption, and low cost, making it suitable for communication and data acquisition across large numbers of terminal devices. By analyzing the Received Signal Strength Indicator (RSSI) of LoRa signals, the distance between the terminal and the receiving node can be estimated, thereby enabling terminal positioning.

[0005] In existing LoRa positioning research, multiple fixed base stations are typically deployed for positioning. However, in remote areas, complex environments, or temporary monitoring scenarios, the deployment cost of a large number of fixed base stations is high, and it also increases the difficulty of system maintenance. To reduce the deployment of fixed infrastructure, existing technologies have proposed using unmanned aerial vehicles (UAVs) as mobile base stations for positioning. UAVs are highly mobile and flexible in deployment, and can collect terminal signals over a large area to achieve positioning. However, in a single UAV positioning system, flight path design has a significant impact on positioning accuracy. If the UAV flight path design is unreasonable, it will lead to insufficient measurement information, thereby reducing positioning accuracy.

[0006] Furthermore, in practical applications, the positioning terminal may be a stationary target or a moving target. For example, in a pasture monitoring scenario, it is necessary to locate moving livestock. Therefore, how to design effective drone flight paths under different terminal motion states to improve positioning accuracy is an important problem that needs to be solved.

[0007] Therefore, existing technologies still need to be improved and developed. Summary of the Invention

[0008] The main objective of this invention is to provide a LoRa terminal positioning method assisted by a single UAV, which aims to solve the problem that in existing single UAV LoRa positioning systems, the flight path lacks adaptive optimization for the terminal's motion state, making it difficult to meet the high-precision positioning requirements of both stationary and mobile terminals within a limited flight time.

[0009] To achieve the above objectives, the present invention provides a single UAV-assisted LoRa terminal positioning method, which includes the following steps: Build a LoRa network and complete the design of LoRa terminal nodes and LoRa gateways; A LoRa ranging model is established using an RSSI-based ranging method, and a matching LoRa signal propagation model is selected. During flight, the drone equipped with a LoRa gateway receives signal packets sent by the LoRa terminal, parses the signal packets and extracts RSSI information. At the same time, the drone obtains its current three-dimensional coordinate position information through its onboard positioning module. For fixed LoRa terminal scenarios or mobile LoRa terminal scenarios, an EKF-based LoRa terminal location information acquisition method is used to obtain the terminal location; For fixed LoRa terminal scenarios, the flight path is optimized using a single UAV path optimization method based on CRLB. For mobile LoRa terminal scenarios, the flight path is optimized using a single UAV path optimization method based on FIM increment and Minkowski inequality. The drone flies according to the optimized flight path and determines in real time whether the return-to-home conditions are met during the flight. If the return-to-home conditions are met, the drone is controlled to return to the starting position along the return-to-home path. During the return-to-home process, the drone continues to acquire the current position information of the drone, receive LoRa signals and perform terminal positioning. If the return-to-home conditions are not met, the drone continues to fly according to the optimized path. Determine whether the total positioning time has been reached. If the total positioning time has not been reached and the drone has not entered the return-to-home state, continue to acquire the drone's current location information, receive LoRa signals, and perform terminal positioning and path optimization. If the total positioning time has been reached or the drone has completed its return-to-home process, output the flight path and positioning results.

[0010] Optionally, in the single-UAV-assisted LoRa terminal positioning method, the LoRa network includes: a terminal device containing a LoRa chip, a LoRa base station, and a server; the LoRa terminal node includes: a LoRa module, a sensor, and an MCU microcontroller; the LoRa gateway includes: a LoRa module and an MCU microcontroller; the LoRa gateway is mounted on a UAV, and the UAV serves as a mobile base station for LoRa positioning.

[0011] Optionally, the single UAV-assisted LoRa terminal positioning method, wherein the step of establishing a LoRa ranging model using an RSSI-based ranging method specifically includes: A positioning model is established using an RSSI-based ranging method. In free space, the relationship between signal strength and propagation distance is expressed as: ; in, For received power, For transmission power, For transmitter antenna gain, For receiver antenna gain, The wavelength of the transmitted signal, d This indicates the distance between the receiver and the transmitter. The LoRa signal propagation process is described using a log-normal shading model, and the LoRa ranging model is as follows: ; in, Indicates the first j The distance between each receiving point and the transmitting source This indicates that the distance from the source is The received signal power, This represents the average path loss at the reference point. For reference distance, This is the path loss factor. This represents the random items caused by the shadowing effect.

[0012] Optionally, in the single-UAV-assisted LoRa terminal positioning method, for fixed LoRa terminal scenarios, the step of obtaining the terminal location using an EKF-based LoRa terminal location information acquisition method specifically includes: Using the coordinates of unknown nodes as the state vector of the positioning system: ; in, To locate the state vector of the system, For node coordinates, T Indicates transpose; The state equation of a fixed LoRa terminal is: ; in, Indicates the number of times the location was located. Indicates the first The state vector of the next location. Indicates the first +1 positioning state vector Here is the state transition matrix. This is process noise; Using the RSSI received by the LoRa terminal node as the observation value, the measurement equation is: ; in, For the first RSSI measurement value of the second positioning. The nonlinear measurement function is composed of RSSI and terminal position. For measuring noise; The measurement equation is linearized using EKF, and the location estimate of the fixed LoRa terminal is obtained through a prediction and update process.

[0013] Optionally, the single-UAV-assisted LoRa terminal positioning method, for mobile LoRa terminal scenarios, specifically includes obtaining the terminal location using an EKF-based LoRa terminal location information acquisition method, which includes: The state vector of the node to be located is: ; in, For node coordinates, For node speed, The direction of movement of the LoRa terminal node. T This indicates transpose.

[0014] The motion of the mobile LoRa terminal is described using a random walk model. The node motion model of the mobile LoRa terminal is as follows: ; in, , and Indicates the current position. , and Indicates the position at the next moment. This represents the time interval between the current moment and the next moment. Indicates the velocity at the next moment. Indicates the direction angle of motion at the next moment. Random noise representing changes in drive speed. Random noise representing changes in the driving direction angle; Using RSSI as the observation and employing the EKF recursive process, the location estimate of the mobile LoRa terminal is obtained.

[0015] Optionally, in the single UAV-assisted LoRa terminal positioning method, if the scenario is a fixed LoRa terminal, a single UAV path optimization method based on CRLB is used to optimize the flight path, specifically including: The estimated coordinate covariance matrix of the node to be located satisfies: ; in, This represents the covariance matrix of the estimated location of the node to be located. Indicates the flight path of the UAV Related FIM; The D-optimal criterion is used for path optimization, and the objective function for path optimization is: ; in, Indicates that the UAV is in the The direction angle at that moment, Indicates that the UAV is in the The pitch angle at any moment, This represents the FIM (First Instruction Metric) at the next moment after the UAV moves according to the candidate flight direction. This represents matrix determinant operations.

[0016] Optionally, in the single UAV-assisted LoRa terminal positioning method, if the scenario is a mobile LoRa terminal, the flight path is optimized using a single UAV path optimization method based on the FIM incremental method and the Minkowski inequality, specifically including: Reconstruct the FIM and calculate the CRLB; Based on the Minkowski inequality for positive definite matrices, the path optimization problem is transformed into optimizing the FIM increment at the next time step. The objective function for path optimization is: ; in, Indicates that the UAV is in the The direction angle at that moment, Indicates that the UAV is in the The pitch angle at any moment, This represents the FIM increment brought about by the UAV flying from the current moment to the next moment. Indicates the flight path, This represents matrix determinant operations.

[0017] Optionally, in the single-UAV-assisted LoRa terminal positioning method, the objective function needs to satisfy the following constraints during the path optimization process: Starting position constraints: ; in, This is the initial position of the UAV; Termination position constraint: ; in, This is the termination point for the UAV mission; Return-to-home constraints: ; in, Indicates the current position of the UAV. Indicates the flight speed of the UAV. Indicates the total task duration. Indicates the single-step flight time. and These represent the number of positioning attempts and the number of times the LoRa terminal node transmitted signals before the UAV achieved positioning, respectively. Lower limit constraints on flight altitude: ; Flight altitude limit constraints: ; in, For UAV in the Flight altitude at any given moment and These are the minimum and maximum permitted flight altitudes, respectively. Pitch angle variation constraints: ; in, This represents the maximum allowable pitch angle change between adjacent moments for the UAV. Azimuth variation constraints: ; in, This represents the maximum allowable change in azimuth angle between adjacent moments of the UAV.

[0018] Optionally, the single-UAV-assisted LoRa terminal positioning method is applicable to multi-LoRa terminal scenarios, which include fixed multi-LoRa terminal scenarios and mobile multi-LoRa terminal scenarios.

[0019] Optionally, in the single-UAV-assisted LoRa terminal positioning method, for a fixed multi-LoRa terminal scenario, the overall FIM is constructed as a block diagonal matrix, and the path optimization objective is to maximize the product of the determinants of the FIMs of all terminals; for a mobile multi-LoRa terminal scenario, the cumulative FIM increment of multiple targets is calculated, and the path optimization objective is to maximize the product of the determinants of the FIM increments of multiple targets.

[0020] In this invention, a LoRa network is constructed, and the design of LoRa terminal nodes and LoRa gateways is completed. A LoRa ranging model is established using an RSSI-based ranging method, and a matching LoRa signal propagation model is selected. During flight, a UAV equipped with a LoRa gateway receives signal packets sent by the LoRa terminal, parses the signal packets, and extracts RSSI information. Simultaneously, the UAV obtains its current three-dimensional coordinate position information through its onboard positioning module. For fixed or mobile LoRa terminal scenarios, an EKF-based LoRa terminal position information acquisition method is used to obtain the terminal position. For fixed LoRa terminal scenarios, a CRLB-based single UAV path optimization method is used to optimize the flight path; for mobile LoRa terminal scenarios… This invention employs a single UAV path optimization method based on the fundamental FIM incremental method and Minkowski inequality to optimize the flight path. The UAV flies according to the optimized flight path, and during flight, it determines in real time whether the return-to-home conditions are met. If the return-to-home conditions are met, the UAV is controlled to return to the starting position along the return-to-home path, and during the return-to-home process, it continues to acquire the UAV's current position information, receive LoRa signals, and perform terminal positioning. If the return-to-home conditions are not met, the UAV continues to fly along the optimized path. The invention then determines whether the total positioning time has been reached. If the total positioning time has not been reached and the UAV has not entered the return-to-home state, it continues to acquire the UAV's current position information, receive LoRa signals, and perform terminal positioning and path optimization. If the total positioning time has been reached or the UAV has completed its return-to-home process, the flight path and positioning result are output. This invention eliminates the need to deploy a large number of fixed base stations, achieving low-power, wide-range, high-precision positioning of IoT terminals within limited energy consumption and time. Attached Figure Description

[0021] Figure 1 This is a flowchart of a preferred embodiment of the LoRa terminal positioning method assisted by a single UAV of the present invention; Figure 2This is a LoRa network architecture diagram in a preferred embodiment of the LoRa terminal positioning method assisted by a single UAV of the present invention; Figure 3 This is a schematic diagram of a single UAV-assisted LoRa terminal positioning model in a preferred embodiment of the single UAV-assisted LoRa terminal positioning method of the present invention; Figure 4 This is a schematic diagram of the structure of a LoRa terminal node in a preferred embodiment of the LoRa terminal positioning method assisted by a single UAV of the present invention; Figure 5 This is a schematic diagram of the LoRa gateway structure in a preferred embodiment of the LoRa terminal positioning method assisted by a single UAV of the present invention; Figure 6 This is a flowchart of the LoRa terminal location information acquisition algorithm based on EKF in a preferred embodiment of the LoRa terminal positioning method assisted by a single UAV of the present invention. Figure 7 This is a flowchart illustrating the overall process of the single-UAV-assisted LoRa terminal positioning method in a preferred embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0023] This invention proposes a single UAV-assisted LoRa terminal positioning method for achieving high-precision terminal location estimation in large-scale low-power IoT scenarios. Specifically, the method utilizes a UAV as a mobile base station, receiving signals transmitted by the LoRa terminal during flight, and establishing a distance measurement model between the terminal and the UAV using RSSI. Based on this, the extended Kalman filter (EKF) algorithm is used to recursively estimate the terminal's position, thereby achieving three-dimensional positioning of the terminal. For different application scenarios, this invention considers both fixed and mobile terminals. (1) In fixed terminal scenarios, a positioning accuracy evaluation model is established based on the distance-dependent Cramér-Rao Lower Bound (CRLB), and the flight path of the UAV is optimized accordingly, thereby improving the positioning accuracy under limited flight time and energy consumption conditions.

[0024] (2) In the mobile terminal scenario, by constructing the Fisher Information Matrix (FIM) and calculating the distance dependency CRLB, the CRLB expression is optimized using the Minkowski inequality, and the path planning of the UAV is carried out with the maximization of the incremental determinant of the FIM as the criterion, thereby further improving the positioning accuracy of the mobile terminal.

[0025] Compared with existing technologies, this invention combines LoRa communication technology with a single UAV mobile base station and uses the theoretical lower bound of positioning accuracy to guide UAV path optimization, achieving efficient positioning of low-power, wide-range IoT terminals without the need to deploy a large number of fixed base stations.

[0026] The preferred embodiment of the LoRa terminal positioning method assisted by a single unmanned aerial vehicle (UAV) of the present invention, such as... Figure 1 As shown, the single UAV-assisted LoRa terminal positioning method includes the following steps: Step S10: Build a LoRa network and complete the design of LoRa terminal nodes and LoRa gateways.

[0027] Specifically, the single-UAV-assisted LoRa terminal positioning model constructed in this embodiment consists of one unmanned aerial vehicle (UAV) and multiple LoRa terminal nodes, belonging to a three-dimensional positioning scenario. As shown in Figures 2 to 5, this invention first constructs a LoRa network architecture and completes the design of LoRa terminal nodes and LoRa gateways. Based on this, a LoRa ranging model is established and a LoRa signal propagation model is selected. A LoRa network generally consists of terminal devices containing LoRa chips, LoRa base stations, and servers; the single-UAV-assisted LoRa terminal positioning model constructed in this invention consists of one UAV and multiple LoRa terminal nodes, belonging to a three-dimensional positioning scenario. A single UAV equipped with a LoRa gateway is used as the mobile base station for LoRa positioning, without setting a fixed hovering point, and the UAV's flight path is optimized.

[0028] In terms of hardware implementation, a LoRa terminal node consists of a LoRa module, a sensor, and an MCU microcontroller. The LoRa module can be the SX1278 RF chip, the sensor can be the DHT11 temperature and humidity sensor, and the MCU microcontroller can be the Arduino Uno R3. A LoRa gateway can be constructed from a LoRa module and an MCU microcontroller and mounted on a drone. The drone platform can be a model with open development interfaces, such as the DJI Matrice M210.

[0029] Step S20: Establish a LoRa ranging model using an RSSI-based ranging method and select a matching LoRa signal propagation model.

[0030] Specifically, in the LoRa terminal positioning process, a positioning model is established using an RSSI-based ranging method. In free space, the relationship between signal strength and propagation distance can be expressed by Friis's equation, as shown in formula (1): (1) in, For received power, For transmission power, For transmitter antenna gain, For receiver antenna gain, The wavelength of the transmitted signal, d This indicates the distance between the receiver and the transmitter.

[0031] Since formula (1) is an idealized propagation model, and the signal propagation process in the actual environment is affected by shadow fading and environmental noise, this invention uses a log-normal shadow model to describe the LoRa signal propagation process. The LoRa ranging model is shown in formula (2): (2) in, Indicates the first j The distance between each receiving point and the transmitting source This indicates that the distance from the source is The received signal power, This represents the average path loss at the reference point. For reference distance, This is the path loss factor. This represents the random items caused by the shadowing effect.

[0032] As can be seen from formula (2), the distance between the LoRa terminal and the UAV can be estimated based on the received RSSI information, thus providing a basis for subsequent location information acquisition.

[0033] Step S30: During flight, the drone equipped with the LoRa gateway receives signal packets sent by the LoRa terminal, parses the signal packets and extracts RSSI information. At the same time, the drone obtains its current three-dimensional coordinate position information through its onboard positioning module.

[0034] Specifically, the LoRa terminal node periodically sends data to the UAV. In a fixed terminal scenario, the LoRa terminal node can be configured to send data every 20 seconds; in a mobile terminal scenario, it can be configured to send data every 1 second. During flight, the UAV receives signal packets sent by the LoRa terminal, parses the data packets, and extracts the RSSI information. Simultaneously, the UAV obtains its current three-dimensional coordinate position information through its onboard positioning module.

[0035] Step S40: For fixed LoRa terminal scenarios or mobile LoRa terminal scenarios, the terminal location is obtained using the EKF-based LoRa terminal location information acquisition method.

[0036] Specifically, in this embodiment, different state vectors are selected for EKF estimation based on the motion state of the terminal: (1) For fixed LoRa terminal scenarios, the coordinates of unknown nodes are used as the state vector of the positioning system, i.e. ,in, To locate the state vector of the system, For node coordinates, T This indicates transpose. Since the position of the fixed LoRa terminal does not change between adjacent positioning times, its state equation can be expressed as formula (3): (3) in, Indicates the number of times the location was located. Indicates the first The state vector of the next location. Indicates the first +1 positioning state vector Here is the state transition matrix. This is process noise.

[0037] Using the RSSI received by the LoRa terminal node as the observation value, the measurement equation can be expressed as formula (4): (4) in, For the first RSSI measurement value of the second positioning. The nonlinear measurement function is composed of RSSI and terminal position. For measuring noise.

[0038] Since the measurement equation in formula (4) is nonlinear, EKF is used to linearize it, and the location estimate of the fixed LoRa terminal is obtained through the prediction and update process.

[0039] (2) For mobile LoRa terminal scenarios, the state vector of the node to be located is ,in, For node coordinates, For node speed, The direction of movement of the LoRa terminal node. T This indicates transpose. A random walk model conforming to the characteristics of herd movement is used to describe the motion of the mobile LoRa terminal, and its node motion model is shown in formula (5): (5) in, , and Indicates the current position. , and Indicates the position at the next moment. This represents the time interval between the current moment and the next moment. Indicates the velocity at the next moment. Indicates the direction angle of motion at the next moment. Random noise representing changes in drive speed. This represents random noise indicating the change in the driving direction angle; where, , .

[0040] In mobile scenarios, RSSI is still used as the observation, and the EKF algorithm recursive process is used to perform dynamic position estimation for mobile LoRa terminals.

[0041] Step S50: If it is a fixed LoRa terminal scenario, the flight path is optimized using a single UAV path optimization method based on CRLB. If it is a mobile LoRa terminal scenario, the flight path is optimized using a single UAV path optimization method based on FIM increment and Minkowski inequality.

[0042] Specifically, this embodiment minimizes the positioning error by optimizing the UAV flight path: (1) In the fixed LoRa terminal scenario, the estimated coordinate covariance matrix of the node to be located is greater than or equal to the inverse matrix of FIM, that is, the CRLB relationship is as shown in formula (6): (6) in, This represents the covariance matrix of the estimated location of the node to be located. Indicates the flight path of the UAV Related FIM.

[0043] The D-optimal criterion is used for path optimization, and the objective function of path optimization can be expressed as formula (7): (7) in, Indicates that the UAV is in the The direction angle at that moment, Indicates that the UAV is in the The pitch angle at any moment, This represents the FIM (First Instruction Metric) at the next moment after the UAV moves according to the candidate flight direction. This represents matrix determinant operations.

[0044] (2) In the scenario of mobile LoRa terminals, since a single base station cannot acquire multiple valid information at the same time, and historical RSSI values ​​cannot all be used for the positioning of the current mobile point, this embodiment reconstructs the FIM and, based on the Minkowski inequality of positive definite matrices, transforms the path optimization problem into an optimization problem of the FIM increment at the next time step. The objective function for path optimization is: (8) in, Indicates that the UAV is in the The direction angle at that moment, Indicates that the UAV is in the The pitch angle at any moment, This represents the FIM increment brought about by the UAV flying from the current moment to the next moment. Indicates the flight path, This represents matrix determinant operations.

[0045] Step S60: The UAV flies according to the optimized flight path and determines in real time whether the return-to-home conditions are met during the flight. If the return-to-home conditions are met, the UAV is controlled to return to the starting position along the return-to-home path. During the return-to-home process, the UAV continues to acquire the current position information of the UAV, receive LoRa signals and perform terminal positioning. If the return-to-home conditions are not met, the UAV continues to fly according to the optimized path.

[0046] Specifically, during the path optimization process, the objective function must satisfy the following constraints: (1) Starting position constraint: (9) in, This is the initial position of the UAV.

[0047] (2) Termination position constraint: (10) in, This is the termination point for the UAV mission.

[0048] (3) Return-to-home constraints: (11) in, Indicates the current position of the UAV. Indicates the flight speed of the UAV. Indicates the total task duration. Indicates the single-step flight time. and These represent the number of positioning attempts and the number of LoRa terminal node signals transmitted before the UAV achieves positioning, respectively. This constraint is used to ensure that the UAV has sufficient remaining energy at the current moment to safely return to the starting point.

[0049] (4) Lower limit constraint of flight altitude: (12) (5) Maximum flight altitude constraint: (13) in, For UAV in the Flight altitude at any given moment and These are the minimum and maximum altitudes that are allowed to fly, respectively.

[0050] (6) Pitch angle variation constraint: (14) in, This represents the maximum allowable pitch angle change between adjacent moments for the UAV.

[0051] (7) Azimuth variation constraints: (15) in, This represents the maximum allowable change in azimuth angle between adjacent moments of the UAV, used to limit the maximum rotation rate of the UAV.

[0052] If the return-to-home conditions are met, the UAV is controlled to return to the starting position along the return-to-home path. During the UAV's return-to-home process, steps S30 and S40 are executed again, but the path optimization in step S50 is not performed to reduce the computational burden.

[0053] Step S70: Determine whether the total positioning time has been reached. If the total positioning time has not been reached and the drone has not entered the return-to-home state, continue to acquire the drone's current location information, receive LoRa signals, and perform terminal positioning and path optimization. If the total positioning time has been reached or the drone has completed its return-to-home process, output the flight path and positioning results.

[0054] Specifically, determine whether the current positioning time has reached the total positioning time T. If the total positioning time has not been reached and the UAV has not entered the return-to-home state, return to step S30 to continue acquiring the UAV's current location information, receiving LoRa signals, and performing terminal positioning and path optimization; if the total positioning time has been reached or the UAV has completed its return-to-home process, output the flight path and positioning results, and the process ends.

[0055] like Figure 6 The diagram illustrates a typical extended Kalman filter (EKF) iteration cycle. Since the relationship between RSSI and distance is non-linear, EKF is used to linearize it, and a two-stage "prediction-update" iterative process is employed to continuously approximate the true position of the terminal, even in the presence of measurement noise. The specific steps are as follows: (1) Start: Start the positioning calculation process for the current moment.

[0056] (2) EKF initialization.

[0057] (3) The UAV obtains its own position: obtains the current position of the UAV in the air, which is the reference point for calculating the distance between the terminal and the UAV.

[0058] (4) Receiving LoRa terminal signals: The LoRa gateway on the UAV receives LoRa wireless signals sent from the ground terminal to be located at this location.

[0059] (5) Extract RSSI signal strength: Extract the key measurement value - Received Signal Strength Indicator (RSSI) from the received physical layer data.

[0060] (6) Calculate distance based on RSSI ranging model: Using the log-normal shading model, the extracted RSSI values ​​are converted into the estimated distance between the terminal and the UAV.

[0061] (7) EKF prediction stage: One-step state prediction: Based on the system's state equations, predict the current state from the state at the previous time step; Covariance-based one-step prediction: Calculate the covariance matrix of the predicted state to reflect the uncertainty brought about by the prediction step.

[0062] (8) EKF Measurement Update Phase: Calculate the Kalman gain: Based on the prediction covariance and measurement noise, calculate the Kalman gain. The gain determines whether the filtering algorithm places more trust in the model prediction or the current measurement when correcting the estimate; State update: Based on the difference between the calculated actual measured distance and the predicted distance, the predicted state is corrected to obtain the optimal state estimate (i.e., terminal position) at the current moment. Covariance Update: Updates the state covariance matrix for prediction at the next time step.

[0063] (9) Determine whether the positioning accuracy or positioning time requirements have been met: Check whether the current positioning result meets the preset accuracy threshold, or whether the positioning time limit assigned to this terminal or this stage has been reached; if "No": it means that new measurement values ​​need to be acquired to improve accuracy or track moving targets, the process returns to the "UAV acquires its own position" step, the UAV moves to the next measurement point (or the next moment arrives), and a new EKF loop begins; If "Yes": The termination condition is met, proceed to the next step.

[0064] (10) Output terminal position estimation: The optimal state estimate obtained after the EKF update at the current time is output as the final (or current time) position information of the terminal.

[0065] (11) End: The EKF calculation process for this terminal or this positioning period ends.

[0066] Figure 6 The process of LoRa terminal location information acquisition based on extended Kalman filtering is demonstrated: In each positioning cycle, the UAV first obtains its own position and receives terminal signals to extract RSSI values. Then, the distance is calculated through the RSSI ranging model. Next, under the EKF framework, state prediction (based on motion model) and measurement update (based on RSSI observations) are performed alternately. Measurement noise is gradually filtered out through continuous iterative recursion, and finally, high-precision estimation of the position of fixed or mobile LoRa terminals is achieved.

[0067] like Figure 7 The diagram illustrates the complete process of the single-UAV-assisted LoRa terminal positioning method proposed in this invention. The core of this method lies in selecting different path optimization strategies based on the terminal's motion state (fixed or moving). The process is as follows: Step 1: After starting, the system initializes; Set the initial position of the drone, the total mission duration, the flight speed, and various constraints (such as flight altitude range, maximum turn rate, return-to-home capability, etc.).

[0068] Step 2: Obtain the current location information of the UAV, receive the LoRa signal, and extract the RSSI information; The drone hovers or passes over its current flight position to obtain its own spatial coordinates; the onboard LoRa gateway receives signals sent from ground LoRa terminal nodes; and extracts the Received Signal Strength Indicator (RSSI) value from the received signals.

[0069] Step 3: Estimate the terminal location based on EKF; Using the RSSI information obtained in step two, combined with a pre-established signal propagation model (such as a log-normal shadowing model), the position of the terminal is recursively estimated using the extended Kalman filter (EKF) algorithm to obtain the terminal position information at the current moment.

[0070] Step 4: Determine the terminal's motion mode; The system determines whether the currently located terminal is a fixed terminal or a mobile terminal based on preset or real-time identification results.

[0071] Step 5: Path optimization strategy selection; Fixed terminal mode: Constructing a positioning accuracy evaluation model: Based on the Fisher Information Matrix (FIM), a Cramer-Rao Lower Bound (CRLB) model is constructed as a theoretical lower limit evaluation standard for positioning accuracy.

[0072] Path optimization: The D-optimal criterion is adopted, which selects the direction from the candidate flight directions that maximizes the FIM determinant at the next moment, thereby minimizing the positioning error.

[0073] Mobile terminal mode: Predicting the terminal's next position: Based on the motion model (such as a random walk model), predict the terminal's possible position at the next moment.

[0074] Construct a positioning accuracy evaluation model: Reconstruct the FIM suitable for mobile scenarios and optimize it using the Minkowski inequality.

[0075] Path optimization: Select the direction from the candidate flight directions that maximizes the determinant of the FIM increment at the next moment, and avoid using expired historical data.

[0076] Step 6: Determine if flight constraints are met; Check whether the optimized flight path (i.e., the flight direction, pitch angle, etc. at the next moment) meets all physical and mission constraints, such as: upper and lower limits of flight altitude, pitch angle change limit, azimuth (turn rate) change limit, etc.

[0077] If not satisfied: Return to step five, perform path optimization again, and select the next best candidate direction.

[0078] If satisfied: Proceed to the next step.

[0079] Step 7: Determine if the conditions for returning to base are met; Based on the remaining energy (or total mission duration) and the current location, determine whether the drone needs to return immediately to ensure a safe return to the starting point according to the return-to-home constraint formula.

[0080] If the return-to-home conditions are met: the drone will no longer fly along the optimized path, but will directly return to the starting position along the shortest path.

[0081] If the return-to-home conditions are not met: the drone continues to fly to the next measurement point according to the path optimized in step five and verified in step six.

[0082] Step 8: Determine if the positioning time has been reached; Check whether the preset total task duration or number of location attempts has been reached.

[0083] If the target is not achieved: the process returns to step two, the drone continues to receive signals, estimate its position, and optimize its path again at the new location, forming a closed-loop iteration.

[0084] If the following conditions are met: terminate the flight mission, output the optimized UAV flight path obtained throughout the process and the positioning results of all terminals, and the process ends.

[0085] Figure 7 The core idea is a closed-loop iterative optimization process. The UAV executes a loop of "signal acquisition - position estimation - mode determination - path planning" at each measurement point. By introducing the theoretical guidance of the base FIM / CRLB and dynamically adjusting the optimization target based on whether the terminal is moving, the UAV can autonomously plan a flight path that maximizes the efficiency of acquiring positioning information within a limited flight time, ultimately achieving high-precision terminal positioning.

[0086] Furthermore, the single-UAV-assisted LoRa terminal positioning method of the present invention is applicable to multi-LoRa terminal scenarios, including fixed multi-LoRa terminal scenarios and mobile multi-LoRa terminal scenarios. For fixed multi-LoRa terminal scenarios, the overall FIM is constructed as a block diagonal matrix, and the path optimization objective is to maximize the product of the determinants of the FIMs of all terminals; for mobile multi-LoRa terminal scenarios, the cumulative FIM increment of multiple targets is calculated, and the path optimization objective is to maximize the product of the determinants of the FIM increments of multiple targets.

[0087] This invention combines the long-range, low-power, and low-cost characteristics of LoRa technology with the maneuverability and mobile deployment of UAVs to achieve the positioning of LoRa terminals. Furthermore, it designs corresponding location information acquisition methods and UAV path optimization methods for two scenarios: fixed LoRa terminals and mobile LoRa terminals, respectively, to improve positioning accuracy under conditions of limited energy consumption and limited positioning time.

[0088] The technical effects that this invention can bring are as follows: (1) Achieving low-cost and low-power positioning: This invention utilizes the propagation characteristics of LoRa signals to establish a ranging model and adopts an RSSI-based ranging method. The distance information between the terminal and the UAV can be obtained without the need for the terminal to be positioned to be equipped with an additional GPS module, thereby reducing the system hardware cost and terminal power consumption.

[0089] (2) Improve the efficiency of positioning solution: This invention constructs a three-dimensional state vector for fixed LoRa terminals and uses EKF to obtain location information. Compared with the traditional maximum likelihood estimation method, it can effectively reduce the computational complexity, improve the positioning solution efficiency, and reduce the impact of measurement noise on the positioning results.

[0090] (3) Enhanced adaptability to moving targets: This invention constructs a five-dimensional state vector containing position, speed and direction for mobile LoRa terminals, and adopts a dynamic position information acquisition method based on EKF to realize continuous position updates of moving targets, thereby improving the applicability of this invention in dynamic scenarios such as livestock tracking.

[0091] (4) Improve positioning accuracy in fixed scenarios: In fixed LoRa terminal scenarios, the present invention adopts a single UAV path optimization method based on distance-dependent CRLB, and comprehensively considers constraints such as starting position, return home, altitude and turn rate, which can improve positioning accuracy and reduce invalid flight paths under limited time and limited energy consumption conditions.

[0092] (5) Improve the rationality of path decision-making and positioning accuracy in mobile scenarios: This invention reconstructs the FIM in mobile LoRa terminal scenarios and adopts a path optimization method based on the maximum incremental determinant of FIM, avoiding the direct use of no longer valid historical RSSI measurement information for current positioning, thereby improving the rationality of path planning and positioning accuracy in mobile scenarios.

[0093] In summary, this invention provides a single UAV-assisted LoRa terminal positioning method. The method includes: constructing a LoRa network and designing LoRa terminal nodes and a LoRa gateway; establishing a LoRa ranging model using an RSSI-based ranging method and selecting a matching LoRa signal propagation model; the UAV equipped with the LoRa gateway receiving signal packets sent by the LoRa terminal during flight, parsing the signal packets and extracting RSSI information; simultaneously, the UAV obtaining its current three-dimensional coordinate position information through its onboard positioning module; for fixed or mobile LoRa terminal scenarios, obtaining the terminal position using an EKF-based LoRa terminal position information acquisition method; and for fixed LoRa terminal scenarios, optimizing the flight path using a CRLB-based single UAV path optimization method. For mobile LoRa terminal scenarios, the flight path is optimized using a single UAV path optimization method based on the base FIM incremental method and the Minkowski inequality. The UAV flies according to the optimized flight path and determines in real time whether the return-to-home conditions are met. If the conditions are met, the UAV is controlled to return to the starting position along the return-to-home path, and continues to acquire the UAV's current position information, receive LoRa signals, and perform terminal positioning during the return-to-home process. If the conditions are not met, the UAV continues to fly along the optimized path. The system then checks whether the total positioning time has been reached. If the total positioning time has not been reached and the UAV has not entered the return-to-home state, the system continues to acquire the UAV's current position information, receive LoRa signals, and perform terminal positioning and path optimization. If the total positioning time has been reached or the UAV has completed its return-to-home process, the flight path and positioning result are output. This invention eliminates the need to deploy a large number of fixed base stations, achieving high-precision positioning of low-power, wide-range IoT terminals within limited energy consumption and time.

[0094] Without departing from the core concept of this invention, those skilled in the art can make several substitutions or adjustments to this invention, and all such substitutions or adjustments should fall within the protection scope of this invention. For example, in some embodiments, flight constraint parameters can be adjusted according to specific application scenarios during path optimization, including but not limited to parameters such as flight altitude range, maximum turn rate, return-to-home threshold, and positioning time interval; in some embodiments, the motion model of the mobile terminal can be adjusted according to the actual target motion characteristics; in some embodiments, the parameters of the LoRa signal propagation model can be calibrated according to different deployment environments to improve RSSI ranging accuracy and positioning accuracy. For example: (1) Regarding the ranging method: This invention employs RSSI ranging to estimate the distance between the terminal and the UAV. In other embodiments, it can also combine measurement methods based on Time of Arrival (TOA), Time Difference of Arrival (TDOA), and Angle of Arrival (AOA) to improve positioning accuracy.

[0095] (2) Regarding positioning algorithms: This invention uses EKF for terminal location estimation. In other embodiments, nonlinear filtering algorithms such as Unscented Kalman Filter (UKF) and Particle Filtering (PF) can also be used to achieve localization.

[0096] (3) Regarding the number of drones: This invention primarily focuses on single-UAV applications. In other embodiments, it can be extended to a multi-UAV cooperative positioning system, improving system positioning efficiency by collaboratively collecting terminal signals from multiple UAVs.

[0097] (4) Regarding path planning methods: This invention optimizes paths based on CRLB and FIM. In other embodiments, heuristic algorithms, reinforcement learning methods, or other optimization algorithms can also be combined to achieve UAV path planning.

[0098] (5) Regarding the hardware platform: In this embodiment of the invention, a drone is used as a mobile base station equipped with a LoRa communication module to receive signal data sent by the terminal node. In other embodiments, the mobile platform can also be other forms of mobile carriers, such as unmanned ground vehicles (UGVs), unmanned surface vehicles (USVs), or other mobile robot platforms. These mobile platforms can also move within the target area and collect terminal signals, thereby achieving terminal positioning.

[0099] (6) In terms of actual deployment: In practical engineering deployments, electronic map information can be used to constrain the flight path of drones to avoid no-fly zones or obstacle areas. By incorporating obstacle avoidance constraints into the drone path optimization model, the safety of drone flight and the practicality of path planning can be improved.

[0100] (7) Regarding the extension of path planning algorithms: Based on the method of this invention, machine learning methods, such as reinforcement learning algorithms, can be introduced to train the UAV flight path strategy, enabling the UAV to autonomously adjust its flight path in complex or unknown dynamic environments, thereby further improving the adaptability of path planning and positioning efficiency.

[0101] (8) Regarding communication timing optimization: Considering the asynchronous communication characteristics of LoRa communication, in practical applications, the synchronization mechanism between data reception time slots and UAV flight positions can be further optimized. By rationally designing communication scheduling strategies, the success rate of UAV receiving terminal signals can be improved, thereby enhancing the overall efficiency of the positioning system.

[0102] (9) Regarding system deployment methods: The positioning and path optimization algorithms of this invention have good hardware adaptability and can be deployed on the flight control modules of various small UAVs, such as embedded platforms like STM32 and FPGA. They can also be deployed on the ground control terminal to send path planning instructions to the UAV via wireless communication, thereby achieving flexible system deployment without the need for large-scale hardware modifications to the UAV platform.

[0103] (10) Regarding terminal power consumption optimization: In terms of terminal design, LoRa terminals can adopt an adaptive data rate (ADR) mechanism to dynamically adjust the transmission power and data rate according to the distance between the terminal and the drone, thereby reducing terminal power consumption and extending the working time of the terminal device, making it more suitable for low-power IoT positioning scenarios.

[0104] (11) Regarding the fusion of positioning results: In practical applications, the positioning results of this invention can be fused with GPS positioning results. When satellite signal conditions are good, the satellite positioning results are used; when satellite signals are blocked or unavailable, the method of this invention is used for positioning, thereby achieving more stable and reliable positioning in all scenarios.

[0105] (12) In terms of engineering applications: In practical engineering applications, parameters in the RSSI ranging model can be corrected through on-site calibration to reduce the impact of environmental factors on signal propagation. Simultaneously, obstacle avoidance modules can be added to the UAV platform to improve flight safety and system reliability in complex environments.

[0106] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal that includes that element.

[0107] Of course, those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware (such as a processor, controller, etc.). The program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The computer-readable storage medium can be a memory, magnetic disk, optical disk, etc.

[0108] It should be understood that the application of the present invention is not limited to the examples above. Those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A LoRa terminal positioning method assisted by a single unmanned aerial vehicle (UAV), characterized in that, The single-UAV-assisted LoRa terminal positioning method includes: Build a LoRa network and complete the design of LoRa terminal nodes and LoRa gateways; A LoRa ranging model is established using an RSSI-based ranging method, and a matching LoRa signal propagation model is selected. During flight, the drone equipped with a LoRa gateway receives signal packets sent by the LoRa terminal, parses the signal packets and extracts RSSI information. At the same time, the drone obtains its current three-dimensional coordinate position information through its onboard positioning module. For fixed LoRa terminal scenarios or mobile LoRa terminal scenarios, an EKF-based LoRa terminal location information acquisition method is used to obtain the terminal location; For fixed LoRa terminal scenarios, the flight path is optimized using a single UAV path optimization method based on CRLB. For mobile LoRa terminal scenarios, the flight path is optimized using a single UAV path optimization method based on FIM increment and Minkowski inequality. The drone flies according to the optimized flight path and determines in real time whether the return-to-home conditions are met during the flight. If the return-to-home conditions are met, the drone is controlled to return to the starting position along the return-to-home path. During the return-to-home process, the drone continues to acquire the current position information of the drone, receive LoRa signals and perform terminal positioning. If the return-to-home conditions are not met, the drone continues to fly according to the optimized path. Determine whether the total positioning time has been reached. If the total positioning time has not been reached and the drone has not entered the return-to-home state, continue to acquire the drone's current location information, receive LoRa signals, and perform terminal positioning and path optimization. If the total positioning time has been reached or the drone has completed its return-to-home process, output the flight path and positioning results.

2. The LoRa terminal positioning method assisted by a single UAV according to claim 1, characterized in that, The LoRa network includes: a terminal device containing a LoRa chip, a LoRa base station, and a server; the LoRa terminal node includes: a LoRa module, a sensor, and an MCU microcontroller; the LoRa gateway includes: a LoRa module and an MCU microcontroller; the LoRa gateway is mounted on a drone, which serves as a mobile base station for LoRa positioning.

3. The LoRa terminal positioning method assisted by a single UAV according to claim 1, characterized in that, The method of establishing the LoRa ranging model using RSSI-based ranging specifically includes: A positioning model is established using an RSSI-based ranging method. In free space, the relationship between signal strength and propagation distance is expressed as: ； in, For received power, For transmission power, For transmitter antenna gain, For receiver antenna gain, The wavelength of the transmitted signal, d This indicates the distance between the receiver and the transmitter. The LoRa signal propagation process is described using a log-normal shading model, and the LoRa ranging model is as follows: ； in, Indicates the first j The distance between each receiving point and the transmitting source This indicates that the distance from the source is The received signal power, This represents the average path loss at the reference point. For reference distance, This is the path loss factor. This represents the random items caused by the shadowing effect.

4. The LoRa terminal positioning method assisted by a single UAV according to claim 1, characterized in that, For fixed LoRa terminal scenarios, the method for obtaining terminal location information using EKF-based LoRa terminal location information acquisition specifically includes: Using the coordinates of unknown nodes as the state vector of the positioning system: ； in, To locate the state vector of the system, For node coordinates, T Indicates transpose; The state equation of a fixed LoRa terminal is: ； in, Indicates the number of times the location was located. Indicates the first The state vector of the next location. Indicates the first +1 positioning state vector Here is the state transition matrix. This is process noise; Using the RSSI received by the LoRa terminal node as the observation value, the measurement equation is: ； in, For the first RSSI measurement value of the second positioning. The nonlinear measurement function is composed of RSSI and terminal position. For measuring noise; The measurement equation is linearized using EKF, and the location estimate of the fixed LoRa terminal is obtained through a prediction and update process.

5. The LoRa terminal positioning method assisted by a single unmanned aerial vehicle (UAV) according to claim 1, characterized in that, For mobile LoRa terminal scenarios, the method for obtaining terminal location information using EKF-based LoRa terminal location information acquisition specifically includes: The state vector of the node to be located is: ； in, For node coordinates, For node speed, The direction of movement of the LoRa terminal node. T This indicates transpose. The motion of the mobile LoRa terminal is described using a random walk model. The node motion model of the mobile LoRa terminal is as follows: ； in, , and Indicates the current position. , and Indicates the position at the next moment. This represents the time interval between the current moment and the next moment. Indicates the velocity at the next moment. Indicates the direction angle of motion at the next moment. Random noise representing changes in drive speed. Random noise representing changes in the driving direction angle; Using RSSI as the observation and employing the EKF recursive process, the location estimate of the mobile LoRa terminal is obtained.

6. The LoRa terminal positioning method assisted by a single UAV according to claim 1, characterized in that, If the scenario involves a fixed LoRa terminal, a single UAV path optimization method based on CRLB is used to optimize the flight path, specifically including: The estimated coordinate covariance matrix of the node to be located satisfies: ； in, This represents the covariance matrix of the estimated location of the node to be located. Indicates the flight path of the UAV Related FIM; The D-optimal criterion is used for path optimization, and the objective function for path optimization is: ； in, Indicates that the UAV is in the The direction angle at that moment, Indicates that the UAV is in the The pitch angle at any moment, This represents the FIM (First Instruction Metric) at the next moment after the UAV moves according to the candidate flight direction. This represents matrix determinant operations.

7. The LoRa terminal positioning method assisted by a single UAV according to claim 1, characterized in that, In the case of a mobile LoRa terminal scenario, a single UAV path optimization method based on the FIM incremental method and the Minkowski inequality is used to optimize the flight path, specifically including: Reconstruct the FIM and calculate the CRLB; Based on the Minkowski inequality for positive definite matrices, the path optimization problem is transformed into optimizing the FIM increment at the next time step. The objective function for path optimization is: ； in, Indicates that the UAV is in the The direction angle at that moment, Indicates that the UAV is in the The pitch angle at any moment, This represents the FIM increment brought about by the UAV flying from the current moment to the next moment. Indicates the flight path, This represents matrix determinant operations.

8. The single UAV-assisted LoRa terminal positioning method according to claim 6 or 7, characterized in that, During path optimization, the objective function must satisfy the following constraints: Starting position constraints: ; in, This is the initial position of the UAV; Termination position constraint: ; in, This is the termination point for the UAV mission; Return-to-home constraints: ; in, Indicates the current position of the UAV. Indicates the flight speed of the UAV. Indicates the total task duration. Indicates the single-step flight time. and These represent the number of positioning attempts and the number of times the LoRa terminal node transmitted signals before the UAV achieved positioning, respectively. Lower limit constraints on flight altitude: ; Flight altitude limit constraints: ; in, For UAV in the Flight altitude at any given moment and These are the minimum and maximum permitted flight altitudes, respectively. Pitch angle variation constraints: ; in, This represents the maximum allowable pitch angle change between adjacent moments for the UAV. Azimuth variation constraints: ; in, This represents the maximum allowable change in azimuth angle between adjacent moments of the UAV.

9. The LoRa terminal positioning method assisted by a single UAV according to claim 1, characterized in that, The single UAV-assisted LoRa terminal positioning method is applicable to multi-LoRa terminal scenarios, which include fixed multi-LoRa terminal scenarios and mobile multi-LoRa terminal scenarios.

10. The LoRa terminal positioning method assisted by a single UAV according to claim 9, characterized in that, For fixed multi-LoRa terminal scenarios, the overall FIM is constructed as a block diagonal matrix, and the path optimization objective is to maximize the product of the determinants of the FIMs of all terminals; for mobile multi-LoRa terminal scenarios, the cumulative FIM increment of multiple targets is calculated, and the path optimization objective is to maximize the product of the determinants of the FIM increments of multiple targets.

Images