Method and system for reconstructing a trajectory of a drone, electronic device and medium

Abstract

The application relates to the technical field of unmanned aerial vehicle monitoring, and specifically discloses a method, system, device and medium for reconstructing the trajectory of an unmanned aerial vehicle, which comprises the following steps: acquiring the received signals of multiple WiFi passive sensing links, wherein the received signals comprise reference signals and monitoring signals; processing the reference signals and the monitoring signals of each sensing link to extract the Doppler observations of each sensing link at each time; generating a coarse reference trajectory according to the Doppler observations of each sensing link at each time; constructing the composite weights of each Doppler observation according to the residual error between the coarse reference trajectory and the Doppler observations, and in combination with the link-level reliability and the local observation quality; selecting multiple trajectory control points as optimization variables based on the coarse reference trajectory; constructing an objective function; optimizing the optimization variables to obtain optimized trajectory control points, and generating the reconstructed trajectory of the unmanned aerial vehicle according to the optimized trajectory control points.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) monitoring technology, specifically to a method, apparatus, device, and medium for reconstructing UAV trajectories. Background Technology

[0002] Low-altitude drones are widely used in logistics delivery, energy inspection, emergency communications, aerial photography and mapping, and urban management. Continuous monitoring of drones is necessary around airports, industrial parks, ports, critical infrastructure, and densely populated areas to prevent collisions, unauthorized filming, payload delivery, and airspace management risks. Existing detection methods include active radar, electro-optical infrared, acoustic, and radio frequency detection. However, active radar is costly and subject to strict spectrum management; electro-optical infrared is greatly affected by weather and obstructions; acoustic detection has a limited range; and radio frequency detection relies on the drone's communication link radiation, making it difficult to cover silent flying targets. Passive WiFi sensing utilizes widely available environmental features... WiFi Signals, as illumination sources, have advantages such as electromagnetic concealment, flexible deployment, and low cost, and can serve as a supplementary means for local low-altitude monitoring.

[0003] However, WiFi The low signal transmission power and weak target scattered echo, coupled with strong direct interference, ground object multipath, and communication noise in the monitoring channel, severely degraded the quality of the received signal. Furthermore, the signal quality varied significantly between different receiving channels and at different times, making it difficult to effectively extract the target echo in the context of strong interference. Summary of the Invention

[0004] This application provides a method, system, electronic device, and medium for reconstructing UAV trajectories. This application acquires Doppler observations through multi-link WiFi passive sensing and generates a coarse reference trajectory. Based on this, it constructs composite weights for each link at each time point, and then constructs an objective function including weighted residual terms, reference residual terms, and motion continuity constraints to perform global trajectory optimization. Thus, without the need for a dedicated active illumination source, it effectively suppresses the impact of weak echoes, strong clutter, and link reliability imbalances on trajectory reconstruction, thereby improving the accuracy and continuity of UAV trajectories.

[0005] On one hand, embodiments of this application provide a method for reconstructing the trajectory of a drone, the method comprising: Acquire received signals from multiple WiFi passive sensing links, the received signals including reference signals and monitoring signals; The reference signal and the monitoring signal of each sensing link are processed to extract the Doppler observations of each sensing link at each time. A coarse baseline trajectory is generated based on the Doppler observations of the sensing link at each time step; Based on the residual between the coarse reference trajectory and the Doppler observation, and combined with link-level reliability and local observation quality, a composite weight for each Doppler observation is constructed. Based on the coarse baseline trajectory, multiple trajectory control points are selected, and the positions of the trajectory control points are used as variables to be optimized. Construct an objective function, which includes a weighted Doppler residual term constructed using the composite weights, a baseline residual term relative to the coarse baseline trajectory, and a motion continuity constraint term; The variables to be optimized are optimized to obtain optimized trajectory control points, and the UAV reconstructed trajectory is generated based on the optimized trajectory control points.

[0006] In some embodiments, processing the reference signal and the monitoring signal of each of the sensing links to extract the Doppler observations of each sensing link at each time step includes: The monitoring signal is subjected to clutter suppression to obtain the clutter-free monitoring signal; The mutual ambiguity function is calculated based on the clutter-suppressed monitoring signal and the reference signal to generate time-Doppler spectra at each time point; The target Doppler ridge is extracted from the time-Doppler spectrum and used as the Doppler observation.

[0007] In some embodiments, the clutter suppression of the monitoring signal includes: The reference signal and the monitoring signal are divided into multiple corresponding signal blocks; For each of the signal blocks, a reference dictionary is constructed based on multiple discrete time-delay replicas of the reference signal, and an orthogonal complementary projection matrix is ​​constructed based on the reference dictionary; The corresponding monitoring signal block is projected using the orthogonal complementary projection matrix to obtain the monitoring signal after clutter suppression.

[0008] In some embodiments, calculating the mutual ambiguity function based on the clutter-suppressed monitoring signal and the reference signal includes generating a time-Doppler spectrum, which includes: The clutter-suppressed monitoring signal and the reference signal are divided into multiple corresponding processing time windows; For each processing time window and each candidate Doppler frequency, calculate the correlation values ​​corresponding to multiple discrete time delays, and determine the corresponding mutual ambiguity function values ​​based on the correlation values; The mutual ambiguity function values ​​are arranged according to the processing time window and candidate Doppler frequencies to generate the time-Doppler spectrum.

[0009] In some embodiments, extracting the target Doppler ridge from the time-Doppler spectrum includes: The time-Doppler spectrum is decomposed into low-rank background components and sparse target components; Candidate Doppler point detection is performed on the sparse target component; A continuous trajectory search is performed on the detected candidate Doppler points to obtain continuous Doppler ridges.

[0010] In some embodiments, the step of decomposing the time-Doppler spectrum into a low-rank background component and a sparse target component includes: The time-Doppler spectrum is decomposed into low-rank background components and sparse target components using low-rank sparse decomposition.

[0011] In some embodiments, the step of detecting candidate Doppler points in the sparse target component includes: Based on the candidate point detection method, candidate Doppler points are extracted according to the signal features of each point in the sparse target component. The candidate point detection method includes one of constant false alarm rate detection, local peak detection, adaptive threshold detection, or learning-based candidate detection.

[0012] In some embodiments, performing a continuous trajectory search on the detected candidate Doppler points to obtain continuous Doppler ridges includes: An optimal path search algorithm is used to search for the optimal path for the candidate Doppler points based on response intensity, time continuity, and Doppler variation smoothness, thereby generating continuous Doppler ridges.

[0013] In some embodiments, generating a coarse reference trajectory based on the Doppler observations of each of the sensing links at each time step includes: Multiple candidate initial states are generated within the monitoring area; Based on each candidate initial state, recursive state estimation and backward smoothing are performed using the Doppler observations to obtain multiple candidate trajectories. Based on the fitting error between each candidate trajectory and the Doppler observation, one of the candidate trajectories is selected as the coarse reference trajectory.

[0014] In some embodiments, the recursive state estimation includes one of extended Kalman filtering, unscented Kalman filtering, particle filtering, batch least squares, and factor graph optimization, and the backward smoothing includes... RTS One of the smoothing methods, namely, fixed hysteresis smoothing.

[0015] In some embodiments, based on the coarse reference trajectory, a composite weight of the Doppler observations of each sensing link at each time step is constructed, including: Based on the residuals between the Doppler observations of each sensing link at each time and the predicted Doppler obtained by mapping the coarse reference trajectory through the bistatic Doppler observation model, the residual attenuation factor is determined. The composite weight is constructed based on one or more of the effective observation mask, link-level reliability factor, and local ridge quality factor, along with the residual attenuation factor.

[0016] In some embodiments, the link-level reliability factor is determined based on the effective observation ratio and / or coarse baseline residual of the corresponding sensing link over the entire time window; the local ridge quality factor is determined based on at least one of ridge energy, confidence level, tracking score, and peak ratio.

[0017] In some embodiments, the step of selecting multiple trajectory control points based on the coarse reference trajectory and using the positions of the trajectory control points as variables to be optimized includes: Multiple coarse reference trajectory points are sampled from the coarse reference trajectory at preset time intervals; The positions of each coarse reference trajectory point are set to the initial values ​​of the corresponding control point positions.

[0018] In some embodiments, optimizing the variables to be optimized based on the objective function to obtain optimized trajectory control points, and generating a UAV reconstructed trajectory based on the optimized trajectory control points, includes: Based on the objective function, a nonlinear optimization algorithm is used to iteratively optimize the variables to be optimized; Based on the optimized trajectory control points, the initial reconstructed trajectory is obtained; The initial reconstructed trajectory is smoothed to obtain the reconstructed trajectory.

[0019] In some embodiments, the motion continuity constraint includes one or more of the following: a second-order difference smoothing term for the control point, a velocity constraint term, and an acceleration constraint term.

[0020] On the other hand, embodiments of this application provide a system for reconstructing the trajectory of a drone, the system comprising: The signal acquisition module is used to acquire the received signals from multiple WiFi passive sensing links, the received signals including reference signals and monitoring signals; The processing module is used to process the reference signal and the monitoring signal of each of the sensing links and extract the Doppler observations of each sensing link at each time. The trajectory generation module is used to generate a coarse reference trajectory based on the Doppler observations of each of the sensing links at each time. The weight construction module is used to construct the composite weight of each effective Doppler observation based on the link-level reliability of each sensing link, the local ridge quality of each Doppler observation, and the residual between each Doppler observation and the predicted Doppler corresponding to the coarse reference trajectory. The setting module is used to select multiple trajectory control points based on the coarse reference trajectory, and to use the positions of the trajectory control points as variables to be optimized; A construction module is used to construct an objective function, which includes a weighted Doppler residual term constructed using the composite weights, a baseline residual term relative to the coarse baseline trajectory, and a motion continuity constraint term. The trajectory optimization module is used to optimize the variables to be optimized, obtain optimized trajectory control points, and generate a UAV reconstructed trajectory based on the optimized trajectory control points.

[0021] In some embodiments, the reconfiguration system further includes multiple WiFi A transmitter and multiple receivers, each receiver being associated with one of the aforementioned WiFi The transmitter forms a sensing link, and the receiver includes a reference channel and a monitoring channel, which are used to acquire reference signals and monitoring signals, respectively.

[0022] On the other hand, embodiments of this application provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it performs a method for reconstructing the trajectory of a drone as described in any of the preceding embodiments.

[0023] On the other hand, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, performs the drone trajectory reconstruction method as described in the above embodiments.

[0024] The UAV trajectory reconstruction method provided in this application, in the reconstruction system, electronic equipment, and storage medium, acquires the received signals of multiple WiFi passive sensing links and extracts Doppler observations of each link. This achieves passive detection of low-altitude UAVs without the need for a dedicated active illumination source, reducing system costs and electromagnetic exposure. Based on this, a coarse reference trajectory is generated, and then a composite weight for each link at each time is constructed according to this reference trajectory. This allows for the quantification of the imbalance in Doppler sensitivity caused by differences in geometric relationships between different links, as well as the fluctuation in observation quality of the same link at different times due to echo fluctuations and clutter interference. The dynamic weights are used to construct weighted residual terms when building the objective function, allowing high-reliability observations to have a larger weight in the optimization process, while low-reliability observations are automatically suppressed. The benchmark deviation term in the objective function ensures that the optimized trajectory does not deviate from the overall trend of the coarse benchmark, and the motion continuity constraint term introduces the physical motion continuity of the UAV into the optimization. Combined with the trajectory optimization parameters set based on the coarse benchmark trajectory, global optimization is performed, which significantly improves the accuracy and continuity of the trajectory reconstruction under conditions of weak echo, strong interference, and severe signal quality degradation, thus achieving stable and reliable UAV trajectory monitoring. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a schematic diagram of a reconstruction system provided as an example of an embodiment of this application.

[0027] Figure 2 This is one of the flowcharts illustrating the method for reconstructing the trajectory of a drone provided in this application.

[0028] Figure 3 This is a second flowchart illustrating the method for reconstructing the trajectory of a drone provided in this application.

[0029] Figure 4 This is the third flowchart illustrating the method for reconstructing the trajectory of a drone provided in this application.

[0030] Figure 5 The fourth flowchart illustrates the method for reconstructing the trajectory of a drone provided in this application.

[0031] Figure 6 The fifth flowchart illustrates the method for reconstructing the trajectory of a drone provided in this application.

[0032] Figure 7 This is the sixth flowchart illustrating the method for reconstructing the trajectory of a drone provided in this application.

[0033] Figure 8 This is the seventh flowchart illustrating the method for reconstructing the trajectory of a drone provided in this application.

[0034] Figure 9 This is the eighth flowchart illustrating the method for reconstructing the trajectory of a drone provided in this application.

[0035] Figure 10 This is the ninth flowchart illustrating the method for reconstructing the trajectory of a drone provided in this application.

[0036] Figure 11 This is a schematic diagram of Doppler ridge recovery obtained after single-link preprocessing in this application.

[0037] Figure 12 This is a schematic diagram comparing the Doppler ridge obtained after single-link preprocessing in this application with the actual Doppler ridge.

[0038] Figure 13 This is one of the schematic diagrams comparing the reconstructed trajectory obtained using the reconstruction method of this application with the UAV trajectory obtained using other techniques.

[0039] Figure 14 This is the second schematic diagram comparing the reconstructed trajectory obtained using the reconstruction method of this application with the UAV trajectory obtained using other technologies.

[0040] Figure 15 This application provides a schematic diagram of the structure of a drone trajectory reconstruction system.

[0041] Figure 16 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0042] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0043] Low-altitude drones have been widely adopted in logistics delivery, energy inspection, emergency communications, aerial surveying and mapping, and urban management. With the development of the low-altitude economy and urban low-altitude air routes, continuous monitoring of low-altitude drones is necessary in industrial parks, airport perimeters, ports, exhibition venues, critical infrastructure, and densely populated areas. Unauthorized or abnormally flying drones may pose risks such as collisions, unauthorized filming, payload delivery, communication interference, and airspace management issues.

[0044] Currently, the main technologies used for UAV detection include active radar, photoelectric / infrared detection, acoustic detection, radio frequency detection, and multi-sensor fusion. Active radar, by actively transmitting detection waveforms and receiving echoes, can directly acquire target distance, speed, or angle information. However, it typically relies on dedicated transmitters, antennas, and radio frequency front-end hardware, resulting in high equipment costs, power consumption, installation environment requirements, and stringent spectrum / electromagnetic emission management requirements. Photoelectric and infrared detection methods are susceptible to light intensity, weather conditions, and obstructions; acoustic detection is limited by environmental noise levels and sound wave propagation distance; radio frequency detection relies on the active radiation of the UAV's communication or remote control link, making it difficult to cover targets in a silent state, pre-programmed flight, or with the link closed. Therefore, relying solely on traditional dedicated sensors makes it difficult to achieve a monitoring effect that balances economy, concealment, and continuous stability in large-scale, localized low-altitude scenarios.

[0045] Passive radar and passive communication sensing technologies utilize existing broadcast, cellular, and other technologies in the environment. WiFi Using communication signals as an opportunity illumination source, the receiver simultaneously acquires the direct reference signal from the transmitter and the monitoring signal of the target's scattered echo. Through reference signal cancellation processing, cross-correlation / mutual ambiguity function calculation, constant false alarm rate (CFAR) detection, and subsequent correlation and tracking processing, the target's motion state is estimated. This type of technology does not require the additional transmission of dedicated radar waveforms and possesses significant advantages such as good electromagnetic concealment, low spectrum occupation, relatively low hardware cost, and flexible deployment. Therefore, it is suitable as a supplementary sensing means in low-altitude security networks.

[0046] Existing research has utilized cellular downlink signals for passive sensing, which can leverage... LTE Alternatively, fifth-generation mobile communication base station signals can be used to extract bistatic distance, Doppler frequency shift, or angle of arrival information, and combined with digital array processing, bistatic geometric relationships, or recursive tracking algorithms to achieve UAV positioning and trajectory estimation. Cellular base stations typically have high transmission power and large coverage areas; however, their location, antenna orientation, installation height, obstruction conditions, and operating parameters are mainly determined by communication coverage and capacity planning, not by geometric optimization for local low-altitude sensing. For local monitoring areas such as campuses, parks, building boundaries, or temporary activity sites, cellular illumination sources may face problems such as uncontrollable geometric configurations, unadjustable obstruction conditions, or insufficient local observability.

[0047] In comparison, WiFi Access points offer advantages such as low cost, large quantity, and the ability to be actively selected by the monitoring area. They can be flexibly deployed on rooftops, near windows, perimeters, and site edges, forming customizable bistatic or multistatic geometries with multiple distributed receiving units. Simultaneously, WiFi The transmitter can operate on different channels to reduce mutual interference, making it suitable as a supplementary illumination source in local low-altitude sensing. However, WiFi The transmission power and coverage radius of the signal are usually lower than those of cellular base stations, and the scattered echo of the target UAV is relatively weak. There are also strong direct wave leakage, ground object / building multipath interference, other communication signal interference and thermal noise in the monitoring channel, which makes the Doppler component of the target easy to be submerged.

[0048] In WiFi passive sensing systems, the receiver typically needs to extract the transmitted signal template from the reference channel and suppress leakage and static multipath components associated with the reference signal in the monitoring channel. Then, it obtains the time-Doppler spectrum through mutual ambiguity function processing. Due to the weak scattered energy of UAVs and the dynamic changes in their flight attitude and bistatic geometry over time, the target ridge line in the time-Doppler spectrum often exhibits instability phenomena such as discontinuity, broadening, peak quality fluctuations, or being pulled by residual clutter.

[0049] In view of this, embodiments of this application provide a method, system, electronic device, and medium for reconstructing the trajectory of a drone. This solves the problem... WiFi Low signal transmission power and weak target scattered echo, coupled with strong direct interference, ground object multipath, and communication noise in the monitoring channel, often cause the target ridge line in the time-Doppler spectrum to appear discontinuous, broadened, or fluctuating in peak value.

[0050] Specifically, the drone trajectory reconstruction method of this application embodiment can be executed by an electronic device, which can be a terminal or a server. The terminal can be a smartphone, tablet, laptop, smart TV, wearable smart device, smart vehicle terminal, etc. The terminal can also include a client, which can be a user client, browser client, instant messaging client, or mini-program, etc.

[0051] A server can be a standalone physical server, a server cluster or a distributed system composed of multiple physical servers, and can also provide cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, and content delivery networks. ContentDeliveryNetwork , CDN Cloud servers that provide basic cloud computing services such as big data and artificial intelligence platforms.

[0052] It should be noted that, in this embodiment, the execution entity of the UAV trajectory reconstruction method can be a terminal device or a server. The terminal device can be a local terminal device or the aforementioned client device. This embodiment does not limit the type of execution entity.

[0053] For example, in conjunction with the above description, Figure 1 This application illustrates a reconstruction system 1000 for implementing a method to reconstruct the trajectory of a drone. The reconstruction system 1000 may include multiple WiFi passive sensing links 101. (WiFi passive sensing links) 101 The number can be 2, 3, 4 or more. The WiFi passive sensing link 101 can be... WiFi The dual-base link, each WiFi passive sensing link 101 includes a transmitter 1001 and a receiver 1002, that is, each receiver 1002 and a transmitter 1001 correspond to form a WiFi passive sensing link 101.

[0054] Transmitter 1001 can be, but is not limited to, a commercial router, access point, hotspot device, or other device capable of providing known communication signals. WiFi The transmitting equipment, and each transmitter 1001 can use different channels. The receiver may be, but is not limited to, [specific types]. USRP Software-defined radio receiver, multi-channel acquisition equipment, or equivalent equipment with reference / monitoring synchronous acquisition capabilities.

[0055] Receiver 1002 includes a reference channel and a monitoring channel. The reference channel is used to acquire a reference signal, and the monitoring channel is used to acquire a monitoring signal. The reference signal is a direct signal from transmitter 1001 to receiver 1002. WiFi Signals. Monitoring signals can be superimposed signals from drone scattering, reference leakage, multipath interference, and noise. Among them, drone scattering signals are weak echoes radiated in all directions after electromagnetic waves radiated by the transmitter hit the surface of the drone; reference leakage (direct wave interference) is the signal that slips into the monitoring channel directly through spatial diffraction, antenna sidelobes, or hardware coupling of the transmitter's direct wave signal; multipath clutter is the component of the transmitted signal that is received by the monitoring antenna after being reflected by large background objects such as the ground, buildings, and trees, whether stationary or moving; noise consists of Gaussian white noise generated by the thermal motion of the receiver's internal electronic components, as well as external radio frequency interference in the environment.

[0056] The reconfiguration system 1000 can also employ a high-precision clock synchronization module to ensure the consistency of the sampling clock between channels, and at a preset sampling frequency (e.g., 20). MHz Up to 40 MHz ,cover WiFi 2.4 GHz / 5 GHz The frequency band bandwidth is used to perform analog-to-digital conversion on the intermediate frequency signal after down-conversion at the RF front-end, thereby acquiring multiple digital baseband signals in parallel. IQ Data refers to the received signals from each sensing link.

[0057] It should be noted that, Figure 1 The schematic diagram of the reconstruction system shown is merely an example. The reconstruction system 1000 described in this application embodiment is intended to more clearly illustrate the technical solutions of this application embodiment and does not constitute a limitation on the technical solutions provided in this application embodiment. As those skilled in the art will know, with the evolution of reconstruction systems and the emergence of new business scenarios, the technical solutions provided in this application embodiment are also applicable to similar technical problems.

[0058] The technical solution of this application will be described in detail below through specific embodiments. It should be noted that the following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0059] Please see Figures 2 to 10 , Figures 2-10 This is a flowchart illustrating a method for reconstructing a drone trajectory provided in an embodiment of this application. It should be noted that the steps shown may be executed in a different logical order than those shown in the flowchart. The method may include the following steps: Step 110: Obtain the received signals from multiple WiFi passive sensing links. The received signals include reference signals and monitoring signals.

[0060] Wherein, the reference signal is the reference signal acquired by the receiver's reference channel, that is, by WiFi Transmitter directly to receiver WiFi The monitoring signal is the monitoring signal collected by the monitoring channel of the receiver. That is, it is the superimposed signal formed by the UAV's scattered signal, reference leakage, multipath, and noise superposition.

[0061] For the Link, the first There are processing time windows, and the reference signal and monitoring signal are denoted as follows: and ,in, .

[0062] In one example, three lines were deployed around the drone's flight area. WiFi Dual-base sensing links, each link including a commercial... WiFi The router has a transmitter and a receiver. The receiver uses... USRPX 310 software radio; one of its directional antennas is oriented towards the corresponding WiFiOne transmitter is used to acquire reference channel signals, and another directional antenna is pointed towards the UAV's flight area to acquire surveillance signals. WiFi The transmitters use different channels to reduce interference between communication signals on different links. The experimental target is a non-cooperative drone. Each receiver collects baseband signals and transmits them to the host for offline or online processing; the sampling rate can be set to 1. MS / s Reference and monitoring signals are acquired within each processing time window.

[0063] Step 120: Process the reference signals and monitoring signals of each sensing link to extract the Doppler observations of each sensing link at each time.

[0064] Among them, the Doppler observations at each moment refer to the instantaneous Doppler frequency shift values ​​calculated for each sensing link according to a preset time sampling interval during the continuous monitoring process of the reconstruction system. These are the core observation data reflecting the motion state of the UAV target. Since the UAV is in continuous motion, its relative velocity with the sensing link changes over time, and the corresponding Doppler frequency shift also exhibits time-varying characteristics. Therefore, each discrete moment corresponds to an independent Doppler observation value. These observation values ​​are arranged in chronological order, which can completely present the change law of the UAV's motion velocity during the monitoring period.

[0065] Please see Figure 3 Optionally, step 120 can be implemented by steps 121-123, that is, step 120 may include: Step 121: Perform clutter suppression on the monitoring signal to obtain the clutter-suppressed monitoring signal; Understandably, surveillance signals are a superposition of multiple components, including target scattering signals (useful), direct wave leakage, ground / building multipath clutter, other communication interference, and thermal noise. Direct wave leakage and static multipath clutter are both strongly correlated with the reference signal, while the target scattering signal has a weaker correlation with the reference signal due to Doppler frequency shift. Therefore, this difference in correlation can be used to suppress reference leakage and multipath clutter in the surveillance signal, removing interference components related to the reference signal from the surveillance signal.

[0066] In actual processing, clutter suppression can be performed for each processing time window. Clutter suppression methods may include, but are not limited to, extended decoupling batch processing algorithms (...). ExtensiveCancellationAlgorithmBatches , ECA - B Least squares projection, subspace projection, adaptive filtering (such as...) LMS , RLS Background modeling or other clutter suppression methods that utilize reference signals to construct destructive bases.

[0067] In this way, by using the reference channel signal to suppress clutter in the monitoring channel signal, the strong direct interference and the masking of the target echo by ground object multipath are effectively eliminated, making the weak target scattering signal stand out from the strong interference background.

[0068] Please see Figure 4 Optionally, the following can be adopted: ECA - B The algorithm implements clutter suppression of the monitored signal. Step 121 can be implemented by steps 1211 and 1213; that is, step 121 may include: Step 1211: Divide the reference signal and the monitoring signal into multiple corresponding signal blocks; That is, the long sequence signal received in the continuous time domain is time-sliced, and the reference signal and the monitoring signal during the entire observation time are synchronously divided into K data segments (i.e. signal blocks) with fixed durations and lengths according to the same start time and time length.

[0069] Step 1212: For each signal block, construct a reference dictionary based on multiple discrete time-delay replicas of the reference signal, and construct an orthogonal complementary projection matrix based on the reference dictionary; Specifically, for the i-th signal block currently being processed, the reference signal within the block is first time-delayed and extended to generate N shifted copies of it at different distance delays. All copies are then arranged and combined into a high-dimensional matrix, which is the reference dictionary (also known as the observation matrix H). Subsequently, based on this dictionary matrix, the projection operator matrix of the orthogonal complement space corresponding to the subspace spanned by its column vectors is calculated through linear algebra operations.

[0070] Step 1213: Project the corresponding monitoring signal block using the orthogonal projection matrix to obtain the monitoring signal after clutter suppression.

[0071] The formula for generating a monitoring signal to suppress clutter can be:

[0072]

[0073] in, For the first Link number k Orthogonal projection matrix of time windows For the first Link number The reference dictionary matrix for each time window is constructed from the reference channel signal and contains multiple delay taps. It is the identity matrix. For reference, the conjugate transpose of the dictionary. For monitoring signal vectors, This is the monitoring vector after clutter suppression. The regularization coefficient is . It is the inverse of the regularized autocorrelation matrix.

[0074] Thus, by using projection cancellation to suppress clutter in the monitoring channel signal using the reference channel signal, it is not necessary to know the specific time delay, amplitude, and phase parameters of the clutter in advance. Only by constructing a cancellation basis using the reference signal, the direct leakage and static multipath components related to the reference signal can be removed from the monitoring signal. This effectively preserves the target scattering signal in strong direct interference and complex multipath environments, thereby improving the signal-to-clutter ratio of the monitoring signal.

[0075] Step 122: Calculate the mutual ambiguity function based on the clutter-suppressed monitoring signal and reference signal to generate time-Doppler spectra at each time point; It is worth noting that the monitoring signal after clutter suppression is still weak and requires further processing to extract the Doppler information of the target. The mutual ambiguity function is physically a two-dimensional correlation function that describes the similarity distribution of two signals in a two-dimensional plane of time delay (τ) and Doppler frequency shift (f). The time-Doppler spectrum is a standard tool for estimating time-delay-Doppler two-dimensional parameters. The time-Doppler spectrum is a two-dimensional matrix reflecting the distribution of Doppler energy over time, where rows correspond to processing time windows (time axis), columns correspond to Doppler frequencies (frequency axis), and the value of each element reflects the signal energy at that Doppler frequency at that moment.

[0076] Specifically, for each segment of the time-domain monitoring signal output after clutter suppression, a two-dimensional time-frequency cross-correlation operation (i.e., cross-ambiguity function, CAF) is performed with the reference signal. This operation calculates the coherent accumulation amplitude under each set of time delay-frequency shift combinations by traversing the preset time delay search range (corresponding to the range dimension) and Doppler frequency shift search range (corresponding to the velocity dimension), thereby constructing the range-Doppler two-dimensional energy spectrum at the current processing time.

[0077] Optionally, please refer to Figure 5 Step 122 can be implemented by steps 1221-1223, that is, step 122 may include: Step 1221: Divide the clutter-suppressed monitoring signal and reference signal into multiple corresponding processing time windows; Two long-time signals can be synchronously divided into k consecutive or partially overlapping processing time windows according to the preset coherent accumulation time (CPI).

[0078] Step 1222: For each processing time window and each candidate Doppler frequency, calculate the correlation values ​​corresponding to multiple discrete time delays, and determine the corresponding mutual ambiguity function values ​​based on the correlation values; Specifically, for a fixed candidate Doppler frequency, the reference signal is first digitally down-converted (frequency compensation), and then the sliding dot multiplication is performed with the monitoring signal to obtain the complex correlation values ​​of different time delays under that frequency shift; the process is then switched to the next candidate Doppler frequency and the above operation is repeated to finally obtain the complex mutual ambiguity function value matrix under the current time window.

[0079] Step 1223: Arrange the mutually blurred function values ​​according to the processing time window and candidate Doppler frequencies to generate a time-Doppler spectrum.

[0080] The formula for generating the time-Doppler spectrum at each moment can be:

[0081]

[0082] in, For the first l Link, the first k Time window, Doppler frequency f place CAF Response amplitude, This is the monitoring signal after clutter suppression. For the reference signal delayed by τ sampling points, For delayed search set, The frequency of the Doppler signal to be detected. The sampling period; This indicates that the time-Doppler spectrum is obtained by aggregating the discrete frequency set after point-by-point calculation. .

[0083] Step 123: Extract the target Doppler ridge from the time-Doppler spectrum as a Doppler observation.

[0084] Among them, the target Doppler ridge is a continuous strip-shaped peak trajectory formed by the signal of a single sensing link on a time-Doppler two-dimensional time-frequency map. It is formed by connecting the time-varying Doppler frequency shift peaks generated by the continuous movement of the UAV target in sequence, and its change over time reflects the dynamic evolution of the target Doppler frequency shift.

[0085] Thus, by extracting the target Doppler ridge from the spectrum, the energy distribution of the two-dimensional spectrum is transformed into a one-dimensional continuous Doppler observation, thereby achieving stable extraction of Doppler information from each sensing link at each moment under conditions of weak echo and strong interference, providing a reliable observational data foundation for subsequent trajectory reconstruction.

[0086] Optionally, please refer to Figure 6 Step 123 can be implemented by steps 1231-1233, that is, step 123 may include: Step 1231: Decompose the time-Doppler spectrum into low-rank background components and sparse target components; Specifically, the time-Doppler spectrum can be normalized and decomposed using low-rank sparsity to separate the low-rank background component and the sparse target component. The low-rank background component is the energy-dominant but slowly changing part of the spectrum. "Low-rank" means the matrix has a low rank—that is, there is a strong linear correlation between the columns (or rows) of the matrix, implying that the changes in the spectrum in the time and frequency dimensions are slow and predictable. The low-rank background component physically originates from the direct waves and multipath leakage remaining after clutter suppression. The sparse target component is the energy-weak but structurally distinct part of the spectrum. "Sparse" means that most elements of this matrix are zero, with only a few positions having non-zero values; that is, the target Doppler ridge occupies only a very small area in the spectrum.

[0087] Furthermore, the time-Doppler spectrum can be decomposed using low-rank sparse decomposition to obtain low-rank background components and sparse target components. Low-rank sparse decomposition methods may include, but are not limited to, robust principal component analysis. RPCA ), low-rank representation ( LRR Time-frequency filtering or sparse enhancement, etc.

[0088] For example, in some cases, robust principal component analysis can be used to analyze time-Doppler spectra, and the formula can be:

[0089] in, Indicates a low-rank background term. Represents sparse target terms, The regularization coefficient is used to balance low-rank terms and sparse terms. This represents the normalized time-Doppler spectrum matrix.

[0090] Thus, by decomposing the time-Doppler spectrum into low-rank background components and sparse target components, the weak target echoes that were originally masked by strong clutter can be separated from the background, significantly improving the signal-to-clutter ratio.

[0091] Step 1232: Detect candidate Doppler points in the sparse target component; Among them, candidate Doppler points can be extracted based on the signal features of each point in the sparse target component, according to a preset candidate point detection method.

[0092] Candidate point detection methods may include, but are not limited to, constant false alarm rate (CFAR) detection, local peak detection, adaptive threshold detection, or learned candidate detection. CFAR detection estimates the local background using the mean amplitude / ranking statistics of reference cells surrounding the target point, calculates the detection threshold, and extracts points exceeding the threshold. Local peak detection searches for local amplitude maxima in sparse target components. Adaptive threshold detection adjusts the threshold adaptively based on the local distribution characteristics of the signal amplitude (variance, quantiles, etc.). Learned candidate detection uses a trained classification model (such as a neural network or support vector machine) to determine whether each point belongs to the target Doppler point.

[0093] For example, in one instance, the candidate point detection method can be constant false alarm rate (CFAR) detection, and the candidate Doppler point set can be represented as:

[0094] in, For the first l Link, the first k A set of candidate Doppler points for a time window. for CFAR Threshold coefficient (set according to the target false alarm rate). For sparse target components in time k ,frequency The amplitude at that point; This is for local background estimation.

[0095] Thus, by detecting candidate Doppler points in sparse target components, a set of candidate points belonging to the target Doppler ridge line is effectively extracted.

[0096] Step 1233: Perform continuous trajectory search on the detected candidate Doppler points to obtain continuous Doppler ridges.

[0097] An optimal path search algorithm can be used to search for the optimal path for candidate Doppler points based on response intensity, time continuity, and smoothness of Doppler variation, thereby generating continuous Doppler ridges.

[0098] The optimal path search algorithm may include, but is not limited to, second-order algorithms. DP Viterbi search, graph search, dynamic programming variants, particle path search, or other continuous trajectory search methods.

[0099] For example, a second-order method can be used. DP A trade-off between candidate response intensity, temporal continuity, and Doppler variation smoothness.

[0100] Therefore, please refer to Figure 11 and Figure 12By performing continuous trajectory search on candidate points and utilizing the triple criteria of response intensity, temporal continuity, and Doppler variation smoothness, continuous and smooth Doppler ridges were recovered from discrete candidate points with missed detections and false alarms. This solved the problem of discontinuity, broadening, or peak jitter of target ridges under weak echo conditions, and provided a complete and reliable Doppler observation sequence for subsequent trajectory reconstruction.

[0101] Step 130: Generate a coarse baseline trajectory based on the Doppler observations of the sensing link at each time point.

[0102] Among them, the coarse reference trajectory is an initial trajectory that can reflect the general movement trend of the target, obtained by multiple initial value recursive estimation under the condition that only Doppler observation is available and range and angle measurement information is lacking.

[0103] Optionally, please refer to Figure 7 Step 130 can be implemented by steps 131-133, that is, step 130 may include: Step 131: Generate multiple candidate initial states within the monitoring area; It is worth noting that candidate initial states can include initial position, initial velocity, and heading. Candidate initial states can be generated using gridded sampling, random sampling, multi-starting-point optimization, or sampling based on prior regions. Gridded sampling involves sampling the initial position at fixed intervals in the position space and enumerating the heading and velocity at fixed step sizes in the velocity space. Random sampling randomly generates a large number of candidate states in the state space. Multi-starting-point optimization starts from a small number of initial points and uses optimization algorithms to find multiple locally optimal starting points. Sampling based on prior regions narrows the sampling range based on prior information about the monitoring area (such as the boundaries of the UAV take-off and landing zone and no-fly zone).

[0104] For example, the initial position is sampled in a grid pattern within the monitoring area, and multiple initial headings and velocities are enumerated to obtain multiple candidate initial states, ultimately resulting in a set of candidate initial states.

[0105] Step 132: Based on each candidate initial state, recursive state estimation and backward smoothing are performed using Doppler observations to obtain multiple candidate trajectories.

[0106] It is worth noting that recursive state estimation refers to the process of updating the target state step by step using current observation data, starting from the initial state and proceeding sequentially over time. Backward smoothing, on the other hand, refers to the process of reversing the historical state estimates from the end point to the starting point, starting from the last moment and using observation information from future moments to revise the historical state estimates.

[0107] Among them, recursive state estimation can employ extended Kalman filtering or unscented Kalman filtering. UKF)Particle filtering, batch least squares, factor graph optimization, etc., can be used for backward smoothing. RTS Smoothing, fixed hysteresis smoothing, etc.

[0108] Specifically, for each candidate initial state, a recursive state estimation can be performed, where at time... The predicted Doppler is based on the bistatic Doppler observation model. The residual is calculated and compared with the measured Doppler to obtain the residual, which is used for recursive state estimation. After all time steps are processed, the forward estimation result can be reverse smoothed by backward smoothing to obtain the candidate trajectory corresponding to the initial value. Finally, multiple candidate trajectories corresponding to each candidate initial state are obtained in sequence.

[0109] Bistatic Doppler observation model The calculation formula can be:

[0110] in, For the first l The link at time k The predicted Doppler frequency shift, For a moment k The target state vector, For a moment k The drone's position vector For a moment k The drone's velocity vector and These are the drone's position and speed, respectively. and These are the locations of the transmitter and receiver, respectively.

[0111] Step 133: Based on the fitting error between each candidate trajectory and the Doppler observation, select one of the candidate trajectories as the coarse baseline trajectory.

[0112] Under multiple initial conditions, different candidate initial states will generate different candidate trajectories. It is necessary to establish evaluation criteria to quantify the degree of matching between each candidate trajectory and Doppler observations, and select the best one as the coarse baseline trajectory.

[0113] For the m Given a smooth state sequence corresponding to each initial value, define the cumulative Doppler fitting error, and select the trajectory with the smallest error as the coarse baseline trajectory. The calculation formula can be:

[0114]

[0115] in, For the first The link at time The recovered Doppler observations This is a bistatic Doppler observation model. The Doppler residual scale.

[0116] Thus, by enumerating multiple candidate initial states in the position space and motion state space, the problems of sensitivity to initial values ​​and geometric ambiguity in state estimation under Doppler observation alone are overcome. On this basis, by screening out abnormal observations that are inconsistent with the prediction, erroneous ridges are prevented from causing trajectory divergence. Then, state estimation and smoothing are performed by combining a bistatic Doppler observation model. The cumulative error of forward recursion is corrected using all observation data to obtain multiple candidate trajectories. Finally, the best candidate trajectory is selected as the coarse baseline trajectory based on the fitting error between each candidate trajectory and the Doppler observation. Under the condition that the initial state is unknown and some link observations are invalid, a reliable initial trajectory that can reflect the general motion trend of the target is provided, providing a unified reference benchmark for subsequent composite weight construction and global trajectory optimization.

[0117] Step 140: Based on the residuals between the coarse baseline trajectory and the Doppler observations, and in combination with link-level reliability and local observation quality, construct the composite weights for each Doppler observation. The composite weight is determined by a combination of multiple factors, including but not limited to: effective observation mask (indicating whether there is a valid observation for the corresponding link at the corresponding time), link-level reliability factor (reflecting the overall reliability of the link during the entire observation period), local ridge quality factor (reflecting the local quality of the ridge at that time), and residual decay factor.

[0118] Link-level reliability factors can be determined based on the proportion of effective observations of the corresponding sensing link across all time windows and / or the coarse baseline residuals, reflecting the overall observation quality of the link. Local ridge quality factors are determined based on one or more of ridge energy, confidence level, tracking score, and peak-to-peak ratio, reflecting the reliability of ridge recovery at that moment. Residual attenuation factors are determined based on the residuals between Doppler observations and the predicted Doppler data from the coarse baseline trajectory, used to reduce the weight of observations that are significantly inconsistent with the coarse baseline motion trend.

[0119] It is worth noting that the specific functional form of the composite weight can be replaced with equivalents, as long as the link efficiency, local ridge quality, and consistency of the relatively coarse baseline trajectory are considered simultaneously or in combination. The number of local quality indicators, weight ratios, normalized quantiles, smoothing windows, and lower limit parameters can all be adjusted according to the scenario.

[0120] Optionally, please refer to Figure 8 Step 140 can be implemented by steps 141 and 142; that is, step 140 may include: Step 141: Determine the residual attenuation factor based on the residual between the Doppler observations of each sensing link at each time point and the predicted Doppler obtained by mapping the coarse reference trajectory through the bistatic Doppler observation model. For the l The link at time Doppler observations The predicted Doppler obtained by mapping the coarse baseline trajectory to a bistatic Doppler observation model The technical formula for the residuals between them can be:

[0121] in, For the coarse baseline trajectory at time... The state (velocity and position). For the first The link at time The recovered Doppler observations This is a bistatic Doppler observation model.

[0122] Step 142: Construct composite weights based on one or more of the effective observation mask, link-level reliability factor, and local ridge quality factor, along with the residual attenuation factor. For example, the composite weight is obtained by multiplying the effective observation mask, link-level reliability, local ridge quality, and residual attenuation factor. The calculation formula can be:

[0123] In the formula, To effectively observe the mask, For link-level reliability, For local ridge quality, This is the attenuation factor based on the coarse reference Doppler residual.

[0124] Thus, by determining the residual attenuation factor based on the residual of Doppler observations of each sensing link at each time point relative to the predicted Doppler of the coarse reference trajectory, abnormal observations deviating from the coarse reference motion trend are automatically suppressed. On this basis, a composite weight is constructed based on one or more of the effective observation mask, link-level reliability factor, and local ridge quality factor, together with the residual attenuation factor. This allows for the direct elimination of unobserved moments, global weighting of links with poor overall quality, local weighting of moments with low local quality, and dynamic suppression of observations deviating from the reference trend. This forms a hierarchical, multi-dimensional observation reliability assessment mechanism, providing accurate adaptive weight input for the weighted residual term in subsequent trajectory optimization.

[0125] Step 150: Based on the coarse baseline trajectory, select multiple trajectory control points and use the positions of the trajectory control points as variables to be optimized; Understandably, the coarse baseline trajectory is a complete initial trajectory (containing the position at each time step), but its accuracy is limited. Directly using it as the final output would result in significant error. Therefore, a trajectory parameterization strategy is needed, which uses a set of parameters—far fewer than the number of time windows—to characterize the coarse baseline trajectory, transforming the high-dimensional point-by-point trajectory optimization problem into a low-dimensional parameter optimization problem. The role of the coarse baseline trajectory here is to provide the initial values ​​for these parameters. The position coordinates of the control points are used as optimization variables, and their initial values ​​are taken from their positions on the coarse baseline trajectory at the corresponding time steps.

[0126] Optionally, please refer to Figure 9 Step 150 can be implemented by steps 151 and 152; that is, step 150 may include: Step 151: Sample multiple coarse reference trajectory points from the coarse reference trajectory according to a preset time interval; Specifically, the coarse baseline trajectory curve in the continuous time domain is discretized and extracted at a fixed time period (determined by the system data rate or pulse repetition interval), resulting in a set of sparse position sequence points representing the overall trajectory trend, arranged in chronological order. Each position sequence point contains time information and position coordinates. An adaptive spacing strategy can be employed for sampling, with denser sampling in areas of rapid trajectory change (such as curves) and sparser sampling in areas of gentle trajectory, to achieve optimal trajectory approximation with a limited number of control points.

[0127] Step 152: Set the position of each coarse reference trajectory point to the initial value of the corresponding control point position.

[0128] That is, the coordinates of each discrete point obtained by sampling are directly mapped one by one to the iterative starting solution of the control variable to be optimized in the trajectory optimization algorithm, that is, the initial value injection channel from the coarse measurement position to the optimization variable space is established.

[0129] Step 160: Construct the objective function, which includes a weighted Doppler residual term constructed using composite weights, a baseline residual term relative to the coarse baseline trajectory, and a motion continuity constraint term.

[0130] Among them, the motion continuity constraint term includes one or more of the second-order difference smoothing term, velocity constraint term, and acceleration constraint term of the control point.

[0131] In addition, the objective function can also simultaneously consider the weighted normalized Doppler residual, the reference trajectory residual, the second-order difference smoothing term of the control point, the velocity over-limit penalty, and the acceleration over-limit penalty, so as to obtain a continuous trajectory that satisfies both the multi-link Doppler geometric consistency and the physical motion constraints of the UAV.

[0132] The objective function can be:

[0133]

[0134] in, To weighted normalize the Doppler residuals, The deviation of the control point relative to the coarse reference control point. This includes second-order difference of control points, penalties for exceeding velocity and acceleration limits. To optimize the predicted Doppler values ​​for the trajectory, For Doppler observation, To predict the residuals between Doppler and Doppler observations. As the normalization factor, For composite weights, For a moment The set of valid links.

[0135] Step 170: Based on the objective function, optimize the variables to be optimized to obtain the optimized trajectory control points, and generate the UAV reconstructed trajectory based on the optimized trajectory control points.

[0136] It is worth noting that optimizing the variable to be optimized based on the objective function means treating the variable to be optimized as an unknown quantity to be solved, using the objective function as the evaluation criterion, and iteratively adjusting the value of the variable to be optimized through numerical optimization algorithms to gradually reduce the value of the objective function.

[0137] Generating a UAV reconstructed trajectory based on optimized trajectory control points refers to obtaining the trajectory control points that minimize the objective function after optimization iteration convergence, and then restoring the complete UAV trajectory through the mapping relationship corresponding to these trajectory control points in a parameterized manner.

[0138] Please see Figure 10 Optionally, step 170 can be implemented by steps 171 and 172, that is, step 170 may include: Step 171: Based on the objective function, a nonlinear optimization algorithm is used to iteratively optimize the variables to be optimized; Since the weighted Doppler residual term in the objective function contains a nonlinear mapping of the bistatic Doppler observation model, and the motion continuity constraint term contains nonlinear penalties for velocity and acceleration constraints, the objective function is a nonlinear function of the variables to be optimized. Therefore, the optimization problem is a nonlinear optimization problem and requires iterative solution using numerical optimization methods. The variables to be optimized are treated as unknowns to be solved, and the objective function is used as the evaluation criterion. Iterative calculations are then used to find the optimal parameters that minimize the objective function.

[0139] In each iteration, a complete planar trajectory is generated from the current variable to be optimized through the mapping relationship corresponding to the parameterization method (such as piecewise interpolation), and then the position, velocity and acceleration at each moment are calculated; the objective function value is calculated based on the complete trajectory, including the weighted Doppler residual term, the baseline deviation term and the motion continuity constraint term; at the same time, the gradient information of the objective function with respect to each variable to be optimized can be calculated through the chain rule.

[0140] The nonlinear optimization algorithm determines the parameter update direction and step size for the current iteration based on the objective function value and gradient information, and generates the updated variables to be optimized.

[0141] Repeat the above process until the preset convergence conditions are met. Convergence conditions include, but are not limited to: the change in the objective function is less than a preset threshold, the change in the variable to be optimized is less than a preset threshold, or the number of iterations reaches the maximum number of iterations.

[0142] The nonlinear optimization algorithm may be any one of the trust region reflection algorithm, Levenberg-Marquardt algorithm, Gauss-Newton algorithm or sequential quadratic programming algorithm, and this application does not limit it.

[0143] Step 172: Obtain the initial reconstructed trajectory based on the optimized trajectory control points; The discrete control point sequence that converges after iterative optimization can be substituted into a preset trajectory generation basis function (such as piecewise interpolation basis function, B-spline curve basis function, or Bézier curve basis function). Through linear combination interpolation of the basis functions, a continuous time domain trajectory curve covering the entire observation period can be generated, which is the initial reconstructed trajectory.

[0144] Taking piecewise interpolation basis functions as an example: one-dimensional linear interpolation is performed on the horizontal and vertical coordinates along the time axis to generate a complete planar trajectory, and the velocity and acceleration are calculated by finite difference to obtain the initial reconstructed trajectory.

[0145] Step 173: Smooth the initial reconstructed trajectory to obtain the reconstructed trajectory.

[0146] Specifically, while maintaining the overall trajectory trend, velocity / acceleration constraints or low-pass filtering are applied to the initial reconstructed trajectory to suppress high-frequency jitter components introduced by measurement noise or optimization residuals, outputting a smooth trajectory that satisfies the UAV's flight mechanics constraints (such as maximum overload limits and bounded jerk), which is the final output reconstructed trajectory. The reconstructed trajectory can be used to display the UAV's flight path on an electronic map or to trigger alarms when the trajectory enters a preset no-fly zone or deviates from the planned route.

[0147] All of the above technical solutions can be combined in any way to form optional embodiments of this application, and will not be described in detail here.

[0148] The UAV trajectory reconstruction method provided in this application acquires the received signals from multiple WiFi passive sensing links and extracts Doppler observations, enabling passive detection of low-altitude UAVs without the need for a dedicated active illumination source, thus reducing system costs and electromagnetic exposure. Based on this, a coarse baseline trajectory is first generated, and then composite weights for each link at each time point are constructed according to this baseline trajectory. This allows the imbalance in observation reliability caused by signal quality differences between different receiving channels and at different times to be quantified as dynamic weights. These composite weights are then used to construct a weighted residual term in the subsequent objective function, ensuring that highly reliable observations occupy a larger proportion in the optimization process while automatically suppressing low-reliability observations. Simultaneously, the baseline deviation term in the objective function ensures that the optimized trajectory does not deviate from the overall trend of the coarse baseline, and the motion continuity constraint term introduces the physical motion continuity of the UAV into the optimization. Combined with the variables to be optimized based on the coarse baseline trajectory, a global optimization solution is performed, significantly improving the accuracy and continuity of the reconstructed trajectory even under conditions of weak echo, strong interference, and severe signal quality degradation, achieving stable and reliable UAV trajectory monitoring.

[0149] In one specific embodiment, three lines are deployed around the drone's flight area. WiFi Bistatic sensing links. Each link includes a commercial WiFi router transmitter and a receiver. The receiver uses the USRP X310 software radio platform. The two channels are equipped with directional antennas, one pointing towards the corresponding transmitter to collect reference signals, and the other pointing towards the UAV's flight area to collect surveillance signals.

[0150] The target of the data acquisition was a non-cooperative UAV. Each receiver acquired baseband signals and transmitted them to the host for processing, with a sampling rate set to 1 MS / s. Reference and monitoring signals were acquired within each processing time window and then processed according to the following procedure: a reference dictionary was constructed using the reference signal; the monitoring signal was projected onto its orthogonal complement space using a projection matrix for clutter suppression; the CAF (Conversion Aspect) was calculated for the clutter-suppressed monitoring and reference signals, and the maximum amplitude was selected within a preset delay search set to obtain the time-Doppler spectrum; the spectrum was decomposed using RPCA (Recursive Path Comparison) to separate low-rank background components and sparse target components; CFAR (Conversion Array Detection) was performed on the sparse target components to extract candidate Doppler points; second-order dynamic programming (DP) was used to search for the optimal path for the candidate points to recover continuous Doppler ridges (e.g., ...). Figure 11 and Figure 12 (as shown) Figure 11 This is a schematic diagram of the Doppler ridge recovery obtained after single-link preprocessing in this application. Figure 12 This is a schematic diagram comparing the Doppler ridge obtained after single-link preprocessing in this application with the actual Doppler ridge.

[0151] Two segments of UAV flight data were selected as validation segments in the experiment. The same parameter settings were used for both segments: Doppler residual scale was set to 5Hz, residual attenuation scale to 20Hz, local ridge quality lower limit parameter was set to 0.85, control point spacing was 40 sampling points, baseline weighting coefficient was 0.6, and smoothing weighting coefficient was 3.0. The effective observation ratios of the Doppler ridges recovered from each link were statistically analyzed. In the first segment, the effective observation ratios for the three links were 77.2%, 84.8%, and 77.8%, respectively; in the second segment, they were 82.8%, 78.8%, and 75.4%, respectively. These data indicate that the observation availability of different links varies significantly during the same flight process, and the ridge quality of the same link also fluctuates over time. Therefore, if all links are equally weighted and fused, low-quality observations will bias the trajectory. This application constructs a composite weighting based on link-level reliability, local ridge quality, and residual attenuation factor, which can effectively reduce the impact of unreliable observations (such as...). Figure 12 (As shown).

[0152] Please see Figure 13 and Figure 14 ,in, Figure 13 To compare the trajectory reconstruction results generated using the first segment of UAV flight data using the method of this application with the coarse baseline trajectory using only EKF and using EKF+RTS, the UAV RTK recorded trajectory is used as the ground truth baseline. Figure 14 The trajectory reconstruction results generated using the second segment of UAV flight data in this application are compared with those generated using only EKF and with coarse baseline trajectories using EKF+RTS, with the UAV RTK recorded trajectory used as the true baseline. It can be seen that the experimental results show that the first segment of data (such as...) Figure 13 The average level error of the method in this application is approximately 0.97m, and the root mean square error is approximately 1.00m; the second segment of data (such as...) Figure 14 The average error of the proposed method is approximately 0.74 m, and the root mean square error is approximately 0.79 m, both significantly lower than the results of pure EKF and EKF+RTS. Averaging the two sets of data, the average error of the proposed method is approximately 0.86 m. Therefore, by combining reliability weighting and control point optimization, this proposed method effectively reduces the disturbances to the trajectory caused by local erroneous Doppler observations and low-quality links while preserving the overall trend of the coarse baseline trajectory, thus obtaining a reconstruction result closer to the actual trajectory.

[0153] To facilitate better implementation of the UAV trajectory reconstruction method of this application, this application also provides a UAV trajectory reconstruction system. Please refer to... Figure 15 , Figure 15 This is a schematic diagram of the structure of a drone trajectory reconstruction system provided in an embodiment of this application. The drone trajectory reconstruction system 200 may include: The signal acquisition module 210 is used to acquire the received signals of multiple WiFi passive sensing links, including reference signals and monitoring signals; The processing module 220 is used to process the reference signals and monitoring signals of each sensing link and extract the Doppler observations of each sensing link at each time. The trajectory generation module 230 is used to generate a coarse reference trajectory based on the Doppler observations of the sensing link at each time. The weight construction module 240 is used to construct the composite weights of each Doppler observation based on the residual between the coarse reference trajectory and the Doppler observation, and in combination with link-level reliability and local observation quality. The setting module 250 is used to select multiple trajectory control points based on the coarse baseline trajectory and use the positions of the trajectory control points as variables to be optimized. Module 260 is used to construct the objective function, which includes a weighted Doppler residual term constructed using composite weights, a baseline residual term relative to the coarse baseline trajectory, and a motion continuity constraint term. The trajectory optimization module 270 is used to optimize the variables to be optimized based on the objective function, obtain the optimized trajectory control points, and generate the UAV reconstructed trajectory based on the optimized trajectory control points.

[0154] Thus, the aforementioned UAV trajectory reconstruction system 200, by acquiring the received signals from multiple WiFi passive sensing links and extracting Doppler observations, achieves passive detection of low-altitude UAVs without the need for a dedicated active illumination source, reducing system costs and electromagnetic exposure. Based on this, a coarse baseline trajectory is first generated, and then composite weights for each link at each time point are constructed according to this baseline trajectory. This allows the imbalance in observation reliability caused by signal quality differences between different receiving channels and at different times to be quantified as dynamic weights. These composite weights are then used to construct a weighted residual term in the subsequent objective function, ensuring that high-reliability observations occupy a larger proportion in the optimization process while low-reliability observations are automatically suppressed. Simultaneously, the baseline deviation term in the objective function ensures that the optimized trajectory does not deviate from the overall trend of the coarse baseline, and the motion continuity constraint term introduces the physical motion continuity of the UAV into the optimization. Combined with trajectory optimization parameters set based on the coarse baseline trajectory, a global optimization solution is performed, significantly improving the accuracy and continuity of the final reconstructed trajectory even under conditions of weak echo, strong interference, and severe signal quality degradation, achieving stable and reliable UAV trajectory monitoring.

[0155] Each unit in the aforementioned UAV trajectory reconstruction system can be implemented entirely or partially through software, hardware, or a combination thereof. These units can be embedded in or independent of the processor within the electronic device, or stored in the memory of the electronic device in software form, so that the processor can invoke and execute the corresponding operations of each unit.

[0156] The drone trajectory reconstruction system 200 can be integrated into a terminal or server that has storage and a processor and thus computing power, or the drone trajectory reconstruction system 200 can be the terminal or server.

[0157] Optionally, this application also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0158] Figure 16 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may be a terminal or a server. Figure 16 As shown, the electronic device 300 includes a processor 301 with one or more processing cores, a memory 302 with one or more computer-readable storage media, and a computer program stored in the memory 302 and executable on the processor. The processor 301 and the memory 302 are electrically connected. Those skilled in the art will understand that the electronic device structure shown in the figures does not constitute a limitation on the electronic device, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0159] The processor 301 is the control center of the electronic device 300. It connects various parts of the electronic device 300 through various interfaces and lines. By running or loading software programs and / or modules stored in the memory 302, and calling data stored in the memory 302, it executes various functions of the electronic device 300 and processes data, thereby performing overall processing of the electronic device 300.

[0160] In this embodiment, the processor 301 in the electronic device 300 loads the instructions corresponding to the processes of one or more computer programs into the memory 302 according to the following steps, and the processor 301 runs the computer programs stored in the memory 302 to realize various functions: The system acquires received signals from multiple WiFi passive sensing links, each link including a reference signal and a monitoring signal; processes the reference and monitoring signals of each sensing link to extract Doppler observations at each time step; generates a coarse baseline trajectory based on the Doppler observations of the sensing links at each time step; constructs a composite weight for the Doppler observations of each sensing link at each time step based on the coarse baseline trajectory; sets trajectory optimization parameters based on the coarse baseline trajectory; constructs an objective function, which includes a weighted residual term constructed using the composite weight and a motion continuity constraint term; optimizes the trajectory optimization parameters based on the objective function, and generates a reconstructed trajectory based on the optimization results.

[0161] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0162] Optional, such as Figure 16 As shown, the electronic device 300 also includes: a display screen 303, a radio frequency circuit 304, an audio circuit 305, an input unit 306, and a power supply 307. The processor 301 is electrically connected to the display screen 303, the radio frequency circuit 304, the audio circuit 305, the input unit 306, and the power supply 307. Those skilled in the art will understand that... Figure 16 The electronic device structure shown does not constitute a limitation on the electronic device and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0163] The display screen 303 can be used to display a graphical user interface (GUI) and receive operation commands generated by the user interacting with the GUI. The display screen 303 may include a display panel and a touch panel. The display panel can be used to display information input by the user or information provided to the user, as well as various graphical user interfaces of the electronic device. These graphical user interfaces can be composed of graphics, text, icons, video, and any combination thereof. The touch panel can be used to collect touch operations performed by the user on or near it (such as operations performed by the user using a finger, stylus, or any suitable object or accessory on or near the touch panel), generate corresponding operation commands, and execute the corresponding program according to the operation commands.

[0164] The radio frequency circuit 304 can be used to transmit and receive radio frequency signals to establish wireless communication with network devices or other electronic devices, and to transmit and receive signals with network devices or other electronic devices.

[0165] Audio circuitry 305 can be used to provide an audio interface between a user and an electronic device via a speaker and a microphone. Audio circuitry 305 converts received audio data into electrical signals, transmits them to the speaker, and the speaker converts them into sound signals for output. Conversely, the microphone converts collected sound signals into electrical signals, which are then received by audio circuitry 305, converted back into audio data, and then processed by processor 301 before being transmitted via radio frequency circuitry 304 to, for example, another electronic device, or output to memory 302 for further processing. Audio circuitry 305 may also include an earphone jack to facilitate communication between peripheral headphones and electronic devices.

[0166] The input unit 306 can be used to receive input numbers, characters, or object feature information (such as fingerprints, irises, facial information, etc.), and to generate keyboard, mouse, joystick, optical, or trackball signal inputs related to user settings and function control.

[0167] Power supply 307 is used to supply power to various components of electronic device 300. Optionally, power supply 307 can be logically connected to processor 301 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system. Power supply 307 may also include one or more DC or AC power supplies, recharging systems, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0168] although Figure 16 As not shown in the diagram, the electronic device 300 may also include a camera, sensor, wireless fidelity module, Bluetooth module, etc., which will not be described in detail here.

[0169] This application also provides a computer-readable storage medium for storing a computer program. This computer-readable storage medium can be applied to an electronic device, and the computer program causes the electronic device to execute the corresponding processes in the drone trajectory reconstruction method of the embodiments of this application; for the sake of brevity, these will not be elaborated further here.

[0170] This application also provides a computer program comprising computer instructions stored in a computer-readable storage medium. The processor of an electronic device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the electronic device to perform the corresponding process in the UAV trajectory reconstruction method of this application. For the sake of brevity, further details are omitted here.

[0171] It should be understood that the processor in this application may be an integrated circuit chip with signal processing capabilities. In implementation, the steps of the above method embodiments can be completed by integrated logic circuits in the processor's hardware or by software instructions. The processor described above may be a general-purpose processor, a digital signal processor (DSP), or a digital signal processor (DSP). DigitalSignalProcessor , DSP Application-Specific Integrated Circuits (ASICs) ApplicationSpecificIntegratedCircu it , ASIC Ready-made programmable gate arrays (OPGs) FieldProgrammableGateArray , FPGAThe processor can be a microprocessor or any conventional processor. The steps of the UAV trajectory reconstruction method disclosed in the embodiments of this application can be directly implemented or executed by a hardware decoding processor, or by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the field. The storage medium is located in memory, and the processor reads information from the memory and combines it with its hardware to complete the steps of the above method.

[0172] It is understood that the memory in the embodiments of this application can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory may be a read-only memory (ROM). Read - OnlyMemory , ROM Programmable read-only memory (PROM) ProgrammableROM , PROM Erasable programmable read-only memory (EPROM) ErasablePROM , EPROM Electrically erasable programmable read-only memory (EEPROM) Electrically EPROM , EEPROM ) or flash memory. Volatile memory can be random access memory (RAM) RandomAccessMemory , RAM It is used as an external cache. Many forms of this are illustrated, but not limited to, descriptions. RAM Available, for example, static random access memory (SRAM) StaticRAM , SRAM Dynamic Random Access Memory (DRAM) DynamicRAM , DRAM Synchronous Dynamic Random Access Memory (SDRAM) SynchronousDRAM , SDRAM Double Data Rate Synchronous Dynamic Random Access Memory (DRAM) DoubleDataRateSDRAM , DDR SDRAM Enhanced Synchronous Dynamic Random Access Memory (ESR) Enhanced SDRAM , ESDRAM Synchronous Link Dynamic Random Access Memory (SRAM) Synchlink DRAM , SLDRAM ) and Direct Memory Bus Random Access Memory (DRAM) DirectRambusRAM , DRRAM It should be noted that the memory of the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0173] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0174] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0175] In the embodiments of this application, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0176] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0177] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0178] In addition, the functional units in this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0179] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause an electronic device (which may be a personal computer or a server) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes: U disk, external hard drive, ROM , RAM Various media that can store program code, such as magnetic disks or optical discs.

[0180] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for reconstructing the trajectory of a drone, characterized in that, include: Acquire received signals from multiple WiFi passive sensing links, the received signals including reference signals and monitoring signals; The reference signal and the monitoring signal of each sensing link are processed to extract Doppler observations of each sensing link at each time step, including: performing clutter suppression on the monitoring signal to obtain a clutter-suppressed monitoring signal; calculating a mutual ambiguity function based on the clutter-suppressed monitoring signal and the reference signal to generate a time-Doppler spectrum at each time step; and extracting the target Doppler ridge from the time-Doppler spectrum as the Doppler observation. A coarse baseline trajectory is generated based on the Doppler observations of the sensing link at each time step; Based on the residuals between the Doppler observations of each sensing link at each time point and the predicted Doppler obtained by mapping the coarse reference trajectory through a bistatic Doppler observation model, a residual attenuation factor is determined; and a composite weight is constructed based on one or more of the effective observation mask, link-level reliability factor, and local ridge quality factor, together with the residual attenuation factor; wherein, the link-level reliability factor is determined based on the effective observation ratio of the corresponding sensing link within the entire time window and / or the coarse reference residual; the local ridge quality factor is determined based on at least one of ridge energy, confidence level, tracking score, and peak ratio; Based on the coarse baseline trajectory, multiple trajectory control points are selected, and the positions of the trajectory control points are used as variables to be optimized. Construct an objective function, which includes a weighted Doppler residual term constructed using the composite weights, a baseline residual term relative to the coarse baseline trajectory, and a motion continuity constraint term; Based on the objective function, the variables to be optimized are optimized to obtain optimized trajectory control points, and the UAV reconstructed trajectory is generated based on the optimized trajectory control points; Extracting the target Doppler ridge from the time-Doppler spectrum includes: The time-Doppler spectrum is decomposed into low-rank background components and sparse target components; Candidate Doppler point detection is performed on the sparse target component; A continuous trajectory search is performed on the detected candidate Doppler points to obtain continuous Doppler ridges.

2. The method according to claim 1, characterized in that, The clutter suppression of the monitoring signal includes: The reference signal and the monitoring signal are divided into multiple corresponding signal blocks; For each of the signal blocks, a reference dictionary is constructed based on multiple discrete time-delay replicas of the reference signal, and an orthogonal complementary projection matrix is ​​constructed based on the reference dictionary; The corresponding monitoring signal block is projected using the orthogonal complementary projection matrix to obtain the monitoring signal after clutter suppression.

3. The method according to claim 1, characterized in that, The step of calculating the mutual ambiguity function based on the clutter-suppressed monitoring signal and the reference signal to generate the time-Doppler spectrum includes: The clutter-suppressed monitoring signal and the reference signal are divided into multiple corresponding processing time windows; For each processing time window and each candidate Doppler frequency, calculate the correlation values ​​corresponding to multiple discrete time delays, and determine the corresponding mutual ambiguity function values ​​based on the correlation values; The mutual ambiguity function values ​​are arranged according to the processing time window and candidate Doppler frequencies to generate the time-Doppler spectrum.

4. The method according to claim 1, characterized in that, The step of decomposing the time-Doppler spectrum into low-rank background components and sparse target components includes: The time-Doppler spectrum is decomposed into low-rank background components and sparse target components using low-rank sparse decomposition.

5. The method according to claim 1, characterized in that, The step of detecting candidate Doppler points in the sparse target component includes: Based on the candidate point detection method, candidate Doppler points are extracted according to the signal features of each point in the sparse target component. The candidate point detection method includes one of constant false alarm rate detection, local peak detection, adaptive threshold detection, or learning-based candidate detection.

6. The method according to claim 1, characterized in that, The step of performing continuous trajectory search on the detected candidate Doppler points to obtain continuous Doppler ridges includes: An optimal path search algorithm is used to search for the optimal path for the candidate Doppler points based on response intensity, time continuity, and Doppler variation smoothness, thereby generating continuous Doppler ridges.

7. The method according to claim 1, characterized in that, The process of generating a coarse baseline trajectory based on the Doppler observations of each of the sensing links at each time step includes: Multiple candidate initial states are generated within the monitoring area; Based on each candidate initial state, recursive state estimation and backward smoothing are performed using the Doppler observations to obtain multiple candidate trajectories. Based on the fitting error between each candidate trajectory and the Doppler observation, one of the candidate trajectories is selected as the coarse reference trajectory.

8. The method according to claim 7, characterized in that, The recursive state estimation includes one of extended Kalman filtering, unscented Kalman filtering, and particle filtering; the backward smoothing includes... RTS One of the smoothing methods, namely, fixed hysteresis smoothing.

9. The method according to claim 1, characterized in that, Based on the coarse baseline trajectory, multiple trajectory control points are selected, and the positions of these control points are used as variables to be optimized, including: Multiple coarse reference trajectory points are sampled from the coarse reference trajectory at preset time intervals; The positions of each coarse reference trajectory point are set to the initial values ​​of the corresponding control point positions.

10. The method according to claim 1, characterized in that, The step of optimizing the variables to be optimized based on the objective function to obtain optimized trajectory control points, and generating a UAV reconstructed trajectory based on the optimized trajectory control points, includes: Based on the objective function, a nonlinear optimization algorithm is used to iteratively optimize the variables to be optimized; Based on the optimized trajectory control points, the initial reconstructed trajectory is obtained; The initial reconstructed trajectory is smoothed to obtain the reconstructed trajectory.

11. The method according to claim 1, characterized in that, The motion continuity constraints include one or more of the following: second-order difference smoothing terms for control points, velocity constraints, and acceleration constraints.

12. A system for reconstructing the trajectory of a drone, characterized in that, include: The signal acquisition module is used to acquire the received signals from multiple WiFi passive sensing links, the received signals including reference signals and monitoring signals; The processing module is used to process the reference signal and the monitoring signal of each sensing link, and extract the Doppler observations of each sensing link at each time, including: performing clutter suppression on the monitoring signal to obtain the clutter-suppressed monitoring signal; calculating a mutual ambiguity function based on the clutter-suppressed monitoring signal and the reference signal to generate a time-Doppler spectrum at each time; and extracting the target Doppler ridge from the time-Doppler spectrum as the Doppler observation. The trajectory generation module is used to generate a coarse reference trajectory based on the Doppler observations of each of the sensing links at each time. The weight construction module is used to determine the residual attenuation factor based on the residual between the Doppler observations of each sensing link at each time point and the predicted Doppler obtained by mapping the coarse reference trajectory through a bistatic Doppler observation model; and to construct composite weights based on one or more of the effective observation mask, link-level reliability factor, and local ridge quality factor, together with the residual attenuation factor; wherein, the link-level reliability factor is determined based on the effective observation ratio and / or coarse reference residual of the corresponding sensing link throughout the entire time window; the local ridge quality factor is determined based on at least one of ridge energy, confidence level, tracking score, and peak ratio; The setting module is used to select multiple trajectory control points based on the coarse reference trajectory, and to use the positions of the trajectory control points as variables to be optimized; A construction module is used to construct an objective function, which includes a weighted Doppler residual term constructed using the composite weights, a baseline residual term relative to the coarse baseline trajectory, and a motion continuity constraint term. The trajectory optimization module is used to optimize the variables to be optimized, obtain optimized trajectory control points, and generate a UAV reconstructed trajectory based on the optimized trajectory control points; Extracting the target Doppler ridge from the time-Doppler spectrum includes: The time-Doppler spectrum is decomposed into low-rank background components and sparse target components; Candidate Doppler point detection is performed on the sparse target component; A continuous trajectory search is performed on the detected candidate Doppler points to obtain continuous Doppler ridges.

13. The reconfiguration system according to claim 12, characterized in that, The reconfiguration system also includes multiple WiFi A transmitter and multiple receivers, each receiver being associated with one of the aforementioned WiFi The transmitter forms a sensing link, and the receiver includes a reference channel and a monitoring channel, which are used to acquire reference signals and monitoring signals, respectively.

14. An electronic device, characterized in that, include: processor; as well as Memory, used to store computer programs; When the computer program is executed by the processor, the processor performs the method as described in any one of claims 1 to 11.

15. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 11.

Images