Unmanned aerial vehicle perception method and system

Abstract

The application discloses a kind of unmanned aerial vehicle sensing method and system.The method comprises: target airspace is detected by first beam pair target airspace, and the target direction of target unmanned aerial vehicle in target airspace is determined;From reference beam codebook, the beamforming vector group matched with target direction is determined, and reference beam codebook includes the association between reference direction and reference beamforming vector group;Reference beamforming vector group is the beamforming weight set solved under the preset constraint condition of beamforming weight factor, which can make reference sensing error reach preset optimization target when;Reference sensing error is the sensing error limit when the position and speed of unmanned aerial vehicle at reference position are sensed under reference system;Second beam is determined by target direction matched beamforming vector group.The technical scheme of the application can realize efficient utilization of communication sensing resource integration under the architecture of large-scale multiple-input multiple-output integrated sensing and communication system without cell.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of communication technology, and in particular to a method and system for sensing unmanned aerial vehicles (UAVs). Background Technology

[0002] With the rapid development of drone technology, its widespread application in civilian and commercial fields has brought about an urgent need for low-altitude safety management. Related solutions for drone detection employ radar detection, collaborative perception and data fusion detection, and visual / infrared sensor-based detection. Radar detection is costly to deploy and operates independently of the communication system, making resource reuse difficult. Collaborative perception and data fusion detection typically have limited fusion capabilities at the signal level, making true data fusion difficult to achieve. Visual / infrared sensor-based detection is greatly affected by ambient light and weather conditions and struggles to provide physical quantities such as distance and speed. When facing long-range or high-speed unauthorized drones, continuous and stable detection is difficult, leading to target loss or inaccurate speed estimation, and failing to effectively meet the actual needs of low-altitude safety management. Summary of the Invention

[0003] This invention provides a UAV perception method and system that can achieve integrated and efficient utilization of communication and perception resources under a non-cellular, large-scale, multi-input, multi-output integrated sensing and communication system architecture. It significantly improves the target detection accuracy, tracking stability, and system robustness of unauthorized UAVs, thereby meeting the actual needs of low-altitude safety management for high-reliability perception.

[0004] According to one aspect of the present invention, a UAV sensing method is provided, the method comprising:

[0005] Under the reference system, the target airspace is detected by the first beam to determine the target location of the target UAV within the target airspace; the reference system is a non-cellular massive MIMO integrated sensing and communication system.

[0006] The beamforming vector set matching the target azimuth is determined from the reference beamcodebook, which includes the correlation between the reference azimuth and the reference beamforming vector set; the reference azimuth is the azimuth angle corresponding to the UAV at the reference position; the reference beamforming vector set is a set of beamforming weights that can achieve a preset optimization target when the reference sensing error is solved under the preset constraint of the beamforming weight factor; the reference sensing error is the sensing error limit when sensing the position and velocity of the UAV at the reference position under the reference system.

[0007] The second beam is determined by the beamforming vector group that matches the target azimuth, so that it can be loaded and used to detect the target airspace during the next detection. The beamforming vector group that matches the target azimuth is used to control the directional convergence of the transmitted signal energy onto the target UAV.

[0008] According to one aspect of the present invention, a drone sensing system is provided, the system performing the following operations:

[0009] Under the reference system, the target airspace is detected by the first beam to determine the target location of the target UAV within the target airspace; the reference system is a non-cellular massive MIMO integrated sensing and communication system.

[0010] The beamforming vector set matching the target azimuth is determined from the reference beamcodebook, which includes the correlation between the reference azimuth and the reference beamforming vector set; the reference azimuth is the azimuth angle corresponding to the UAV at the reference position; the reference beamforming vector set is a set of beamforming weights that can achieve a preset optimization target when the reference sensing error is solved under the preset constraint of the beamforming weight factor; the reference sensing error is the sensing error limit when sensing the position and velocity of the UAV at the reference position under the reference system.

[0011] The second beam is determined by the beamforming vector group that matches the target azimuth, so that it can be loaded and used to detect the target airspace during the next detection. The beamforming vector group that matches the target azimuth is used to control the directional convergence of the transmitted signal energy onto the target UAV.

[0012] The technical solution of this invention, by scanning and detecting the target airspace with the first beam, can quickly narrow down the target search range, avoid wasting energy in irrelevant airspace, significantly improve the efficiency and accuracy of target positioning, ensure that subsequent beam manipulation has a clear target direction, and avoid ineffective consumption of detection resources; the reference beam codebook pre-establishes the association between the reference azimuth and the beamforming vector group, and relies on a non-cellular large-scale multi-input multi-output integrated sensing and communication system to effectively obtain the theoretical value of the perception error for sensing the position and velocity of the UAV, and the beamforming vector group, under the preset constraint of the beamforming weight factor, can make the reference perception error meet the perception error optimization. The target solution yields beamforming weights, enabling precise matching between the target's azimuth and beamforming resources. Subsequent complex real-time calculations are unnecessary; the reference beamcode can be directly invoked to match a beamforming vector set suitable for the target's azimuth, significantly reducing computational latency while ensuring beamforming accuracy. Furthermore, the second beam determined using the matched beamforming vector set allows for directional focusing of the transmitted signal energy during subsequent detection. Compared to traditional diffused signal transmission, this enhances the signal strength in the target airspace, improves the sensitivity and accuracy of target UAV detection, reduces external interference, and meets the requirements for dynamic tracking and precise detection of UAVs.

[0013] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention, 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 the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a flowchart of a drone perception method provided according to an embodiment of the present invention;

[0016] Figure 2 This is a schematic diagram of a non-cellular, large-scale multiple-input multiple-output integrated sensing and communication system according to an embodiment of the present invention;

[0017] Figure 3 This is a flowchart of another UAV perception method provided according to an embodiment of the present invention. Detailed Implementation

[0018] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0019] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0020] Figure 1 The present invention provides a flowchart of a drone perception method. This embodiment is applicable to the situation of target detection of drones in low-altitude areas. The method can be applied to a drone perception system, which can be implemented in hardware and / or software.

[0021] like Figure 1 As shown, the UAV perception method provided in this embodiment of the invention may include the following processes:

[0022] S110. Under the reference system, the target airspace is detected by the first beam to determine the target location of the target UAV in the target airspace. The reference system is a non-cellular, large-scale multiple-input multiple-output integrated sensing and communication system.

[0023] See Figure 2This application relates to a technology for integrating sensing and communication (ISAC) in a cell-free massive multiple-input multiple-output (CF mMIMO) system. The reference system is a cell-free CF mMIMO integrated sensing and communication system, which adopts a distributed access point architecture. In this system, each access point (AP) can be configured as a transmitting access point for the integrated sensing and communication signal (ISAC) or a receiving access point for the echo signal. The multiple transmitting access points in the cell-free CF mMIMO integrated sensing and communication system include... Multiple antenna transmitters and multiple receiver access points A multi-antenna receiver, the first The transmitter and the first The locations of the receivers are respectively denoted as and All transmitting and receiving access points can be centrally scheduled and managed through core networks such as 5G and edge cloud platforms. For the drone to be detected, the drone's position and velocity state vectors are defined as follows: and .

[0024] See Figure 2 For non-cellular massive MIMO integrated sensing and communication systems, the receiving antenna at the receiving access point... The target received from the first The low-pass equivalent value of the transmitter reflected signal corresponding to each transmit access point is .in, The attenuation coefficient representing channel fading and target scattering characteristics is represented by the Swerling-I sensing model, which means that the radar cross-section (RCS) remains determined but unknown during observation. Indicates the total available energy; Indicates the energy distribution factor; This represents the normalized transmit beamforming weight; This represents clutter plus noise, with a mean of zero and a covariance matrix of... The joint complex Gaussian distribution; Indicates the signal to be transmitted, unit energy. Considered for use in the first Sensing and communication of individual transmitters; Indicates Doppler frequency shift; This indicates the signal propagation delay.

[0025] The drone target is relative to any launcher and receiver The Doppler frequency introduced by the relative radial velocity is calculated as follows: ,in, This represents the unit vector pointing from the drone target to the base station. Indicates wavelength. The transmitter and the first The propagation delay between the receivers is , It is the speed of light.

[0026] Define with The incident (exit) angle of each element is The linear array steering vector is as follows:

[0027] ;

[0028] Among them, unit energy Considered for use in the first In a non-cellular, large-scale multi-input multi-output integrated sensing and communication system, the base station obtains the target's distance and speed by processing the echo of the downlink transmitted signal.

[0029] Assuming a non-cellular, massively multi-input multiple-output integrated sensing and communication system is used There are subcarriers, and each OFDM symbol period is . Frequency interval In a non-cellular, massively multi-input multiple-output integrated sensing and communication system, the transmitter at the access point (AP) continuously transmits data. The OFDM symbol, the first The first base station sent the first The baseband transmitted signal (in the time domain) of one Orthogonal Frequency-Division Multiplexing (OFDM) symbol can be represented as: ;in, Indicates the first The symbol, the first Pilot symbols modulated on each subcarrier; The rectangular window function is defined as a function that operates within the interval [0, 1] [0, 1]. The value is 1 if it is inside, otherwise it is 0. Then the... The complete transmission signal sequence of each base station is as follows: .

[0030] The target airspace can be a low-altitude region, which is defined as airspace with an altitude less than a preset height above the ground. Low-altitude airspace is the primary carrier for unmanned aerial vehicles (UAVs), general aviation, and urban air mobility (UAM). The target UAV can be any UAV entering the target airspace that requires identification and location. For example, the target UAV could be an unregistered UAV flying illegally or a UAV flying legally but requiring tracking. When a target UAV is detected, it acts as a source of echo signal reflection. That is, when the first beam of signal illuminates the back of the target UAV, an echo signal is generated, carrying information such as the target UAV's position and speed. The target azimuth refers to the spatial orientation angle of the target UAV relative to a non-cellular massive MIMO integrated sensing and communication system (e.g., with the center of the access point (AP) array as a reference point).

[0031] Optionally, when conducting the initial detection of the target airspace, the first beam is a preset wide-coverage beam.

[0032] During the initial detection of the target airspace, the first beam is the beam used for initial airspace scanning in a non-cellular, large-scale MIMO integrated sensing and communication system. It is typically a wide-coverage beam (large beamwidth, such as azimuth coverage of 30°-90° and elevation coverage of 10°-30°). By using a wide-coverage beam as the first beam to detect the target airspace and determine the target's location during the initial detection phase, the initial perception loop of airspace scanning, target screening, and location locking can be quickly completed. The wide coverage characteristic allows the first beam to cover the preset full-domain detection range of the non-cellular, large-scale MIMO integrated sensing and communication system, avoiding target misses due to narrow beams. Simultaneously, no complex parameter calculations are required; the target UAV's location can be initially locked based solely on the energy distribution and angle of arrival of the echo signal. This achieves target position perception from scratch with low complexity in the initial detection phase, providing clear azimuth information for subsequent precise beam matching and addressing the core need for rapid target search over a large airspace.

[0033] S120. Determine the beamforming vector set for target azimuth matching from the reference beamcodebook. The reference beamcodebook includes the correlation between the reference azimuth and the reference beamforming vector set. The reference azimuth is the azimuth angle of the UAV at the reference position. The reference beamforming vector set is the set of beamforming weights that can make the reference sensing error reach the preset optimization target under the preset constraint of the beamforming weight factor. The reference sensing error is the sensing error limit when sensing the position and velocity of the UAV at the reference position under the reference system. The reference system is a non-cellular large-scale multiple-input multiple-output integrated sensing and communication system.

[0034] Optionally, the reference sensing error is represented by Cramer-Rao lower bound (CRLB), and the preset optimization objective for the reference sensing error is to minimize the Cramer-Rao lower bound (CRLB); or, the reference sensing error is represented by mutual information, and the preset optimization objective for the reference sensing error is to maximize mutual information; or, the reference sensing error is represented by detection probability, and the preset optimization objective for the reference sensing error is to maximize detection probability; or, the reference sensing error is represented by mean square error, and the preset optimization objective for the reference sensing error is to minimize mean square error.

[0035] As an optional but not limited implementation, the reference beam codebook is generated using the following steps A1-A3:

[0036] Step A1: For each reference location, based on the geometric distribution of the transmit access points and receive access points in space in the reference system, determine the received signal vector of each receive access point to each transmit access point. The received signal vector of each receive access point to each transmit access point is a likelihood function constructed based on discrete sampled signals, obtained by sampling the signals received by each receive access point from each transmit access point in the reference system at a specified sampling rate and effective pulse time width.

[0037] Step A2: Based on the signal model and noise assumptions of the reference system, derive and calculate the reference sensing error when estimating the position and velocity of the UAV at the reference position.

[0038] Step A3: Under the preset constraints of the beamforming weighting factor, the beamforming vector set that enables the reference sensing error to reach the preset optimization target is obtained through numerical iteration, and the azimuth angle corresponding to the reference position is associated with the obtained beamforming vector set to generate a callable reference beam codebook.

[0039] The core of this scheme is to minimize the theoretical lower limit of the reference sensing error under the preset constraint of the beamforming weighting factor, and then solve for the optimal beamforming weighting factor, so as to provide a high-performance beamforming control basis for UAV detection in non-cellular large-scale multiple-input multiple-output integrated sensing and communication (ISAC) systems.

[0040] Within the low-altitude detection coverage area defined by a non-cellular, large-scale multi-input multi-output integrated sensing and communication system, a multi-dimensional reference position scenario is constructed by combining actual control requirements with the flight characteristics of UAVs. Optionally, the airspace can be divided into several discrete location points according to a uniform grid density (the grid density can be dynamically adjusted according to the detection accuracy requirements; the higher the accuracy requirement, the finer the grid division) to obtain multiple reference positions. Alternatively, by referring to common intrusion trajectories of unauthorized UAVs (such as low-altitude skimming, hovering and turning, straight-line penetration, etc.), key access points on the trajectory can be extracted as multiple reference positions. The combination of these two methods ensures that the reference positions can not only cover the entire area without detection blind spots, but also accurately match the actual flight position distribution characteristics of the UAVs.

[0041] For each pre-set reference position, a theoretical received signal model is constructed and the received signal vector is calculated based on the spatial geometric distribution of the transmitting access points (APs) and receiving access points (APs) in a non-cellular massive MIMO integrated sensing and communication system (e.g., known parameters such as the three-dimensional coordinates of each AP and the antenna array arrangement). Optionally, the signal propagation delay is derived based on the relative distance between the target UAV and the transmitting and receiving APs, the signal amplitude attenuation is calculated using an antenna gain model, and the phase offset is determined based on the spatial geometric distribution of the transmitting and receiving APs, ultimately forming a complex received signal vector containing amplitude and phase information. This received signal vector reflects the signal characteristics that the receiver should theoretically acquire at the reference position.

[0042] Based on the established signal model of the system (such as the ISAC signal model of OFDM modulation) and noise assumptions (e.g., additive white Gaussian noise assumption, specifying statistical characteristics such as noise mean and variance), the reference sensing error for estimating the position and velocity of the UAV at each reference position is solved through mathematical derivation and matrix operations. Taking the Cramer-Rao lower bound (CRLB) as the representation of the reference sensing error, and minimizing the CRLB as the preset optimization objective, the Cramer-Rao lower bound can serve as a theoretical lower limit for the error in estimating the position and velocity of the UAV at the reference position. The calculation of the Cramer-Rao lower bound requires first constructing a signal likelihood function to quantify the probability distribution of the observed received signal at a given position and velocity; then, the Fisher information matrix (FIM) is constructed by differentiating the likelihood function, reflecting the information correlation strength between the position and velocity parameters and the received signal; finally, the inverse of the Fisher information matrix is ​​used to obtain the CRLB matrix. The diagonal elements of the CRLB matrix correspond to the Cramer-Rao lower bounds for position and velocity estimation, respectively, while the off-diagonal elements represent the correlation between the estimation errors of the two types of parameters.

[0043] Optionally, taking the reference sensing error as represented by the Cramer-Rao lower bound (CRLB), and taking the minimization of the Cramer-Rao lower bound (CRLB) as an example where the preset optimization objective is to minimize the reference sensing error, the calculation process of the Cramer-Rao lower bound is as follows:

[0044] First, a likelihood function is constructed based on the discrete sampled signal, given a sampling rate. and effective pulse time width Next, for the first The receiver received from the first The signals from each transmitter are sampled as follows:

[0045] ;

[0046] in, The sampling interval is the total number of samples. Define the sampling result vector as ,in .

[0047] In the scenario of detecting unauthorized drone flights, the set of system parameters required for calculation is represented as follows: The set of intermediate parameters required for the calculation is ,in For the set of attenuation coefficient parameters, It is a set of angle parameters.

[0048] Based on this, the log-likelihood function in the single-objective case can be written as:

[0049] ;

[0050] Based on the likelihood function, the dimension can be obtained as follows: The single-objective CRLB matrix is:

[0051] , ;in, for The FIM matrix.

[0052] The final form of CRLB calculation can be expressed as:

[0053] ;

[0054] Next, extract the first four elements from the diagonal of the CRLB matrix. , which correspond to the Cramer-Rhodes lower bound (CRLB) for the position and velocity estimation of the black-flying drone, respectively.

[0055] Under the preset constraints of beamforming weighting factors, numerical iterative algorithms (such as gradient descent and particle swarm optimization) are used to solve for the optimal weighting factors. The iterative process takes the goal of achieving a preset optimization target for the reference sensing error as the objective function. By continuously adjusting the weighting factor values, the corresponding trend of change in the reference sensing error is dynamically calculated until the beamforming weight combination that achieves the preset optimization target is found. Under the preset constraints, numerical iterative solution makes The smallest set of beamforming weighting factors The optimal beamforming weight factor is selected and stored together with the corresponding reference azimuth information to form a callable reference beam codebook.

[0056] Optionally, a one-to-one correspondence can be established between the optimal beamforming weight combination obtained from the solution and the reference azimuth corresponding to the reference position of the beamforming weight combination, and stored in a unified data format to form a callable reference beam codebook. The reference beam codebook contains the mapping relationship between the reference azimuth and the reference beamforming vector group, ensuring that in actual detection, the matching beamforming vector group can be quickly called to generate a directional beam based on the target azimuth obtained in real time.

[0057] As an optional but not limited implementation, determining the target azimuth matching beamforming vector set from the reference beam codebook includes the following steps:

[0058] The target azimuth is matched with the reference azimuth in the reference beam codebook, and the beamforming vector group corresponding to the successfully matched reference azimuth is determined as the beamforming vector group matched with the target azimuth.

[0059] After obtaining the target UAV's azimuth within the target airspace, the angle parameters of the target azimuth are used as the retrieval basis. A comparison is performed in the reference beamcodebook. If the azimuth difference between the reference azimuth and the target azimuth is less than a preset threshold, a successful match is determined. The beamforming vector set associated with the successfully matched reference azimuth is then retrieved from the reference beamcodebook and used as the beamforming vector set for the target azimuth match. If the target azimuth is within the azimuth interval formed by the reference azimuths included in the reference beamcodebook, the reference azimuth corresponding to the azimuth interval containing the target azimuth is determined as the successfully matched reference azimuth. The beamforming vector set associated with the successfully matched reference azimuth is then retrieved from the reference beamcodebook and used as the beamforming vector set for the target azimuth match.

[0060] The reference azimuth in the reference beam codebook is related to the optimal beamforming vector set. After a successful match, the extracted vector set can accurately control the direction of the transmitted signal energy to converge at the target azimuth, avoiding energy waste or detection blind spots caused by beam pointing deviation. Furthermore, there is no need to calculate the beamforming vector set in real time. The appropriate beamforming vector set can be quickly extracted from the codebook through azimuth matching, eliminating the complex numerical iterative solution process, shortening the time interval from determining the target azimuth to generating the beam, and avoiding the large amount of computational resources required for real-time solution of the beamforming vector set.

[0061] S130. The second beam is determined by the beamforming vector group matched with the target azimuth, so that the second beam can be loaded and used to detect the target airspace during the next detection. The beamforming vector group matched with the target azimuth is used to control the directional convergence of the transmitted signal energy to the target UAV.

[0062] The beamforming vector set matched with the target azimuth from the reference beamcodebook is synchronously sent to all transmit access points (APs) in the non-cellular massive MIMO integrated sensing and communication system (such as distributed transmit access points in the non-cellular massive MIMO integrated sensing and communication system). The antenna array of each transmit access point AP forms a second beam according to the beamforming weights included in the beamforming vector set matched with the target azimuth. The phase and amplitude of the transmitted signal of the antenna array of the transmit access point AP are adjusted by the second beam, so that the signal transmitted by the antenna array of the transmit access point AP forms in-phase superposition in the azimuth of the target UAV and phase cancellation in the non-target azimuth, and finally controls the energy of the transmitted signal to be directionally focused on the target UAV.

[0063] The beamforming vector group matched to the target azimuth enables directional energy convergence control, concentrating the second beam on the target azimuth. Compared to the first beam, the signal gain in the target area is significantly improved, and the signal-to-noise ratio of the echo signal is greatly enhanced. This provides a high-quality signal source for subsequent high-precision calculation of the target's position and velocity, and also prevents energy from diffusing into areas without the target UAV, reducing ineffective energy consumption at the transmitter and extending the equipment's endurance. Simultaneously, the second beam provides a detection direction adapted to the target azimuth for the next detection, transitioning the detection process from coarse scanning across the entire domain to precise directional scanning, forming a closed loop of azimuth determination, beam matching, and accurate detection. The technical solution of this application feeds back the target azimuth of the UAV within the target airspace to the beamforming module in real time. The optimal beamforming weights are dynamically selected from the optimized reference beam codebook through a lookup table, forming a closed loop of perception, communication, and control. This achieves synergistic optimization between perception accuracy and communication resource utilization, comprehensively improving the system's tracking performance and energy efficiency.

[0064] The technical solution of this invention, by scanning and detecting the target airspace with the first beam, can quickly narrow down the target search range, avoid wasting energy in irrelevant airspace, significantly improve the efficiency and accuracy of target positioning, ensure that subsequent beam manipulation has a clear target direction, and avoid ineffective consumption of detection resources; the reference beam codebook pre-establishes the association between the reference azimuth and the beamforming vector group, and relies on a non-cellular large-scale multi-input multi-output integrated sensing and communication system to effectively obtain the theoretical value of the perception error for sensing the position and velocity of the UAV, and the beamforming vector group, under the preset constraint of the beamforming weight factor, can make the reference perception error meet the perception error optimization. The target solution yields beamforming weights, enabling precise matching between the target's azimuth and beamforming resources. Subsequent complex real-time calculations are unnecessary; the reference beamcode can be directly invoked to match a beamforming vector set suitable for the target's azimuth, significantly reducing computational latency while ensuring beamforming accuracy. Furthermore, the second beam determined using the matched beamforming vector set allows for directional focusing of the transmitted signal energy during subsequent detection. Compared to traditional diffused signal transmission, this enhances the signal strength in the target airspace, improves the sensitivity and accuracy of target UAV detection, reduces external interference, and meets the requirements for dynamic tracking and precise detection of UAVs.

[0065] Figure 3 This is a flowchart illustrating another UAV perception method provided by an embodiment of the present invention. The technical solution of this embodiment further optimizes the process of detecting the target airspace and determining the target location of the UAV within the target airspace by using the first beam in the aforementioned embodiments, based on the technical solutions of the above embodiments. This embodiment can be combined with various optional solutions in one or more of the above embodiments.

[0066] like Figure 3 As shown, the drone perception method in this embodiment may include the following process:

[0067] S310. Under the reference system, multiple transmitting access points in the reference system use the first beam to transmit integrated sensing and communication signals to the target airspace for detection, and obtain the first echo signal corresponding to each receiving access point in the reference system. The reference system is a non-cellular large-scale multiple-input multiple-output integrated sensing and communication system.

[0068] Optionally, by using a first beam to transmit integrated sensing signals into the target airspace through multiple transmitting access points in the reference system for detection, the first echo signals corresponding to each receiving access point in the reference system can be obtained, which may include:

[0069] Through multiple transmitting access points in the reference system, a first beam is used to transmit a synthetic signal to the target airspace. For each receiving access point in the reference system, multiple second echo signals are received at each receiving access point, and the multiple second echo signals are combined using a preset beamforming vector group corresponding to the first beam to obtain a first echo signal corresponding to each receiving access point. Each second echo signal is the echo signal received by the receiving access point for the synthetic signal transmitted to each transmitting access point.

[0070] In a non-cellular, large-scale MIMO integrated sensing and communication system, multiple transmitting access points (APs) synchronously initiate the detection process. Each AP generates and transmits an integrated sensing signal (ISAC) to the target airspace using a first beam according to preset parameters. The ISAC signal (such as an OFDM signal with pilot) transmitted by the APs has both sensing and communication functions; it is used to illuminate the target airspace to obtain echoes and can also carry communication commands. After being reflected by the target UAV, the ISAC signal is acquired by each receiving access point (AP) within the non-cellular, large-scale MIMO integrated sensing and communication system, forming a first echo signal corresponding to each AP.

[0071] In a non-cellular, large-scale MIMO integrated sensing and communication system, all access points (APs) synchronously receive the echo signals corresponding to the integrated sensing signals reflected by the target UAV. Because the APs in the reference system are distributed, they can acquire echoes from different spatial angles, forming multi-source signal data. Each AP calls a preset beamforming vector corresponding to the first beam to combine multiple second echo signals received by its multiple antennas. This allows the multiple second echo signals received by the multiple antennas to form energy superposition, suppressing interference and improving the signal-to-noise ratio. After beamforming and combining, each AP converts the multi-channel signals from multiple antennas into a single-channel signal, obtaining a more ideal first echo signal. This significantly reduces the complexity of subsequent data transmission and processing. Simultaneously, the single-channel signal retains the core characteristics of the second echo, and the signal-to-noise ratio is significantly improved compared to the original signal, preventing target features from being obscured by excessive noise. Each receiving access point (AP) uploads the pre-processed single-channel signal to the central processing unit of the non-cellular, large-scale MIMO integrated sensing and communication system.

[0072] S320. The first echo signal acquired by each receiving access point is down-converted and sampled to obtain the two-dimensional signal matrix corresponding to each receiving access point. The two-dimensional discrete Fourier transform is performed on the two-dimensional signal matrix corresponding to each receiving access point to obtain the local range-velocity spectrum matrix corresponding to each receiving access point.

[0073] For the first echo signals acquired by each receiving access point, after down-conversion and sampling processing, a two-dimensional signal matrix corresponding to each receiving access point is constructed. For each By performing a two-dimensional discrete Fourier transform (2D DFT), the local range-velocity spectrum matrix corresponding to each receiving access point can be obtained. Local range-velocity spectrum matrix The calculation is as follows:

[0074] ;in, Indicates the distance cell index, corresponding to the time delay. ; Indicates the velocity element index, corresponding to the Doppler frequency shift. .

[0075] This application uses a signal processing scheme based on two-dimensional discrete Fourier transform (2D-DFT) to obtain the range-velocity spectrum matrix. Its advantages lie in its simple and clear algorithm structure, lack of complex redundant design, and suitability for real-time parallel processing of multi-node signals in a distributed non-cellular architecture. This significantly reduces processing latency, meets the real-time requirements for rapid detection and tracking of low-altitude UAVs, and is particularly well-suited to the characteristics of distributed non-cellular architectures. It supports real-time parallel processing of multi-node signals, with each receiving access point independently performing 2D-DFT operations. The results are then rapidly fused through a central controller, significantly shortening the signal processing cycle and effectively reducing overall processing latency. This fully meets the needs of low-altitude UAV detection and tracking scenarios for rapid response and real-time updates, ensuring that targets are not missed or lost.

[0076] Optionally, in addition to using two-dimensional DFT as the joint estimation of the range-velocity spectrum matrix, the present invention can also use the MUSIC spectrum estimation algorithm and the ESPRIT spectrum estimation algorithm to estimate the range-velocity spectrum matrix.

[0077] S330. Based on the local range-velocity spectrum matrix corresponding to each receiving access point, determine the target orientation of the target UAV in the target airspace.

[0078] Optionally, the target bearing of the UAV within the target airspace is determined based on the local range-velocity spectrum matrix corresponding to each receiving access point, including the following steps:

[0079] The local range-velocity spectrum matrices corresponding to each receiving access point are incoherently fused to obtain the global range-velocity spectrum matrix. The distance and velocity of the target UAV in the target airspace are determined by detecting the peak positions in the global range-velocity spectrum matrix. Based on the distance and velocity of the target UAV in the target airspace, the peak values ​​of the range-velocity spectrum of each pair of transmitting and receiving access points, and the geometric distribution of transmitting and receiving access points in space in the reference system, the target azimuth of the target UAV in the target airspace is inversely deduced by azimuth estimation.

[0080] In a non-cellular, large-scale multiple-input multiple-output integrated sensing and communication system, all receiving access points (APs) will use the local range-velocity spectrum matrix of the AP. Uploaded to the central processing unit (CPU), the CPU processes the local range-velocity spectrum matrix of all receiving access points (APs). Perform incoherent fusion to generate a global range-velocity spectrum. Global range-velocity spectrum The calculation is as follows: By superimposing the signal energy of the same distance-velocity coordinate points in the local distance-velocity spectrum of different receiving access points (APs), while preserving the peak characteristics of each local spectrum, this method does not require strict phase synchronization compared to coherent fusion and incoherent fusion. It is more adaptable to the hardware differences of distributed nodes, can effectively aggregate the detection energy of multiple APs, suppress local noise interference, and make the target energy peak in the global spectrum more significant.

[0081] Through testing peak position It can provide a preliminary estimate of the distance to the target UAV within the target airspace. speed as follows: , Each pair of transmitting access points (APs) and receiving access points (APs) has an independent local range-velocity spectrum. The local peak value (i.e., the deviation between the distance and velocity parameters of the local and global range-velocity spectra of each pair of APs is less than a preset threshold) is found. This determines the local distance and local radial velocity between each pair of APs and the target UAV. Furthermore, by combining the geometric distribution of the transmitting and receiving access points in space within the reference system, the target azimuth angle of the UAV can be preliminarily estimated, and the most likely initial position of the target UAV can be deduced. and speed .

[0082] In non-cellular, large-scale MIMO integrated sensing and communication systems, the fusion of sensing signals from multiple distributed access points (APs) can fully leverage the advantages of spatial diversity gain to effectively address the challenges of complex low-altitude environments. On one hand, multipath effects and obstruction in low-altitude environments can easily lead to distortion and energy attenuation of echo signals received by a single AP. Distributed APs receive signals from different spatial locations, and the degree of multipath and obstruction affects each AP differently. Some APs may not receive effective signals due to obstruction, while other unobstructed APs can still capture clear echoes. Signal fusion can filter and aggregate effective signals, suppressing interference from distorted signals. On the other hand, the fusion process performs energy superposition and noise cancellation on the sensing signals from multiple APs, significantly improving the echo signal-to-noise ratio. This not only enables stable target detection in complex environments, avoiding missed or false detections, but also improves the accuracy of target distance, velocity, and orientation estimation.

[0083] S340. Determine the beamforming vector set for target azimuth matching from the reference beamcodebook. The reference beamcodebook includes the correlation between the reference azimuth and the reference beamforming vector set. The reference azimuth is the azimuth angle of the UAV at the reference position. The reference beamforming vector set is the set of beamforming weights that can make the reference sensing error reach the preset optimization target under the preset constraint of the beamforming weight factor. The reference sensing error is the sensing error limit when sensing the position and velocity of the UAV at the reference position under the reference system. The reference system is a non-cellular large-scale multiple-input multiple-output integrated sensing and communication system.

[0084] S350. The second beam is determined by the beamforming vector group matched with the target azimuth, so that the second beam can be loaded and used to detect the target airspace during the next detection. The beamforming vector group matched with the target azimuth is used to control the directional convergence of the transmitted signal energy to the target UAV.

[0085] As an optional but not limited implementation, after determining the second beam using the beamforming vector set matched to the target azimuth, the following steps are also included:

[0086] The beamforming weights corresponding to each transmit access point in the reference system, which are included in the beamforming vector group for target azimuth matching, are distributed to each transmit access point in the reference system so that each transmit access point can detect the target airspace transmitted signal by using the second beam corresponding to the beamforming weight of each transmit access point in the reference system during the next detection.

[0087] The target azimuth matching beamforming vector set is a collection of control parameters corresponding to all transmit access points (APs) in a non-cellular, large-scale MIMO integrated sensing and communication system. Each AP has a unique beamforming weight within the beamforming vector set. The central processing unit (CPU) distributes the corresponding beamforming weights from the target azimuth matching beamforming vector set to each AP individually, based on the AP's identifier. This distribution process relies on the system's low-latency communication link to ensure synchronous reception by all APs, avoiding beam pointing deviations due to differences in arrival time.

[0088] After each AP receives and stores the corresponding beamforming weights, it can directly modulate the transmitted signal using the beamforming weights during the next detection initiation without recalculating or requesting parameters. By adjusting the signal phase and amplitude of its own antenna array, it generates a second beam adapted to the target's azimuth. All APs collaboratively transmit their respective second beams, directing the overall signal energy towards the airspace where the target UAV is located, achieving precise target illumination and efficient detection. For example, the central processing unit sets the beamforming vectors matched to the target's azimuth. The mid-wave beamforming weights are distributed to all transmit access points (APs). Each AP then transmits a second beam using the new beamforming weights during the next detection, concentrating the energy of the second beam in the direction of the target UAV. This process is repeated iteratively until the estimated change in the target UAV's azimuth obtained from each detection is less than a preset threshold, indicating convergence, at which point the iterative detection stops.

[0089] The technical solution of this invention, by scanning and detecting the target airspace with the first beam, can quickly narrow down the target search range, avoid wasting energy in irrelevant airspace, significantly improve the efficiency and accuracy of target positioning, ensure that subsequent beam manipulation has a clear target direction, and avoid ineffective consumption of detection resources; the reference beam codebook pre-establishes the association between the reference azimuth and the beamforming vector group, and relies on a non-cellular large-scale multi-input multi-output integrated sensing and communication system to effectively obtain the theoretical value of the perception error for sensing the position and velocity of the UAV, and the beamforming vector group, under the preset constraint of the beamforming weight factor, can make the reference perception error meet the perception error optimization. The target solution yields beamforming weights, enabling precise matching between the target's azimuth and beamforming resources. Subsequent complex real-time calculations are unnecessary; the reference beamcode can be directly invoked to match a beamforming vector set suitable for the target's azimuth, significantly reducing computational latency while ensuring beamforming accuracy. Furthermore, the second beam determined using the matched beamforming vector set allows for directional focusing of the transmitted signal energy during subsequent detection. Compared to traditional diffused signal transmission, this enhances the signal strength in the target airspace, improves the sensitivity and accuracy of target UAV detection, reduces external interference, and meets the requirements for dynamic tracking and precise detection of UAVs.

[0090] This invention provides a drone perception system, applicable to target detection of drones in low-altitude areas. The drone perception system can be implemented in hardware and / or software. This invention also provides a drone perception method that performs the following operations:

[0091] Under the reference system, the target airspace is detected by the first beam to determine the target location of the target UAV within the target airspace; the reference system is a non-cellular massive MIMO integrated sensing and communication system.

[0092] The beamforming vector set matching the target azimuth is determined from the reference beamcodebook, which includes the correlation between the reference azimuth and the reference beamforming vector set; the reference azimuth is the azimuth angle corresponding to the UAV at the reference position; the reference beamforming vector set is a set of beamforming weights that can achieve a preset optimization target when the reference sensing error is solved under the preset constraint of the beamforming weight factor; the reference sensing error is the sensing error limit when sensing the position and velocity of the UAV at the reference position under the reference system.

[0093] The second beam is determined by the beamforming vector group that matches the target azimuth, so that it can be loaded and used to detect the target airspace during the next detection. The beamforming vector group that matches the target azimuth is used to control the directional convergence of the transmitted signal energy onto the target UAV.

[0094] Based on the above embodiments, optionally, when conducting the first detection of the target airspace, the first beam is a preset wide-coverage beam.

[0095] Based on the above embodiments, optionally, the target airspace is detected by the first beam to determine the target location of the target UAV within the target airspace, including:

[0096] By using the first beam to transmit integrated sensing signals into the target airspace through multiple transmitting access points in the reference system, the first echo signals corresponding to each receiving access point in the reference system are obtained.

[0097] The first echo signal acquired by each receiving access point is down-converted and sampled to obtain the two-dimensional signal matrix corresponding to each receiving access point. The two-dimensional discrete Fourier transform is performed on the two-dimensional signal matrix corresponding to each receiving access point to obtain the local range-velocity spectrum matrix corresponding to each receiving access point.

[0098] Based on the local range-velocity spectrum matrix corresponding to each receiving access point, the target orientation of the target UAV within the target airspace is determined.

[0099] Based on the above embodiments, optionally, by using a first beam to transmit integrated sensing signals into the target airspace through multiple transmitting access points in the reference system, the first echo signals corresponding to each receiving access point in the reference system are obtained, including:

[0100] Through multiple transmission access points in the reference system, a first beam is used to transmit an integrated sensing signal to the target airspace.

[0101] For each receiving access point in the reference system, multiple second echo signals are received at each receiving access point, and the multiple second echo signals are combined using a preset beamforming vector group corresponding to the first beam to obtain a first echo signal corresponding to each receiving access point. Each second echo signal is an echo signal received by the receiving access point for the integrated inductive signal transmitted by the transmitting access point to each transmitting access point.

[0102] Based on the above embodiments, optionally, the target bearing of the target UAV within the target airspace is determined based on the local range-velocity spectrum matrix corresponding to each receiving access point, including:

[0103] The global range-velocity spectrum matrix is ​​obtained by incoherently fusing the local range-velocity spectrum matrices corresponding to each receiving access point.

[0104] The distance and velocity of the target UAV within the target airspace are determined by detecting the peak positions in the global range-velocity spectrum matrix.

[0105] Based on the distance and velocity of the target UAV within the target airspace, the peak distance-velocity spectrum of each pair of transmit access points and receive access points, and the geometric distribution of the transmit access points and receive access points in space in the reference system, the target azimuth of the target UAV within the target airspace is inversely estimated by azimuth angle.

[0106] Based on the above embodiments, optionally, the reference sensing error is represented by the Cramer-Rao lower bound (CRLB), and the preset optimization objective for the reference sensing error is to minimize the Cramer-Rao lower bound (CRLB); or...

[0107] The reference sensing error is represented using mutual information, and the preset optimization objective for the reference sensing error is to maximize the mutual information; or...

[0108] The reference sensing error is represented by a detection probability, and the preset optimization objective for the reference sensing error is to maximize the detection probability; or...

[0109] The reference sensing error is represented by mean square error, and the preset optimization objective for the reference sensing error is to minimize the mean square error.

[0110] Based on the above embodiments, optionally, the reference beam codebook is generated in the following manner:

[0111] For each of the reference locations, based on the geometric distribution of the transmit access points and receive access points in space in the reference system, the received signal vector of each receive access point to each transmit access point is determined. The received signal vector of each receive access point to each transmit access point is a likelihood function constructed based on discrete sampled signals, and is the result of sampling the signals received by each receive access point from each transmit access point in the reference system at a specified sampling rate and effective pulse time width.

[0112] Based on the signal model and noise assumptions of the reference system, the reference perception error for estimating the position and velocity of the UAV at the reference location is derived and calculated.

[0113] Under the preset constraint of beamforming weighting factor, a beamforming vector set that enables the reference sensing error to reach the preset optimization target is obtained through numerical iteration, and the azimuth angle corresponding to the reference position is associated with the obtained beamforming vector set to generate a callable reference beam codebook.

[0114] Based on the above embodiments, optionally, determining the target azimuth matching beamforming vector set from the reference beam codebook includes:

[0115] The target azimuth is matched with the reference azimuth in the reference beam codebook, and the beamforming vector group corresponding to the successfully matched reference azimuth is determined as the beamforming vector group matched with the target azimuth.

[0116] Based on the above embodiments, optionally, after determining the second beam through the beamforming vector group matched to the target azimuth, the method further includes:

[0117] The beamforming weights corresponding to each transmit access point in the reference system, which are included in the beamforming vector group for target azimuth matching, are distributed to each transmit access point in the reference system, so that each transmit access point can detect the target airspace transmitted signal by using the second beam corresponding to the beamforming weight of each transmit access point in the reference system during the next detection.

[0118] The technical solution of this invention, by scanning and detecting the target airspace with the first beam, can quickly narrow down the target search range, avoid wasting energy in irrelevant airspace, significantly improve the efficiency and accuracy of target positioning, ensure that subsequent beam manipulation has a clear target direction, and avoid ineffective consumption of detection resources; the reference beam codebook pre-establishes the association between the reference azimuth and the beamforming vector group, and relies on a non-cellular large-scale multi-input multi-output integrated sensing and communication system to effectively obtain the theoretical value of the perception error for sensing the position and velocity of the UAV, and the beamforming vector group, under the preset constraint of the beamforming weight factor, can make the reference perception error meet the perception error optimization. The target solution yields beamforming weights, enabling precise matching between the target's azimuth and beamforming resources. Subsequent complex real-time calculations are unnecessary; the reference beamcode can be directly invoked to match a beamforming vector set suitable for the target's azimuth, significantly reducing computational latency while ensuring beamforming accuracy. Furthermore, the second beam determined using the matched beamforming vector set allows for directional focusing of the transmitted signal energy during subsequent detection. Compared to traditional diffused signal transmission, this enhances the signal strength in the target airspace, improves the sensitivity and accuracy of target UAV detection, reduces external interference, and meets the requirements for dynamic tracking and precise detection of UAVs.

[0119] The UAV perception system provided in the embodiments of the present invention can execute the UAV perception method provided in any of the embodiments of the present invention, and has the corresponding functions and beneficial effects of executing the UAV perception method. For details, please refer to the relevant operations of the UAV perception method in the foregoing embodiments.

[0120] It is worth noting that the various units and modules included in the above-mentioned device are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be realized; in addition, the specific names of each functional unit are only for easy differentiation and are not used to limit the protection scope of the embodiments of the present invention.

[0121] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0122] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for sensing unmanned aerial vehicles (UAVs), characterized in that, The method includes: Under the reference system, the target airspace is detected by the first beam to determine the target location of the target UAV within the target airspace; the reference system is a non-cellular massive MIMO integrated sensing and communication system. The beamforming vector set matching the target azimuth is determined from the reference beamcodebook, which includes the correlation between the reference azimuth and the reference beamforming vector set; the reference azimuth is the azimuth angle corresponding to the UAV at the reference position; the reference beamforming vector set is a set of beamforming weights that can achieve a preset optimization target when the reference sensing error is solved under the preset constraint of the beamforming weight factor; the reference sensing error is the sensing error limit when sensing the position and velocity of the UAV at the reference position under the reference system. The second beam is determined by the beamforming vector group that matches the target azimuth, so that it can be loaded and used to detect the target airspace during the next detection. The beamforming vector group that matches the target azimuth is used to control the directional convergence of the transmitted signal energy onto the target UAV.

2. The method according to claim 1, characterized in that, When conducting the initial detection of the target airspace, the first beam is a preset wide-coverage beam.

3. The method according to claim 1, characterized in that, The first beam is used to detect the target airspace and determine the target location of the target UAV within the target airspace, including: By using the first beam to transmit integrated sensing signals into the target airspace through multiple transmitting access points in the reference system, the first echo signals corresponding to each receiving access point in the reference system are obtained. The first echo signal acquired by each receiving access point is down-converted and sampled to obtain the two-dimensional signal matrix corresponding to each receiving access point. The two-dimensional discrete Fourier transform is performed on the two-dimensional signal matrix corresponding to each receiving access point to obtain the local range-velocity spectrum matrix corresponding to each receiving access point. Based on the local range-velocity spectrum matrix corresponding to each receiving access point, the target orientation of the target UAV within the target airspace is determined.

4. The method according to claim 3, characterized in that, By transmitting integrated sensing signals into the target airspace using a first beam from multiple transmitting access points in the reference system, the first echo signals corresponding to each receiving access point in the reference system are obtained, including: Through multiple transmission access points in the reference system, a first beam is used to transmit an integrated sensing signal to the target airspace. For each receiving access point in the reference system, multiple second echo signals are received at each receiving access point, and the multiple second echo signals are combined using a preset beamforming vector group corresponding to the first beam to obtain a first echo signal corresponding to each receiving access point. Each second echo signal is an echo signal received by the receiving access point for the integrated inductive signal transmitted by the transmitting access point to each transmitting access point.

5. The method according to claim 3, characterized in that, Based on the local range-velocity spectrum matrix corresponding to each receiving access point, the target bearing of the target UAV within the target airspace is determined, including: The global range-velocity spectrum matrix is ​​obtained by incoherently fusing the local range-velocity spectrum matrices corresponding to each receiving access point. The distance and velocity of the target UAV within the target airspace are determined by detecting the peak positions in the global range-velocity spectrum matrix. Based on the distance and velocity of the target UAV within the target airspace, the peak distance-velocity spectrum of each pair of transmit access points and receive access points, and the geometric distribution of the transmit access points and receive access points in space in the reference system, the target azimuth of the target UAV within the target airspace is inversely estimated by azimuth angle.

6. The method according to claim 1, characterized in that, The reference sensing error is represented by the Cramer-Rao lower bound (CRLB), and the preset optimization objective for the reference sensing error is to minimize the Cramer-Rao lower bound (CRLB); or... The reference sensing error is represented using mutual information, and the preset optimization objective for the reference sensing error is to maximize the mutual information; or... The reference sensing error is represented by a detection probability, and the preset optimization objective for the reference sensing error is to maximize the detection probability; or... The reference sensing error is represented by mean square error, and the preset optimization objective for the reference sensing error is to minimize the mean square error.

7. The method according to claim 1 or 6, characterized in that, The reference beam codebook is generated in the following manner: For each of the reference locations, based on the geometric distribution of the transmit access points and receive access points in space in the reference system, the received signal vector of each receive access point to each transmit access point is determined. The received signal vector of each receive access point to each transmit access point is a likelihood function constructed based on discrete sampled signals, and is the result of sampling the signals received by each receive access point from each transmit access point in the reference system at a specified sampling rate and effective pulse time width. Based on the signal model and noise assumptions of the reference system, the reference perception error for estimating the position and velocity of the UAV at the reference location is derived and calculated. Under the preset constraint of beamforming weighting factor, a beamforming vector set that enables the reference sensing error to reach the preset optimization target is obtained through numerical iteration, and the azimuth angle corresponding to the reference position is associated with the obtained beamforming vector set to generate a callable reference beam codebook.

8. The method according to claim 1 or 6, characterized in that, Determining the target azimuth-matching beamforming vector set from the reference beam codebook includes: The target azimuth is matched with the reference azimuth in the reference beam codebook, and the beamforming vector group corresponding to the successfully matched reference azimuth is determined as the beamforming vector group matched with the target azimuth.

9. The method according to claim 1 or 6, characterized in that, After determining the second beam using the beamforming vector set matched to the target azimuth, the method further includes: The beamforming weights corresponding to each transmit access point in the reference system, which are included in the beamforming vector group for target azimuth matching, are distributed to each transmit access point in the reference system, so that each transmit access point can detect the target airspace transmitted signal by using the second beam corresponding to the beamforming weight of each transmit access point in the reference system during the next detection.

10. A drone sensing system, characterized in that, The system performs the following operations: Under the reference system, the target airspace is detected by the first beam to determine the target location of the target UAV within the target airspace; The reference system is a non-cellular, large-scale, multiple-input multiple-output integrated sensing and communication system. The beamforming vector group matching the target azimuth is determined from the reference beam codebook, wherein the reference beam codebook includes the association between the reference azimuth and the reference beamforming vector group; The reference azimuth is the azimuth angle of the UAV at the reference position, and the reference beamforming vector set is the beamforming weight set that can make the reference sensing error reach the preset optimization target when solved under the preset constraint of the beamforming weight factor. The reference sensing error is the sensing error limit when sensing the position and velocity of the UAV at the reference position under the reference system; The second beam is determined by the beamforming vector group that matches the target azimuth, so that it can be loaded and used to detect the target airspace during the next detection. The beamforming vector group that matches the target azimuth is used to control the directional convergence of the transmitted signal energy onto the target UAV.

Images