Unmanned Aerial Vehicle (UAV) Defense System and Control Method Based on Multi-Sensor Fusion and Adaptive Interference Cancellation

By working in concert with a four-sided phased array radar, a broadband radio frequency detection module, and an optical tracking unit, and combining adaptive interference cancellation technology, the problems of receiver saturation, weak target discrimination capability, and long response time caused by interference in existing anti-drone technologies have been solved, achieving precise interference and rapid response.

CN123074632APending Publication Date: 2026-11-10SHENZHEN EWARE INFORMATION TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610946577.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-06-29
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing anti-drone technologies suffer from problems such as wideband jamming which can easily saturate the receiver itself and paralyze legitimate communications in the surrounding area, weak ability to distinguish and selectively suppress frequency-hopping drone swarms, and long detection and jamming response times in complex electromagnetic environments.

Method used

The system employs a four-sided phased array radar and a broadband radio frequency detection module to perform coarse target correlation and identification. Combined with stereo vision positioning by three optical tracking units, it generates precise interference signals through an adaptive interference cancellation device and uses digital beamforming technology for directional radiation. Simultaneously, it adjusts interference parameters in real time to adapt to the dynamic environment.

Benefits of technology

It achieves precise targeted jamming of specific UAVs, improves the ability to distinguish and selectively suppress multiple targets, shortens the detection-jamming closed-loop response time, and ensures that a suppressive advantage is maintained in dynamic confrontation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123074632A_ABST
    Figure CN123074632A_ABST
Patent Text Reader

Abstract

This invention belongs to the field of low-altitude security and anti-drone defense, and relates to a drone defense system and its control method based on multi-sensor fusion and adaptive interference cancellation. The method includes: establishing a unique identifier for each target through the collaboration of a four-sided phased array radar and a broadband radio frequency detection module; analyzing the drone's uplink signal, extracting parameters such as frequency hopping sequence and modulation type, and constructing a frequency hopping synchronization table and interference copy waveform; fusing radar ranging and velocity measurement with stereo vision information from three circular optical units to obtain the target's three-dimensional coordinates and velocity vector and predict its trajectory; generating a baseband interference cancellation signal with the same amplitude but opposite phase as the original signal, which is then radiated directionally towards the target after digital beamforming; sampling the residual signal, quantifying the interference effect, and adjusting parameters in a closed loop. This achieves precise targeted countermeasures, enhances the ability to distinguish and selectively suppress multiple targets, shortens the detection-interference closed-loop response time, and improves adaptability and reliability in dynamic environments.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of low-altitude security and anti-drone technology, and in particular to a drone defense system and its control method based on multi-sensor fusion and adaptive interference cancellation. Background Technology

[0002] Existing anti-drone technologies have several major drawbacks: broadband jamming can easily lead to receiver saturation and paralysis of legitimate communications in the surrounding area; they have weak ability to distinguish and selectively suppress frequency-hopping and multi-model drone swarms; and they have long detection and jamming response times in complex electromagnetic environments. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides a control method for a drone defense system based on multi-sensor fusion and adaptive interference cancellation, employing the following technical solution, including the following steps: By working in conjunction with a four-sided phased array radar and a broadband radio frequency detection module, suspicious drones can be quickly detected within a 360° range. The initial distance, azimuth, pitch and velocity information of the target can be obtained, and the radar spot and radio frequency detection results can be coarsely correlated to establish a unique identity for each target. Deep analysis of the intercepted UAV uplink radio frequency signal is performed to extract the frequency hopping sequence, hopping time, modulation type and symbol rate parameters, and to construct the frequency hopping synchronization table and interference copy waveform, providing accurate waveform parameters for the subsequent generation of anti-phase cancellation signal; By using the ranging and velocity measurement information of the four-sided phased array radar and the stereo vision positioning information of three optical tracking units arranged in a 120° ring, multi-target association and fusion are performed to obtain the three-dimensional spatial coordinates and velocity vector of the target and predict the target's motion trajectory. Extract the clean UAV uplink reference waveform from the aliased received signal, generate a baseband interference cancellation signal with equal amplitude, opposite phase, and synchronized frequency to the original signal according to the frequency hopping synchronization table and modulation parameters, and adaptively control the transmit power and suppress intermodulation distortion. The digital beamforming weights are calculated based on the target's three-dimensional coordinates obtained from the fusion positioning. The baseband interference cancellation signal is upconverted to the current frequency hopping point and amplified in power before being directionally radiated to the target direction through the array antenna. Simultaneously, a pattern null is formed in the direction of the legitimate communication equipment. While transmitting interference, the system samples the received spatial signal, calculates the interference residual ratio to quantify the interference effect, adjusts the filtering coefficient and transmission parameters of the interference cancellation signal in reverse according to the residual signal, and automatically switches to relearning mode or falls back to omnidirectional scanning state when an abnormal target state or interference failure is detected.

[0004] Preferably, the step of rapidly detecting suspicious UAVs within a 360° range by coordinating the four-sided phased array radar with a broadband radio frequency detection module, acquiring the target's initial distance, azimuth, elevation, and velocity information, and coarsely correlating the radar spot and radio frequency detection results to establish a unique identifier for each target specifically includes: Each array independently controls the T / R component and performs digital beamforming, transmits linear frequency modulated pulse trains, performs pulse compression and constant false alarm rate detection on the received waves, and outputs target traces including range, azimuth angle, elevation angle and radial velocity; The communication frequency band of the UAV is segmented and swept, the power spectrum is calculated and the background noise power estimate is updated in real time. When a burst signal exceeding the threshold is detected, its start frequency, end frequency, bandwidth and duration are recorded and the frequency hopping signal is marked. A three-level association strategy of time window, range window and angle window is adopted to fuse and associate radar spot and radio frequency detection results. When the association conditions are met, a new target track is created and a unique ID is assigned. Targets with only radar spot or only radio frequency signal are marked separately.

[0005] Preferably, the step of performing deep analysis on the intercepted UAV uplink radio frequency signal to extract the frequency hopping sequence, hopping time, modulation type, and symbol rate parameters, and constructing a frequency hopping synchronization table and interference copy waveform to provide accurate waveform parameters for subsequent generation of the anti-phase cancellation signal specifically includes: The intercepted IQ data is input into a digital channelized receiver. The frequency hopping rate is extracted through cyclic spectrum analysis, and the hopping time point and corresponding frequency are identified. A frequency hopping table is established, and a prediction model is used to predict the next hop frequency and hopping time. The high-order cumulant is calculated on the de-hopping baseband signal to identify the modulation type, and symbol synchronization is achieved through timing error detection to obtain accurate symbol period and sampling phase offset; The extracted frequency hopping sequence and modulation parameters are formatted into a structured data table. A frequency hopping waveform storage matrix is ​​constructed in the FPGA to pre-store the copied waveform for each hop and generate a frequency hopping prediction table for real-time retrieval.

[0006] Preferably, the step of using the ranging and velocity measurement information of the four-sided phased array radar and the stereo vision positioning information of three optical tracking units arranged in a 120° ring to perform multi-target association and fusion, obtaining the three-dimensional spatial coordinates and velocity vector of the target, and predicting the target's trajectory specifically includes: Kalman filtering is applied to radar measurements to predict the target's motion state at the next moment, smoothing measurement jumps caused by multipath effects or RCS scintillation, and outputting filtered position and velocity estimates. The target is observed synchronously using three optical units with pre-calibrated intrinsic and extrinsic parameters. An overdetermined linear system is constructed based on the multi-view projection equation, and the three-dimensional coordinates of the target are solved to output the optical positioning results. A cost matrix between radar targets and optical targets is constructed, and the optimal matching relationship is solved using the Hungarian algorithm. Variance-weighted fusion is performed on the successfully matched target pairs, and the fused three-dimensional position information is output.

[0007] Preferably, the steps of extracting a clean UAV uplink reference waveform from the aliased received signal, generating a baseband interference cancellation signal with equal amplitude, opposite phase, and synchronized frequency to the original signal according to the frequency hopping synchronization table and modulation parameters, and adaptively controlling the transmit power and suppressing intermodulation distortion specifically include: The reference signal is extracted from the received signal coupled from the radio frequency detection module, and after frequency domain noise reduction preprocessing, it is used as the desired response. The pure UAV uplink waveform is recovered from the aliased signal by minimum mean square error adaptive filtering. Based on the frequency hopping prediction table and the pure waveform parameters, a local carrier is generated using a direct digital frequency synthesizer. After the baseband waveform is inverted and time-reversed, an interference baseband signal that is opposite to the original signal is transmitted according to the predicted hopping time. The power of the reference signal is measured in real time, the interference transmission power is set to the uplink signal power of the UAV plus a compensation margin, and the third-order intermodulation products at the power amplifier output are monitored. When the intermodulation products exceed the standard, digital predistortion processing is initiated.

[0008] Preferably, the steps of calculating digital beamforming weights based on the target's three-dimensional coordinates obtained from fusion positioning, upconverting the baseband interference cancellation signal to the current frequency hopping point and amplifying its power, and then directionally radiating it to the target direction through an array antenna, while simultaneously forming a pattern null in the direction of the legitimate communication equipment, specifically include: The array steering vector at the working wavelength is calculated based on the three-dimensional coordinates of the target. The optimal complex weight vector is calculated using the minimum variance distortion-free response criterion, so that the main lobe of the array radiation pattern is aligned with the target direction. The baseband interference cancellation signal is digitally up-converted and mixed to output an RF signal, and the propagation delay is pre-compensated according to the target distance. After passing through a variable gain amplifier and a power amplifier, it is fed to the array antenna. A constraint matrix is ​​constructed based on the preset location coordinates of legitimate communication equipment. A minimum variance criterion for linear constraints is applied to form a pattern null in the protection direction, so that interference energy is not radiated in the direction of legitimate equipment.

[0009] Preferably, the steps of sampling the received spatial signal while transmitting interference, calculating the interference residual ratio to quantify the interference effect, adjusting the filtering coefficient and transmission parameters of the interference cancellation signal in reverse according to the residual signal, and automatically switching to relearning mode or falling back to omnidirectional scanning state when an abnormal target state or interference failure is detected, specifically include: While jamming is being transmitted, the radio frequency detection module samples and receives spatial signals, performs de-hopping and coherent accumulation on the sampled signals, measures the residual signal power at the target frequency hopping point, and compares it with the original signal power before jamming is initiated. The jamming residual ratio is then calculated as a quantitative indicator of the jamming effect. The sampled residual signal is down-converted and used as the new desired signal. The adaptive filtering algorithm is rerun with a reduced step size to dynamically update the filtering coefficients to deepen the interference cancellation depth. The current coefficients are locked when the residual ratio improvement is lower than the threshold. The system continuously monitors the residual ratio, demodulation error rate, and target track consistency. When at least two indicators trigger abnormal conditions, the jamming transmission is shut down and the system enters relearning mode to re-extract the frequency hopping spectrum and modulation features. If the relearning is successful, the jamming transmission is restored; otherwise, the system reverts to omnidirectional scanning.

[0010] To address the aforementioned technical problems, this invention also provides a control system for a drone defense system based on multi-sensor fusion and adaptive interference cancellation, employing the control method described above, including: A four-sided phased array radar is used to actively detect targets in the airspace and obtain information on the target's range, angle, and velocity. Three optical tracking units arranged in a 120° ring are used for stereo vision positioning of the target to obtain high-precision angle information and three-dimensional spatial coordinates of the target; The broadband radio frequency detection module is used to passively detect uplink radio frequency signals in the communication frequency band of UAVs and extract the frequency, bandwidth and frequency hopping characteristics of the signals. An adaptive interference cancellation device, comprising an FPGA and a multi-tap FIR filter, is used to extract a reference waveform from the aliased signal, synthesize an interference cancellation signal that is inversely phase to the original signal, and perform digital beamforming and adaptive filter updates. It also includes a central fusion processing unit, which is used to correlate, fuse, and manage the trajectory of multi-source data from radar, optical, and radio frequency detection, and to coordinate the timing and parameter configuration of each module.

[0011] To address the aforementioned technical problems, the present invention also provides a computer device that employs the technical solution described below, comprising a memory and a processor. The memory stores computer-readable instructions, and the processor, when executing the computer-readable instructions, implements the steps of the aforementioned control method for a drone defense system based on multi-sensor fusion and adaptive interference cancellation.

[0012] To address the aforementioned technical problems, the present invention also provides a computer-readable storage medium, which employs the technical solution described below. The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the aforementioned control method for a drone defense system based on multi-sensor fusion and adaptive interference cancellation.

[0013] Compared with the prior art, the present invention has the following main advantages: (1) Overcoming the drawbacks of broadband jamming and achieving precise targeted countermeasures: By deeply analyzing the uplink frequency hopping signal of the UAV, the frequency hopping sequence and modulation parameters are accurately extracted, and a cancellation signal with opposite phase is generated only at the current frequency hopping point of the target, and radiated directionally to the target direction by array beams. This mechanism forms a pattern null in the direction of legitimate equipment, fundamentally avoiding indiscriminate attacks on the surrounding electromagnetic environment and achieving clever jamming of specific targets.

[0014] (2) Enhanced multi-target differentiation and selective suppression capabilities in complex environments: By utilizing a four-sided phased array radar and radio frequency detection to complete coarse correlation and unique identification of targets, and then fusing the stereo vision positioning information of three 120° circumferential optical units, accurate multi-target differentiation and three-dimensional trajectory prediction can be achieved. Combined with beamforming technology, interference waveforms and transmission directions can be independently assigned to different targets, significantly improving the selective suppression effectiveness in cluster scenarios.

[0015] (3) Shorten the detection-jamming closed-loop response time and improve dynamic adaptability: By establishing a frequency hopping synchronization table and jamming copy waveform, the jamming signal is generated and synchronized in real time. At the same time, the residual signal is continuously sampled and the jamming effect is quantified during the jamming transmission process, and the filter coefficient and transmission parameters are adjusted in a closed loop. Once the target state is abnormal or the jamming fails, the system can quickly switch to relearning or omnidirectional scanning state. This full-process adaptive mechanism significantly shortens the reaction time and ensures that the suppressive advantage is always maintained in dynamic confrontation. Attached Figure Description

[0016] To more clearly illustrate the solutions in this invention, the accompanying drawings used in the description of the embodiments of this invention will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a flowchart of an embodiment of the control method for a drone defense system based on multi-sensor fusion and adaptive interference cancellation according to the present invention; Figure 2This is a schematic diagram of the structure of an embodiment of the control device of the UAV defense system based on multi-sensor fusion and adaptive interference cancellation of the present invention; Figure 3 This is a schematic diagram of the structure of an embodiment of the computer device of the present invention. Detailed Implementation

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains; the terminology used herein in the specification is for the purpose of describing particular embodiments only and is not intended to limit the invention; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings are used to distinguish different objects and not to describe a particular order.

[0019] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0020] To enable those skilled in the art to better understand the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0021] It should be noted that the control method of the UAV defense system based on multi-sensor fusion and adaptive interference cancellation provided in the embodiments of the present invention is generally executed by a server / terminal device. Correspondingly, the control device of the UAV defense system based on multi-sensor fusion and adaptive interference cancellation is generally set in the server / terminal device.

[0022] It should be understood that the number of terminal devices, networks, and servers is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be used. Example

[0023] Please refer to Figure 1 The diagram illustrates a flowchart of an embodiment of the control method for a drone defense system based on multi-sensor fusion and adaptive interference cancellation according to the present invention. The control method for the drone defense system based on multi-sensor fusion and adaptive interference cancellation includes the following steps: Step S1: Through the coordinated operation of the four-sided phased array radar and the broadband radio frequency detection module, suspicious drones are quickly detected within a 360° range. The initial distance, azimuth, pitch angle and velocity information of the target are obtained, and the radar spot and radio frequency detection results are coarsely correlated to establish a unique identity for each target.

[0024] In this embodiment, the electronic device (e.g., a server / terminal device) running on the control method of the drone defense system based on multi-sensor fusion and adaptive interference cancellation can receive control requests from the drone defense system based on multi-sensor fusion and adaptive interference cancellation via wired or wireless connection. It should be noted that the aforementioned wireless connection methods may include, but are not limited to, 3G / 4G / 5G connections, WiFi connections, Bluetooth connections, WiMAX connections, Zigbee connections, UWB (ultra-wideband) connections, and other currently known or future known wireless connection methods.

[0025] In this embodiment, step S1 may specifically include the following steps: S11: Each array independently controls the T / R component and performs digital beamforming, transmits linear frequency modulated pulse trains, performs pulse compression and constant false alarm rate detection on the received waves, and outputs target traces including range, azimuth, elevation and radial velocity.

[0026] Each of the four arrays independently controls 64 T / R modules. Each array performs digital beamforming within the horizontal range of 0° to 90° and the elevation range of -10° to 60°. The system uses a 20ms scheduling cycle, during which each array sequentially transmits at a center frequency of f. c A linear frequency modulated pulse train with a frequency response (FFR) of 9.5 GHz and a bandwidth (B) of 50 MHz. Pulse width T p =100μs, pulse repetition period PRI=1ms. After receiving the echo, the IQ data within each range gate is subjected to pulse compression processing: .

[0027] Where: s rc (t) — The complex signal after pulse compression (time domain), whose envelope peak corresponds to the target distance; s rx (τ) — The original echo signal acquired by the receiving antenna, where τ is the integration variable; s tx (t-τ) — A copy of the transmitted signal, consistent with the transmitted linear frequency modulated waveform; * — Complex conjugate operation; t — Current time variable; τ — Integral time variable, ranging from the entire pulse width.

[0028] The signal-to-noise ratio is improved by approximately BT after pulse compression. p Times, where B = 50MHz, T p =100μs, therefore BT p =5000, which is an improvement of approximately 37dB. A constant false alarm rate (CFAR) test is performed on the signal after pulse compression, with the threshold set to V. T =13dB (relative to the noise mean), points exceeding this threshold are considered potential targets. Each point contains range R, azimuth θ, elevation ϕ, and Doppler velocity. .

[0029] The purpose of step S11 is to use electronic scanning instead of mechanical rotation to achieve instantaneous all-round coverage, shortening the scanning cycle from the second level of traditional mechanical radar to the millisecond level, thus laying the foundation for subsequent rapid response.

[0030] S12 performs segmented frequency sweeping of the UAV communication frequency band, calculates the power spectrum and updates the background noise power estimate in real time. When a burst signal exceeding the threshold is detected, its start frequency, end frequency, bandwidth and duration are recorded, and the frequency hopping signal is marked.

[0031] The radio frequency module uses a sampling rate f s =200Msps Segmented frequency sweep of the 20MHz~6GHz band, each segment with a bandwidth of B seg =40MHz, with 50% overlap between adjacent segments. The power spectrum of each frame is calculated using short-time Fourier transform, with a frame length of N. FFT =1024 points, Hanning window suppresses sidelobes. Background noise power estimate. First-order recursive smooth real-time update: .

[0032] in: — Estimated background noise power at the k-th frequency point in the nth frame, in dBm; α — learning rate, a dimensionless constant. We take α = 0.05 to make the background estimation insensitive to sudden signals. P current [k] — Instantaneous power at the k-th frequency point measured in the current frame; — Background noise power estimate from the previous frame;

[0033] When the power at a certain frequency point meets If this state continues for more than 3 consecutive cycles (each cycle lasts for the duration of a complete frequency band sweep, approximately 60ms), then the presence of a drone radio frequency signal is determined. Simultaneously, the starting frequency f of this signal is recorded. start Termination frequency f end Bandwidth BW = f end -f start and signal duration T duration If periodic pulses appear in the same frequency band, they are further marked as frequency hopping signals.

[0034] The purpose of step S12 is to passively detect burst signals in the communication frequency bands (2.4GHz, 5.8GHz, 900MHz, etc.) commonly used by UAVs, to compensate for the radar's insufficient ability to detect low, slow, and small targets (hovering or slow-moving), while avoiding active radiation that exposes its own position.

[0035] S13 employs a three-level association strategy of time window, range window, and angle window to fuse and associate radar traces with radio frequency detection results. When the association conditions are met, a new target track is created and a unique ID is assigned. Targets with only radar traces or only radio frequency signals are marked separately.

[0036] A three-level correlation strategy is adopted, consisting of time window, range window, and angle window. Let the radar detect the target at time t. rad The distance is R rad The azimuth angle is θ rad The radio frequency direction finding unit provides the signal incident direction θ using the amplitude comparison method of the array antenna. rf and angular uncertainty Δθ rf =±3°. The correlation condition is: Time difference: |t rad -t rf |<0.5s; Angular difference: |θ rad -θ rf | <max(5°,2Δθ rf );

[0037] Range consistency: Assuming the radio frequency signals are emitted from the same target, if the radar ranging Rrad and the distance estimated based on the signal strength R... est Satisfy |R rad -R est |<50m.

[0038] When all three conditions are met, a new target track is created and a unique ID is assigned. If only the time difference and angle difference are met but the distance is not, it is marked as a suspected radio frequency-radar correlation and confirmed by the optical unit in subsequent step S3. If there is only a radar track but no corresponding radio frequency signal, it is marked as a silent target without communication and is still sent to step S3 for tracking, but the subsequent jamming strategy adopts incoherent suppression.

[0039] The purpose of step S13 is to fuse the frequency / direction information of radar spot and radio frequency detection, reduce false alarms by utilizing the complementarity of heterogeneous sensors, and establish a unique identifier for each target.

[0040] The purpose of step S1 is to quickly detect suspicious drones within a 360° range, obtain initial distance, orientation, and speed, and trigger subsequent fine-tuning. This step combines active radar detection with passive radio frequency detection to reduce the false alarm rate and missed detection rate of a single sensor.

[0041] Step S2 involves performing in-depth analysis on the intercepted UAV uplink radio frequency signal, extracting the frequency hopping sequence, hopping time, modulation type, and symbol rate parameters, constructing a frequency hopping synchronization table and interference copy waveform, and providing accurate waveform parameters for the subsequent generation of the anti-phase cancellation signal.

[0042] In this embodiment, step S2 may specifically include the following steps: S21, the intercepted IQ data is input into the digital channelization receiver, the frequency hopping rate is extracted through cyclic spectrum analysis and the hopping time point and corresponding frequency are identified, a frequency hopping table is established and the prediction model is used to predict the next hop frequency and hopping time.

[0043] The IQ data intercepted by S12 is input into the digital channelization receiver. The channelization structure employs a polyphase filter bank, N ch =512, covering a total bandwidth of B total =20MHz. The output of each channel is a complex baseband signal. Cyclic spectral analysis is used to detect the cyclic stationary characteristics of the signal to distinguish between frequency-hopping signals and fixed-frequency noise. .

[0044] in: — The cyclic spectral density function of signal x(t) at the cyclic frequency α, in units of power / Hz; α — Cyclic frequency. For frequency-hopping signals, α equals the frequency hopping rate or symbol rate, in Hz. f — Spectral frequency, in Hz; — The cyclic autocorrelation function is defined as: .

[0045] Where x(t) is the received complex baseband signal, τ is the time delay variable, and T is the integration time length. For frequency-hopping GFSK or OFDM signals, A peak value appears where α equals the symbol rate or the reciprocal of the frequency hopping cycle. Extract the α value corresponding to the peak value. peak The frequency hopping rate f can then be obtained. hop =α pea .

[0046] Extracting the jump time point t hop [m] and corresponding frequency f m Establish a frequency hopping table {(f1,t1),(f2,t2),…,(f M ,t M The next-hop frequency is predicted using a first-order hidden Markov model. , in: — The predicted next-hop frequency, in Hz; — A complete set of frequency hopping frequencies (obtained from historical statistics, typically 32-64 frequency points), in Hz; P(f∣f M ,f M-1 — Given the frequencies f of the first two hops M and f M-1 Under the condition that the current hopping frequency is f; f M — The frequency of the current jump; f M-1 — The frequency of the previous jump.

[0047] Transition probability matrix The prediction accuracy was obtained through offline training (reverse engineering of the frequency hopping protocols of common drone models). The prediction accuracy can reach over 95% when the frequency hopping rate is below 1000 hops / second.

[0048] The purpose of step S21 is to identify the hopping pattern between the drone and the remote controller, predict the next hop frequency and hopping time, so that the interference signal can follow the frequency hopping.

[0049] S22 performs high-order cumulative calculations on the de-hopping baseband signal to identify the modulation type, and achieves symbol synchronization through timing error detection to obtain accurate symbol period and sampling phase offset.

[0050] The single-hop baseband signal obtained after de-hopping Perform higher-order cumulants calculations, where Sampling interval ( =1 / fs f s =50MHz), where n is the discrete-time index. Fourth-order cumulant C 42 Defined as: .

[0051] Where: C 42 — A fourth-order second-order cumulant, a dimensionless complex number whose amplitude and phase are sensitive to the modulation type; z[n] — Discrete samples (complex numbers) of the baseband signal after de-hopping; |z[n]| — the amplitude of the sample; E[⋅] — the mathematical expectation operator, which is replaced by time averaging in practical systems. ; N — The number of samples in the statistical window, taken as a multiple of one symbol period, with a typical value of N=1024; z[n] 2 — The square of the sample (complex number operations).

[0052] For BPSK signals, C 42 The real part is approximately -1, and the imaginary part is close to 0; for a QPSK signal, C 42 The real part is approximately 0, and the imaginary part is close to 0; for 16QAM, C 42 The real part is approximately -0.68. The sixth-order cumulants are also calculated to further distinguish OFDM and DSSS.

[0053] Symbol synchronization uses a Gardner timing error detector, and its error signal is calculated using the following formula: .

[0054] Where: e[k] — timing error signal of the kth symbol, dimensionless, positive sign indicates that the sampling time is too late; k — symbolic index, an integer; T s — Symbol period estimate, in seconds, initial values ​​given by coarse spectral analysis; x I (t) — In-phase component of the baseband signal; x Q (t) — the quadrature components of the baseband signal; x I (kT s -T s / 2) — The I-path value at the midpoint of the symbol (between two optimal sampling points); x I (kT s ) and x I ((k-1) Ts — The I-path value of two adjacent optimal sampling points.

[0055] The error signal e[k] is input to a digital loop filter (proportional + integral), and the output controls a numerically controlled oscillator to adjust the sampling clock phase until e[k] converges to near 0. After convergence, the precise symbol period T is obtained. s (opt) and sampling phase offset τ offset .

[0056] The purpose of step S22 is to determine the modulation type (such as BPSK, QPSK, DSSS, OFDM) and precise symbol rate of the UAV uplink signal in order to generate an inverted baseband waveform that is phase-coherent with the original waveform.

[0057] S23, the extracted frequency hopping sequence and modulation parameters are formatted into a structured data table, a frequency hopping waveform storage matrix is ​​constructed in the FPGA, the copied waveform of each hop is pre-stored, and a frequency hopping prediction table is generated for real-time retrieval.

[0058] Construct a block RAM of size N in the FPGA. hop A matrix H of size ×L, where: N hop — Frequency hopping sequence length, set to 64 (covering a complete frequency hopping cycle of a common drone). L — Number of baseband waveform sampling points per hop, calculated using the formula L=T hop ×f s .

[0059] Where T hop The duration of each jump (from t in step S21) hop [m] obtained by difference), f s =50MHz is the baseband sampling rate. Typical value T hop =400μs, then L=20000 points.

[0060] Each row of matrix H stores a copy of the waveform for that jump: For t=n / f s .

[0061] Where: w m [n] — Discrete sample of the replicated waveform for the m-th hop, a complex number; f m — The center frequency of the m-th hop, extracted from step S21, in Hz; ϕ m — The initial phase of the m-th hop is estimated in units of radians through the carrier synchronization loop (Costas loop); m(t) — The synchronized baseband modulation waveform, generated according to the modulation type identified in step S22 (e.g., m(t) is a ±1 sequence in BPSK and ±1±j in QPSK). n — Discrete-time index, with a value range of 0 ≤ n ≤ L-1; t=n / f s — A continuous time variable, in seconds.

[0062] Simultaneously generate a frequency hopping prediction table. ,in t is obtained from the HMM prediction in step S21. next The start time for the next hop is determined (locked via a time-locked loop). The timing prediction accuracy is better than ±0.1μs, ensuring that the error between the inverted signal and the original signal in step S4 is within 1 / 10 of a symbol period.

[0063] The purpose of step S23 is to format the extracted frequency hopping spectrum and modulation parameters into a structured data table for the adaptive interference cancellation device in step S4 to call in real time, ensuring that frequency hopping tracking and inverted waveform generation are completed within microseconds.

[0064] The purpose of step S2 is to perform in-depth analysis on the radio frequency signal intercepted in step S12, extracting features such as frequency hopping sequence, modulation type, and symbol rate, to provide accurate waveform parameters for generating the anti-phase interference signal in step S4. This is the core innovative premise of this invention.

[0065] Step S3: Use the ranging and velocity measurement information of the four-sided phased array radar and the stereo vision positioning information of the three optical tracking units arranged in a 120° ring to perform multi-target association and fusion, obtain the three-dimensional spatial coordinates and velocity vector of the target, and predict the target's motion trajectory.

[0066] In this embodiment, step S3 may specifically include the following steps: S31 performs Kalman filtering recursion on radar measurements to predict the target's motion state at the next moment, smooths measurement jumps caused by multipath effects or RCS scintillation, and outputs filtered position and velocity estimates.

[0067] The radar continuously tracks the same target, and the measurement vector is... , where R k θ is the slant distance. k ϕ is the azimuth angle. k The pitch angle, Let be the radial velocity. The subscript k indicates the k-th time moment. A uniform motion model is established, and the state vector is defined as... , where x k ,y k ,z kThe position is in Cartesian coordinates. This corresponds to the velocity component.

[0068] The Kalman filter iterative process is as follows: Prediction steps: ;

[0069] in: : The state vector at time k predicted based on the estimate at time k-1.

[0070] — The optimal estimated state vector at time k-1.

[0071] F — the state transition matrix, defined as .

[0072] Where Δt is the sampling interval (taken as 0.05s, consistent with the radar frame rate).

[0073] P k|k-1 — Covariance matrix of the predicted state (6×6).

[0074] P k-1|k-1 — The covariance matrix of the estimated state at the previous time step.

[0075] Q — Process noise covariance matrix, reflecting the uncertainty of target maneuvering, set as a diagonal matrix Q=diag(0.1,1.0,0.1,1.0,0.1,1.0), with units corresponding to the physical quantities in meters. 2 and m 2 / s 2 .

[0076] Update steps: ; ; .

[0077] Where: K k — The Kalman gain matrix (6×4) determines the weights of prediction and measurement.

[0078] H — Measurement matrix (4×6), which maps the state vector to the measurement space: .

[0079] It should be noted that: radial velocity Here, H is approximately equal to and The contribution of this technology should be strictly implemented using an extended Kalman filter.

[0080] R — Measurement noise covariance matrix (4×4), obtained from radar static calibration, typical value R=diag(1.0,0.01) 2 0.01 2 0.1 2 The unit is m. 2 rad 2 rad 2 (m / s) 2 .

[0081] I — 6×6 identity matrix.

[0082] The significance of Kalman filtering is that, after fusing historical information, the root mean square error of position estimation can be reduced from 3m in the original radar measurement to less than 1.2m, and the velocity estimation error can be reduced from 0.5m / s to 0.15m / s.

[0083] The function of step S31 is to make the optimal estimate of radar measurement noise, predict the target state at the next moment, and smooth the jump caused by multipath effect or RCS scintillation.

[0084] S32 uses three optical units with pre-calibrated intrinsic and extrinsic parameters to simultaneously observe the target, constructs an overdetermined linear system based on the multi-view projection equation, solves the three-dimensional coordinates of the target, and outputs the optical positioning results.

[0085] The three optical units have been pre-calibrated for intrinsic and extrinsic parameters. Intrinsic parameter matrix K i (3×3) Including focal length f x ,f y Principal point coordinates c x ,c y And distortion coefficients. Extrinsic parameters include position C in the world coordinate system. i and rotation matrix R i Each unit detects the center point pixel coordinates (u) of the same target. i ,v i The data was obtained through a deep learning object detection network (YOLOv8-Drone Special Edition).

[0086] The 3D reconstruction uses the direct linear transformation method. For the i-th camera, the projection equation can be written as: .

[0087] Where: u i ,v i — The target pixel coordinates observed by the i-th camera (in pixels); K i — The intrinsic parameter matrix of the i-th camera (3×3). , where f x,i ,fy,i c is the focal length in pixels. x,i ,c y,i Principal point coordinates; R i — The rotation matrix (3×3) of the i-th camera rotates the points in the world coordinate system to the camera coordinate system; t i — Translation vector (3×1), the position of the world coordinate system origin in the camera coordinate system; [R i |t i — A 3×4 extrinsic parameter matrix; (X,Y,Z) — The three-dimensional coordinates of the target object in the world coordinate system; The symbol ∼ indicates equality in homogeneous coordinates (difference of one non-zero scale factor).

[0088] Expanding the above equations and eliminating the scale factor, each camera contributes two linear equations. The three cameras provide a total of six equations with three unknowns (X, Y, Z). Construct an overdetermined linear system AP = b, where P = [X, Y, Z]. T Solve using the least squares method: P = (A T A) -1 A T b.

[0089] In the formula, A is a 6×3 matrix and b is a 6×1 vector, both derived from the projection equation.

[0090] To prevent false detections in a single frame, median filtering is applied to the positioning results of five consecutive frames (approximately 0.17 seconds, 30fps) to output a smoothed (X, Y, Z) value. The final optical positioning output achieves an angular accuracy of 0.005° (lateral error <0.09m at 1km) and a distance accuracy better than 0.5m within 500m.

[0091] The purpose of step S32 is to utilize the high angular resolution (better than 0.01°) of the visual sensor to provide better elevation angle accuracy than radar at close range (<1km) and to help solve the problem of ambiguity in radar height measurement of low elevation angle targets.

[0092] S33 constructs the cost matrix between radar targets and optical targets, uses the Hungarian algorithm to solve for the optimal matching relationship, performs variance-weighted fusion on the successfully matched target pairs, and outputs the fused three-dimensional position information.

[0093] Let the radar output N r There are N target trajectories, each containing the latest filtered state (position, velocity); optical output N cThere are N target trajectories, each obtained from the stereo intersection in step S32. A cost matrix C of size N is constructed. r ×N c , where element c ij Defined as: c ij =α⋅d pos (i,j)+β⋅d vel (i,j).

[0094] Where: c ij — The cost of matching the i-th radar target with the j-th optical target, dimensionless; The unit is meters; The unit is meters per second; α, β — weighting coefficients, with α=1 (position weight) and β=0.5 (velocity weight, with units normalized through empirical measurements). (x rad,i ,y rad,i ,z rad,i ) and — The position and velocity of the i-th target on the radar; (x opt,j ,y opt,j ,z opt,j ) and — The position and velocity of the j-th optical target (the velocity is obtained by the position difference between two consecutive frames).

[0095] The Kuhn-Munkres algorithm is used to find the minimum total cost matching for C, obtaining a one-to-one correspondence M={(i1,j1),(i2,j2),…}. Unmatched radar targets are marked as radar-only targets (which may be non-communication UAVs or flocks of birds), and unmatched optical targets are marked as optical-only targets (which may be hovering UAVs with extremely small RCS).

[0096] For successfully matched target pairs, perform sensor fusion: .

[0097] Where: P fus — The fused 3D position vector (x fus ,y fus ,z fus ); P rad — The position vector estimated by radar; P opt — The position vector estimated by optical estimation; σ rad— Standard deviation of radar position measurement, typically 3 meters (at 500m). σ opt — Standard deviation of optical position measurement, typically 0.5 meters (at 500m).

[0098] This fusion method is a variance-weighted inverse form, which can be proven to achieve the minimum mean square error under the independent Gaussian noise assumption. The fused position output frequency is 30Hz (synchronized with the optical frame rate) and is directly sent to step S5 for beam control.

[0099] The purpose of step S33 is to distinguish multiple nearby targets, establish a correspondence between radar spot and optical observation for each target, prevent interference signals from pointing to the wrong target, and at the same time, fuse information from the two types of sensors to improve the robustness of positioning.

[0100] The purpose of step S3 is to obtain the target's three-dimensional position (x, y, z) and velocity vector with centimeter-level accuracy. This is used to guide the pointing of the interference beam and compensate for propagation delay in step S5. Radar provides long-range detection and velocity information, while optics provides high angular resolution; the fusion of the two compensates for the shortcomings of a single sensor.

[0101] Step S4: Extract the clean UAV uplink reference waveform from the aliased received signal, generate a baseband interference cancellation signal with equal amplitude, opposite phase and frequency to the original signal according to the frequency hopping synchronization table and modulation parameters, and adaptively control the transmit power and suppress intermodulation distortion.

[0102] In this embodiment, step S4 may specifically include the following steps: S41 extracts the reference signal from the received signal coupled from the RF detection module, and uses it as the desired response after frequency domain noise reduction preprocessing. The pure UAV uplink waveform is then recovered from the aliased signal through minimum mean square error adaptive filtering.

[0103] The RF detection module extracts a coupled signal (coupling degree -20dB) from the receiving path via a directional coupler, and inputs it to the adaptive interference cancellation device after passing through an adjustable attenuator. The device has a built-in FPGA and first performs frequency domain noise reduction preprocessing: dividing the coupled signal x(n) into frames, each frame having a length of N. frame =2048 points, with 50% overlap between adjacent frames. Perform FFT transformation on each frame: .

[0104] Where k is the frequency index. Soft thresholding is applied to the amplitude spectrum |X(k)|, with the threshold set to... ,in This is an estimate of the noise amplitude (obtained statistically from periods without signal). The threshold function is: .

[0105] The phase remains unchanged. Then, the time-domain signal is reconstructed using inverse FFT. , as a clean reference d(n).

[0106] Simultaneously construct the input vector Where M=128 is the filter order. The least mean square algorithm is applied to update the FIR filter coefficients: w(n)=[w0,w1,…,w…]. M-1 ] T : y(n)=w H (n)u(n); e(n) = d(n) - y(n) w(n+1)=w(n)+μu(n) e* (n).

[0107] Where: y(n) — the output (complex value) of the adaptive filter at the nth sampling point; w H w(n) — the conjugate transpose of the filter coefficient vector w(n); u(n) — Input vector (M×1), consisting of historical samples of the reference signal; d(n) — Desired response (the nth sample of the clean reference signal); e(n) — error signal; e*(n) — the complex conjugate of e(n); μ — convergence step size. To maintain stability, μ = 0.01 / (M⋅var(u)), where var(u) is the variance estimate of the input signal. n — Discrete-time index.

[0108] The output y(n) of the LMS filter is the clean UAV uplink waveform recovered from the aliased signal, and its normalized mean square error with the original signal is less than 0.01.

[0109] The purpose of step S41 is to extract a clean reference waveform from the aliased signal (including UAV uplink signal, environmental noise, multipath reflection, etc.) received by the radio frequency detection module as the desired response of the adaptive filter.

[0110] S42, based on the frequency hopping prediction table and the pure waveform parameters, a local carrier is generated using a direct digital frequency synthesizer. After the baseband waveform is inverted and time-reversed, an interference baseband signal that is opposite to the original signal is transmitted according to the predicted transition time.

[0111] The local carrier is generated in parallel by multiple direct digital frequency synthesizers integrated within the FPGA. For the predicted m-th hop, the carrier phase is set to θ. m +π, that is: carrier m (t)=cos(2πf m t+θ m +π).

[0112] Where f m The value of θm is given by step S21, and θm is given by step S23. The baseband waveform is derived from w in step S23. m [n] But perform conjugate inversion: s base,m [n]=-Re {w m * [L-1-n]}.

[0113] Where: s base,m [n] — The baseband interference waveform sample of the m-th hop (real number, for easy DAC output); wm* [L-1-n] — Takes the complex conjugate of the stored copy waveform and reverses the time (L is the number of sampling points per hop, n=0,…,L-1); Re{⋅} — Take the real part; The negative sign achieves a 180° phase inversion.

[0114] The significance of time reversal lies in compensating for the propagation path difference between the detection receiving antenna and the jamming transmitting antenna, aligning the time delay of the jamming waveform with that of the original signal at the target receiver. The final transmitted radio frequency signal is: .

[0115] Where τ pre To pre-compensate for the time delay, it consists of two parts: τ pre =τ prop,self +τ align .

[0116] τ prop,self — The physical delay required for the interference signal to propagate from the transmitting array to the target is 2R / c, where R is the target distance output by the fusion in step S3, and c = 3 × 10⁻⁶. 8 m / s; τ align — Fine-tune the time delay and minimize the residual power after cancellation through closed-loop iteration (adjusted in step S6).

[0117] The accuracy of time synchronization relies on a GPS-disciplined, temperature-controlled crystal oscillator (frequency stability ±0.01ppm) and a delay-locked loop (delay resolution 0.1ns) within the FPGA. The total time alignment error is controlled within T... s (opt)Within / 10, for a 1MHz symbol rate (T s =1μs), which is less than 0.1μs.

[0118] The function of step S42 is to generate an interference baseband signal that is exactly 180° out of phase with the original signal based on the frequency hopping prediction table in S23 and the clean waveform learned in step S41, and to transmit it at the correct hopping time.

[0119] S43 measures the power of the reference signal in real time, sets the interference transmission power to the uplink signal power of the UAV plus a compensation margin, and monitors the third-order intermodulation products at the power amplifier output. When the intermodulation products exceed the standard, digital predistortion processing is initiated.

[0120] The root mean square power P of the reference signal d(n) in real-time measurement step S41 drone : .

[0121] Where N is the total number of samples within one frequency hopping cycle. The interference transmit power P is set. jam For: P jam =P drone +ΔP.

[0122] In the formula, ΔP = 3dB is used to compensate for power loss caused by multipath propagation. This value is adaptively adjusted based on the historical suppression effect, and ΔP is increased (maximum not exceeding 6dB) when the residual ratio η fed back in step S6 is too high.

[0123] To prevent the power amplifier from generating third-order intermodulation products, the spectrum at the output of the power amplifier in the transmit link is monitored in real time. Third-order intermodulation products appear at frequencies 2f1-f2 and 2f2-f1 (f1 and f2 are the main signal frequency components). The power PIM3 at these frequencies is measured. If P... IM3 >P drone If the value drops to -20dB, the digital predistortion module is activated. Predistortion uses a memory polynomial model. .

[0124] Where: y DPD (n) — Baseband signal sample after pre-distortion; x(n) — Original baseband interference signal sample; K — nonlinear order, take K=5; M — Memory depth, take M=3; h k,m — The predistortion coefficients (complex values) are obtained by training the inverse model of the power amplifier offline and are adaptively updated during operation through indirect learning of the structure.

[0125] y DPD(n) After being fed into the DAC and upconversion link, the third-order intermodulation products of the power amplifier output are suppressed by more than 20dB.

[0126] The purpose of step S43 is to ensure that the transmission power of the interference cancellation signal is precisely matched to the power level of the UAV uplink signal (slightly higher by 3dB), so as to avoid over-transmission causing the power amplifier to enter the nonlinear region and generate intermodulation interference, and at the same time to prevent under-transmission from causing incomplete cancellation.

[0127] The purpose of step S4 is to generate an interference cancellation signal with equal amplitude, opposite phase, and synchronized frequency with the UAV's uplink signal, achieving precise suppression without blocking its own reception. This step fully utilizes the frequency hopping spectrum extracted in step S2 and the target information provided in step S3, which is the core innovation of this invention that distinguishes it from traditional blocking jamming.

[0128] Step S5: Calculate the digital beamforming weights based on the target's three-dimensional coordinates obtained from the fusion positioning, upconvert the baseband interference cancellation signal to the current frequency hopping point and amplify its power, then radiate it directionally to the target direction through the array antenna, while simultaneously forming a pattern null in the direction of the legitimate communication equipment.

[0129] In this embodiment, step S5 may specifically include the following steps: S51 calculates the array steering vector at the working wavelength based on the target's three-dimensional coordinates, and uses the minimum variance distortion-free response criterion to calculate the optimal complex weight vector, so that the main lobe of the array radiation pattern is aligned with the target direction.

[0130] The purpose of step S51 is to generate a complex weighted vector pointing to the three-dimensional coordinates of the target, so that the main lobe of the radiation pattern of the array antenna is aligned with the target and the side lobes are as low as possible.

[0131] The adaptive interference cancellation device is connected to a uniform circular array with N=16 elements and a radius r=0.12m, operating in the 2.4~5.8GHz frequency band. For each target direction (θ) t ,ϕ t The array response vector (steering vector) is defined as follows: .

[0132] Where: a(θ) t ,ϕ t — A guide vector of N×1; λ=c / f c — Operating wavelength, f c Given the current frequency hopping frequency, c = 3 × 10⁻⁶. 8 m / s; p i =[x i ,y i ,zi ] T — The position vector of the i-th element (in meters); r t =[sinθ t cosϕ t sinθ t sinϕ t cosθ t ] T — The unit vector in the direction of the target; p i ⋅r t — Dot product, representing the projected distance from the element position to the target direction.

[0133] The optimal weights are calculated using the Minimum Variance Distortionless Response (MVDR) criterion: .

[0134] Where: w opt — The optimal complex weight vector of N×1; R nn — The covariance matrix (N×N) of noise plus interference is estimated by sampling through the receiver link during quiet periods (no interference transmission, only environmental interference): .

[0135] Where: z(p) is the p-th sampling snapshot (the received signal of each array element); a H (θ t ,ϕ t — The conjugate transpose of the directional vector; Denominator a H (θ t ,ϕ t R nn -1 a(θ t ,ϕ t ) is a normalization factor to ensure that the gain for the target direction is 1.

[0136] Calculated w opt The amplitude and phase of the transmitted signal of each array element are controlled via 16 DACs and phase shifters / attenuators. The beamwidth is approximately 20°, and the sidelobes are below -25dB.

[0137] S52 outputs a radio frequency signal after digital up-conversion and mixing of the baseband interference cancellation signal, and performs propagation delay pre-compensation according to the target distance. The signal is then fed to the array antenna after passing through a variable gain amplifier and a power amplifier.

[0138] The baseband signal generated in step S42 (which already includes phase inversion and time reversal) is first modulated to the intermediate frequency by a digital upconverter (DUC), then mixed with the local oscillator (LO) frequency, and output as a radio frequency signal via a bandpass filter.

[0139] The propagation delay is calculated based on the target distance obtained from the fusion in step S33.

[0140] Fractional time delay interpolation is achieved in the digital domain using a Farrow filter, with a delay resolution of 0.1 ns.

[0141] The signal is then passed through a variable gain amplifier (60dB dynamic range, 0.5dB step) and fed into a gallium nitride solid-state power amplifier (maximum output power 20W, linear output power 10W). The output of the power amplifier is connected to a circulator (for isolating transmission and reception) and then sent to a 16-element antenna array.

[0142] Depending on the target distance R, the transmission power can be further adjusted to meet free-space propagation loss: P tx =P jam +20log 10 (R)+20log 10 (f m -147.55, in dBm, simplified form.

[0143] Ensure that the power of the interference signal at the target receiver is approximately 3 dB higher than the original uplink signal.

[0144] The function of step S52 is to upconvert the baseband cancellation signal to the current frequency hopping frequency f. m And perform time delay pre-compensation according to the target distance to ensure that the signal is aligned with the original uplink signal when it arrives at the target.

[0145] S53: Construct a constraint matrix based on the preset location coordinates of legitimate communication equipment, apply the minimum variance criterion of linear constraints, and form a pattern null in the protection direction so that interference energy is not radiated in the direction of legitimate equipment.

[0146] The system pre-enters the location coordinates of nearby legally protected equipment via an electronic compass or pre-configured settings. Assume there are K protection directions, with corresponding azimuth and elevation angles (θ). p,1 ,ϕ p,1 ), (θ p,2 ,ϕ p,2 ), …, (θ p,K ,ϕ p,K Construct the constraint matrix: C=[a(θ)] p,1 ,ϕ p,1 ),a(θ(p,2),ϕ(p,2) ),…,a(θ(p,K),ϕ(p,K))].

[0147] Where a(⋅) is the guiding vector (size N×1) defined in step S51. Apply the minimum variance criterion for linear constraints: st C H w=0.

[0148] Where: w — the weight vector with zero trap constraints to be determined; R xx — The covariance matrix of the received signal (available in R) nn (alternatives) C H w=0 — constraint condition, indicating that the array gain is zero in the protection direction (vector 0 is a K×1 all-zero vector).

[0149] The solution to this constraint minimization problem is: .

[0150] in: .

[0151] — The projection matrix (N×N) onto the column space of C, which is orthogonally filled in. I — N×N identity matrix; w opt — The MVDR weights calculated in step S51.

[0152] Using w null As a transmit weight, the radiated power generated in the protection direction can be attenuated by more than 30 dB compared to the main lobe direction. This null is generated and updated every 100 ms within the FPGA to cope with dynamic environments.

[0153] The purpose of step S53 is to create a pattern null in the direction of known legitimate communication equipment (such as airport towers and prison walkie-talkie base stations) so that interference energy does not radiate in these directions and avoids collateral damage to legitimate communications.

[0154] The purpose of step S5 is to radiate the cancellation signal only in the target direction determined by S3, and to use spatial filtering to concentrate the interference energy at the target angle, so as to avoid causing collateral interference to surrounding legitimate communication equipment (such as airport towers and prison base stations).

[0155] Step S6: While transmitting interference, sample the received spatial signal, calculate the interference residual ratio to quantify the interference effect, adjust the filtering coefficient and transmission parameters of the interference cancellation signal in reverse according to the residual signal, and automatically switch to relearning mode or fall back to omnidirectional scanning state when the target state is detected to be abnormal or the interference fails.

[0156] In this embodiment, step S6 may specifically include the following steps: S61, while transmitting jamming signals, samples and receives spatial signals through the radio frequency detection module, performs de-hopping and coherent accumulation on the sampled signals, measures the residual signal power at the target frequency hopping point, compares it with the original signal power before jamming is initiated, and calculates the jamming residual ratio as a quantitative indicator of the jamming effect.

[0157] While interfering with the transmission, the radio frequency detection module samples and receives spatial signals through a circulator and a limiter (the limiter protects the receiver from being burned out by strong interference). The sampled signal s res (t) includes residual drone signals (if incomplete cancellation) and ambient noise.

[0158] For s res (t) De-hopping and coherent accumulation are performed according to the frequency hopping synchronization table extracted in step S21. At the predicted frequency hopping point f m The measured signal power P res Meanwhile, the original signal power P when memory interference is off. before (Measured in step S41). Define the residual interference ratio: η=10log 10 (P res / P before (Unit: dB)

[0159] Where: η — interference residual ratio, a negative value indicates that the interference is effective (the residual power is lower than the original power). P res — Residual power at the target frequency after interference is enabled (unit: mW or W, ratio is dimensionless). P before — The original drone signal power on the same frequency before interference was enabled.

[0160] When η < -20dB, the interference effect is considered excellent (residual signal is less than 1% of the original signal). When η > -10dB, the suppression is considered insufficient, and adjustment needs to be initiated. Measurements are performed once per frequency hopping cycle (typically 400μs), with a delay of less than 1ms.

[0161] The purpose of step S61 is to quantify the interference effect and provide an objective numerical indicator as a basis for subsequent adjustments.

[0162] S62 takes the sampled residual signal as the new desired signal after down-conversion, and re-runs the adaptive filtering algorithm with a reduced step size to dynamically update the filtering coefficients to deepen the interference cancellation depth, and locks the current coefficients when the residual ratio improvement is lower than the threshold.

[0163] The residual signal e sampled in step S61 RF(n) After down-conversion and digitization, the baseband residual error e is obtained. base (n). In the FPGA, this baseband remnant is used as the new desired signal d. new (n)=e base (n), rerun the LMS algorithm. To avoid coefficient oscillations, use a smaller step size: μ new =0.5⋅μ old .

[0164] Where μ old This is the step size used in step S41. A leakage factor γ = 0.999 is also introduced to prevent coefficient drift: w(n+1) = γw(n) + μ new u(n) e base * (n).

[0165] Where: w(n+1) — the updated filter coefficient vector; γ — Leakage factor (slightly less than 1), which causes the coefficient to decay slowly and prevents it from becoming too large; u(n) — Input vector (from the reference signal, same as step S41); e base * (n) — The complex conjugate of the baseband residual signal.

[0166] Each update iteration is performed 50 times (approximately 50 sampling points, corresponding to 1 μs at a 50 MHz sampling rate), and then η is re-evaluated. If the improvement in η is less than 0.5 dB, the current coefficients are locked and updates are stopped to avoid over-adjustment introducing new non-convergence.

[0167] The purpose of step S62 is to adjust the coefficients of the LMS filter in step S41 in reverse according to the residual signal, so as to further improve the interference cancellation depth, especially to compensate for multipath and environmental changes.

[0168] S63 continuously monitors residual ratio, demodulation bit error rate, and target track consistency. When at least two indicators trigger abnormal conditions, jamming transmission is shut down and relearning mode is entered to re-extract frequency hopping spectrum and modulation features. If relearning is successful, jamming transmission is restored; otherwise, it reverts to omnidirectional scanning state.

[0169] Continuous monitoring of three indicators: Residual ratio η: If η > -3dB and lasts for more than 5 frequency hopping cycles (typically 5 × 400μs = 2ms), it indicates that the interference is almost ineffective.

[0170] Demodulation bit error rate estimation: Attempt to demodulate the residual signal (based on the modulation type in step S22) and calculate the cyclic redundancy check (CRC) error rate BER. est If BERest A value >0.3 indicates that the drone may have switched protocols or been powered off.

[0171] Target trajectory consistency: Calculating the innovation of the Kalman filter: ν k =z k -Hx ̂ k|k-1 .

[0172] Where z k For the current measurement, x ̂ k|k-1 This is the predicted state. If... (S) k (where is the new information covariance matrix), indicating abnormal target movement (possibly the release of a decoy).

[0173] If any two of the above three metrics are triggered simultaneously, the system will enter relearning mode. Immediately shut down the interference transmitter (via RF switch, response time <1μs) to prevent interference with its own receiver; Return to step S2 and re-extract the frequency hopping spectrum and modulation features; Simultaneously, step S3 temporarily increases the weight of the optical sensor (σ). opt (Reduced to 0.3 times the original value), with radar assistance to ensure the target is not lost; If relearning is successful (the frequency hopping pattern is relocked within 500ms), the jamming transmission is automatically restored; if relearning fails three times in a row, it is determined that the target has escaped or shut down, and the system returns to the omnidirectional scanning state in step S1.

[0174] If the target signal completely disappears for more than 5 seconds (determined by the end of the track in S3), the system will automatically return to step S1 and wait for the next target to appear.

[0175] The purpose of step S63 is to enable the system to quickly detect and revert to step S1 or step S2 to relearn when the drone changes frequency hopping mode, shuts down, releases decoy signals, or releases deceptive navigation signals, thus avoiding continuous ineffective interference and wasting power.

[0176] The purpose of step S6 is to detect the residual signal after interference (i.e., the uncancelled UAV uplink signal) through the receiving channel, quantitatively assess the interference effect, and dynamically adjust the filter coefficients of step S4 and the beam parameters of step S5 to achieve optimal suppression. This step enables the system to have self-learning and adaptive capabilities.

[0177] This invention can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This invention can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This invention can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0178] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing related hardware through computer-readable instructions. These computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, or read-only memory (ROM), or random access memory (RAM).

[0179] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0180] Example 2 Further reference Figure 2 As a response to the above Figure 1 The present invention provides an embodiment of a drone defense system based on multi-sensor fusion and adaptive interference cancellation, which is similar to the method shown. Figure 1 Corresponding to the illustrated method embodiment, this system can be specifically applied to various electronic devices. The system includes: A four-sided phased array radar is used to actively detect targets in the airspace and obtain information on the target's range, angle, and velocity. Three optical tracking units arranged in a 120° ring are used for stereo vision positioning of the target to obtain high-precision angle information and three-dimensional spatial coordinates of the target; The broadband radio frequency detection module is used to passively detect uplink radio frequency signals in the communication frequency band of UAVs and extract the frequency, bandwidth and frequency hopping characteristics of the signals. An adaptive interference cancellation device, comprising an FPGA and a multi-tap FIR filter, is used to extract a reference waveform from the aliased signal, synthesize an interference cancellation signal that is inversely phase to the original signal, and perform digital beamforming and adaptive filter updates. It also includes a central fusion processing unit, which is used to correlate, fuse, and manage the trajectory of multi-source data from radar, optical, and radio frequency detection, and to coordinate the timing and parameter configuration of each module.

[0181] In some optional implementations of this embodiment, the adaptive interference cancellation device in this system further includes: Multiple direct digital frequency synthesizers are used to generate local carriers synchronized with the target frequency hopping frequency in parallel; Digital upconverters and mixers are used to modulate baseband interference signals to the radio frequency operating frequency. Variable gain amplifiers and power amplifiers are used for power regulation and amplification of radio frequency signals; In addition, circulators and limiters are used to isolate the transmit and receive channels and protect the receiver from damage caused by strong interference signals.

[0182] In some optional implementations of this embodiment, each array of the four-sided phased array radar independently controls multiple T / R components and performs digital beamforming. The optical tracking unit includes a deep learning target detection network for automatically identifying and tracking UAV targets. The broadband radio frequency detection module covers the commonly used communication frequency bands of UAVs in a segmented frequency sweep manner and has the ability to learn background noise in real time. The central fusion processing unit uses Kalman filtering and Hungarian algorithm to achieve multi-target tracking and association.

[0183] Example 3 To address the aforementioned technical problems, embodiments of the present invention also provide a computer device. Please refer to [link / reference needed]. Figure 3 , Figure 3 This is a basic structural block diagram of the computer device in this embodiment.

[0184] The aforementioned computer device 8 includes a memory 81, a processor 82, and a network interface 83 that are interconnected via a system bus. It should be noted that only the computer device 8 with components 81, 82, and 83 is shown in the figure; however, it should be understood that it is not required to implement all the shown components, and more or fewer components can be implemented alternatively. Those skilled in the art will understand that the computer device described herein is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0185] The aforementioned computer devices can be desktop computers, laptops, handheld computers, and cloud servers, among other computing devices. These devices can facilitate human-computer interaction with users through keyboards, mice, remote controls, touchpads, or voice-activated devices.

[0186] The aforementioned memory 81 includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the aforementioned memory 81 may be an internal storage unit of the aforementioned computer device 8, such as the hard disk or memory of the computer device 8. In other embodiments, the aforementioned memory 81 may also be an external storage device of the aforementioned computer device 8, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 8. Of course, the aforementioned memory 81 may also include both the internal storage unit and its external storage device of the aforementioned computer device 8. In this embodiment, the aforementioned memory 81 is typically used to store the operating system and various application software installed on the aforementioned computer device 8, such as computer-readable instructions for a control method of a drone defense system based on multi-sensor fusion and adaptive interference cancellation. In addition, the aforementioned memory 81 can also be used to temporarily store various types of data that have been output or will be output.

[0187] In some embodiments, the processor 82 may be a central processing unit (CPU), controller, microcontroller, microprocessor, or other data processing chip. The processor 82 is typically used to control the overall operation of the computer device 8. In this embodiment, the processor 82 is used to execute computer-readable instructions stored in the memory 81 or to process data, such as executing computer-readable instructions for the control method of the unmanned aerial vehicle (UAV) defense system based on multi-sensor fusion and adaptive interference cancellation.

[0188] The network interface 83 may include a wireless network interface or a wired network interface, which is typically used to establish a communication connection between the computer device 8 and other electronic devices.

[0189] Example 4 The present invention also provides another embodiment, namely, providing a computer-readable storage medium storing computer-readable instructions that can be executed by at least one processor to cause the at least one processor to perform the steps of the control method for an unmanned aerial vehicle defense system based on multi-sensor fusion and adaptive interference cancellation as described above.

[0190] The beneficial effects of implementing the above embodiments are as follows: (1) Overcoming the drawbacks of broadband jamming and achieving precise targeted countermeasures: By deeply analyzing the uplink frequency hopping signal of the UAV, the frequency hopping sequence and modulation parameters are accurately extracted, and a cancellation signal with opposite phase is generated only at the current frequency hopping point of the target, and radiated directionally to the target direction by array beams. This mechanism forms a pattern null in the direction of legitimate equipment, fundamentally avoiding indiscriminate attacks on the surrounding electromagnetic environment and achieving clever jamming of specific targets.

[0191] (2) Enhanced multi-target differentiation and selective suppression capabilities in complex environments: By utilizing a four-sided phased array radar and radio frequency detection to complete coarse correlation and unique identification of targets, and then fusing the stereo vision positioning information of three 120° circumferential optical units, accurate multi-target differentiation and three-dimensional trajectory prediction can be achieved. Combined with beamforming technology, interference waveforms and transmission directions can be independently assigned to different targets, significantly improving the selective suppression effectiveness in cluster scenarios.

[0192] (3) Shorten the detection-jamming closed-loop response time and improve dynamic adaptability: By establishing a frequency hopping synchronization table and jamming copy waveform, the jamming signal is generated and synchronized in real time. At the same time, the residual signal is continuously sampled and the jamming effect is quantified during the jamming transmission process, and the filter coefficient and transmission parameters are adjusted in a closed loop. Once the target state is abnormal or the jamming fails, the system can quickly switch to relearning or omnidirectional scanning state. This full-process adaptive mechanism significantly shortens the reaction time and ensures that the suppressive advantage is always maintained in dynamic confrontation.

[0193] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0194] Obviously, the embodiments described above are merely some embodiments of the present invention, not all embodiments. The accompanying drawings show preferred embodiments of the present invention, but do not limit the patent scope of the present invention. The present invention can be implemented in many different forms; rather, these embodiments are provided to provide a more thorough and complete understanding of the disclosure of the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of this specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the patent protection scope of this invention.

Claims

1. A control method for a drone defense system based on multi-sensor fusion and adaptive interference cancellation, characterized in that, Includes the following steps: By working in conjunction with a four-sided phased array radar and a broadband radio frequency detection module, suspicious drones can be quickly detected within a 360° range. The initial distance, azimuth, pitch and velocity information of the target can be obtained, and the radar spot and radio frequency detection results can be coarsely correlated to establish a unique identity for each target. Deep analysis of the intercepted UAV uplink radio frequency signal is performed to extract the frequency hopping sequence, hopping time, modulation type and symbol rate parameters, and to construct the frequency hopping synchronization table and interference copy waveform, providing accurate waveform parameters for the subsequent generation of anti-phase cancellation signal; By using the ranging and velocity measurement information of the four-sided phased array radar and the stereo vision positioning information of three optical tracking units arranged in a 120° ring, multi-target association and fusion are performed to obtain the three-dimensional spatial coordinates and velocity vector of the target and predict the target's motion trajectory. Extract the clean UAV uplink reference waveform from the aliased received signal, generate a baseband interference cancellation signal with equal amplitude, opposite phase, and synchronized frequency to the original signal according to the frequency hopping synchronization table and modulation parameters, and adaptively control the transmit power and suppress intermodulation distortion. The digital beamforming weights are calculated based on the target's three-dimensional coordinates obtained from the fusion positioning. The baseband interference cancellation signal is upconverted to the current frequency hopping point and amplified in power before being directionally radiated to the target direction through the array antenna. Simultaneously, a pattern null is formed in the direction of the legitimate communication equipment. While transmitting interference, the system samples the received spatial signal, calculates the interference residual ratio to quantify the interference effect, adjusts the filtering coefficient and transmission parameters of the interference cancellation signal in reverse according to the residual signal, and automatically switches to relearning mode or falls back to omnidirectional scanning state when an abnormal target state or interference failure is detected.

2. The control method for a UAV defense system based on multi-sensor fusion and adaptive interference cancellation as described in claim 1, characterized in that, The steps involved in rapidly detecting suspicious drones within a 360° range by coordinating a four-sided phased array radar with a broadband radio frequency detection module, acquiring the target's initial distance, azimuth, elevation, and velocity information, and coarsely correlating the radar spots with the radio frequency detection results to establish a unique identifier for each target specifically include: Each array independently controls the T / R component and performs digital beamforming, transmits linear frequency modulated pulse trains, performs pulse compression and constant false alarm rate detection on the received waves, and outputs target traces including range, azimuth angle, elevation angle and radial velocity; The communication frequency band of the UAV is segmented and swept, the power spectrum is calculated and the background noise power estimate is updated in real time. When a burst signal exceeding the threshold is detected, its start frequency, end frequency, bandwidth and duration are recorded and the frequency hopping signal is marked. A three-level association strategy of time window, range window and angle window is adopted to fuse and associate radar spot and radio frequency detection results. When the association conditions are met, a new target track is created and a unique ID is assigned. Targets with only radar spot or only radio frequency signal are marked separately.

3. The control method for a UAV defense system based on multi-sensor fusion and adaptive interference cancellation according to claim 1, characterized in that, The steps of performing deep analysis on the intercepted UAV uplink radio frequency signal to extract frequency hopping sequence, hopping time, modulation type and symbol rate parameters, and constructing a frequency hopping synchronization table and interference copy waveform to provide accurate waveform parameters for subsequent generation of anti-phase cancellation signal specifically include: The intercepted IQ data is input into a digital channelized receiver. The frequency hopping rate is extracted through cyclic spectrum analysis, and the hopping time point and corresponding frequency are identified. A frequency hopping table is established, and a prediction model is used to predict the next hop frequency and hopping time. The high-order cumulant is calculated on the de-hopping baseband signal to identify the modulation type, and symbol synchronization is achieved through timing error detection to obtain accurate symbol period and sampling phase offset; The extracted frequency hopping sequence and modulation parameters are formatted into a structured data table. A frequency hopping waveform storage matrix is ​​constructed in the FPGA to pre-store the copied waveform for each hop and generate a frequency hopping prediction table for real-time retrieval.

4. The control method for a UAV defense system based on multi-sensor fusion and adaptive interference cancellation according to claim 1, characterized in that, The steps of using the ranging and velocity measurement information of a four-sided phased array radar and the stereo vision positioning information of three optical tracking units arranged in a 120° ring to perform multi-target correlation and fusion, obtain the three-dimensional spatial coordinates and velocity vector of the target, and predict the target's motion trajectory specifically include: Kalman filtering is applied to radar measurements to predict the target's motion state at the next moment, smoothing measurement jumps caused by multipath effects or RCS scintillation, and outputting filtered position and velocity estimates. The target is observed synchronously using three optical units with pre-calibrated intrinsic and extrinsic parameters. An overdetermined linear system is constructed based on the multi-view projection equation, and the three-dimensional coordinates of the target are solved to output the optical positioning results. A cost matrix between radar targets and optical targets is constructed, and the optimal matching relationship is solved using the Hungarian algorithm. Variance-weighted fusion is performed on the successfully matched target pairs, and the fused three-dimensional position information is output.

5. The control method for a UAV defense system based on multi-sensor fusion and adaptive interference cancellation according to claim 1, characterized in that, The steps of extracting a clean UAV uplink reference waveform from the aliased received signal, generating a baseband interference cancellation signal with equal amplitude, opposite phase, and synchronized frequency to the original signal based on the frequency hopping synchronization table and modulation parameters, and adaptively controlling the transmit power and suppressing intermodulation distortion specifically include: The reference signal is extracted from the received signal coupled from the radio frequency detection module, and after frequency domain noise reduction preprocessing, it is used as the desired response. The pure UAV uplink waveform is recovered from the aliased signal by minimum mean square error adaptive filtering. Based on the frequency hopping prediction table and the pure waveform parameters, a local carrier is generated using a direct digital frequency synthesizer. After the baseband waveform is inverted and time-reversed, an interference baseband signal that is opposite to the original signal is transmitted according to the predicted hopping time. The power of the reference signal is measured in real time, the interference transmission power is set to the uplink signal power of the UAV plus a compensation margin, and the third-order intermodulation products at the power amplifier output are monitored. When the intermodulation products exceed the standard, digital predistortion processing is initiated.

6. The control method for a UAV defense system based on multi-sensor fusion and adaptive interference cancellation according to claim 1, characterized in that, The steps of calculating digital beamforming weights based on the target's three-dimensional coordinates obtained from fusion positioning, upconverting the baseband interference cancellation signal to the current frequency hopping point and amplifying its power, and then directionally radiating it to the target direction through an array antenna, while simultaneously forming a pattern null in the direction of the legitimate communication equipment, specifically include: The array steering vector at the working wavelength is calculated based on the three-dimensional coordinates of the target. The optimal complex weight vector is calculated using the minimum variance distortion-free response criterion, so that the main lobe of the array radiation pattern is aligned with the target direction. The baseband interference cancellation signal is digitally up-converted and mixed to output an RF signal, and the propagation delay is pre-compensated according to the target distance. After passing through a variable gain amplifier and a power amplifier, it is fed to the array antenna. A constraint matrix is ​​constructed based on the preset location coordinates of legitimate communication equipment. A minimum variance criterion for linear constraints is applied to form a pattern null in the protection direction, so that interference energy is not radiated in the direction of legitimate equipment.

7. The control method for a UAV defense system based on multi-sensor fusion and adaptive interference cancellation according to any one of claims 1 to 6, characterized in that, The steps of sampling the received spatial signal while transmitting interference, calculating the interference residual ratio to quantify the interference effect, adjusting the filtering coefficient and transmission parameters of the interference cancellation signal in reverse according to the residual signal, and automatically switching to relearning mode or falling back to omnidirectional scanning state when an abnormal target state or interference failure is detected, specifically include: While jamming is being transmitted, the radio frequency detection module samples and receives spatial signals, performs de-hopping and coherent accumulation on the sampled signals, measures the residual signal power at the target frequency hopping point, and compares it with the original signal power before jamming is initiated. The jamming residual ratio is then calculated as a quantitative indicator of the jamming effect. The sampled residual signal is down-converted and used as the new desired signal. The adaptive filtering algorithm is rerun with a reduced step size to dynamically update the filtering coefficients to deepen the interference cancellation depth. The current coefficients are locked when the residual ratio improvement is lower than the threshold. The system continuously monitors the residual ratio, demodulation error rate, and target track consistency. When at least two indicators trigger abnormal conditions, the jamming transmission is shut down and the system enters relearning mode to re-extract the frequency hopping spectrum and modulation features. If the relearning is successful, the jamming transmission is restored; otherwise, the system reverts to omnidirectional scanning.

8. A control system for a drone defense system based on multi-sensor fusion and adaptive interference cancellation, characterized in that, The control method described in any one of claims 1 to 7 includes: A four-sided phased array radar is used to actively detect targets in the airspace and obtain information on the target's range, angle, and velocity. Three optical tracking units arranged in a 120° ring are used for stereo vision positioning of the target to obtain high-precision angle information and three-dimensional spatial coordinates of the target; The broadband radio frequency detection module is used to passively detect uplink radio frequency signals in the communication frequency band of UAVs and extract the frequency, bandwidth and frequency hopping characteristics of the signals. An adaptive interference cancellation device, comprising an FPGA and a multi-tap FIR filter, is used to extract a reference waveform from the aliased signal, synthesize an interference cancellation signal that is inversely phase to the original signal, and perform digital beamforming and adaptive filter updates. It also includes a central fusion processing unit, which is used to correlate, fuse, and manage the trajectory of multi-source data from radar, optical, and radio frequency detection, and to coordinate the timing and parameter configuration of each module.

9. A computer device, characterized in that, The system includes a memory and a processor, wherein the memory stores computer-readable instructions, and the processor executes the computer-readable instructions to implement the steps of the control method for an unmanned aerial vehicle defense system based on multi-sensor fusion and adaptive interference cancellation as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the control method for an unmanned aerial vehicle (UAV) defense system based on multi-sensor fusion and adaptive interference cancellation as described in any one of claims 1 to 7.