Unmanned aerial vehicle radio frequency countermeasure monitoring system based on multi-source signal state perception

Abstract

The application relates to the technical field of radio orientation, and discloses a UAV radio frequency countermeasure monitoring system based on multi-source signal state sensing, which comprises the following steps: a radio frequency acquisition subsystem synchronously acquires target and reference signals by using a double-channel common-source differential mechanism; a state synchronization bus obtains a UAV motion state vector; a virtual aperture synthesis processor calculates an instantaneous phase difference between channels to offset a local oscillator frequency drift, generates a correction phase sequence, and constructs a virtual antenna array model in combination with a motion trajectory to solve a target space incident angle, the application solves a phase non-coherent problem in the construction of a virtual aperture by low-cost hardware by using a reference channel differential mechanism, and realizes high-precision direction finding of long-wave signals on a small platform.

Description

TECHNICAL FIELD

[0001] The present application relates to a kind of unmanned aerial vehicle radio frequency countermeasure monitoring system based on multi-source signal state perception, belong to radio direction technology field. BACKGROUND

[0002] Current radio monitoring is particularly dependent on the physical aperture of antenna array to measure the direction of high-precision interference signals of radiation source and unmanned aerial vehicle countermeasure equipment, and according to Rayleigh limit, the direction finding resolution is proportional to the array size and inversely proportional to the wavelength. For long-distance radio frequency countermeasure signal monitoring, a meter-level aperture is needed to obtain the phase difference. There are physical contradictions in the monitoring scene of micro unmanned aerial vehicle payload. The platform load, volume and aerodynamic characteristics strictly limit the antenna size and element spacing. The long-wave signal monitoring conflicts with the small carrying platform, making the technology fall into a dilemma. The amplitude direction finding system has low precision and ambiguity. The multi-element interferometer reduces the endurance and is prone to mutual coupling distortion. Platform high-frequency vibration and attitude change worsen the convergence of static direction finding algorithm. The existing technology is difficult to break through the aperture limit without changing the load.

[0003] In addition to the constraints of hardware form, the existing signal processing logic also faces challenges in complex dynamic scenarios. For example, a Chinese patent application with publication number CN120596993A discloses a method and system for identifying and counteracting unmanned aerial vehicles based on radio frequency hardware gene recognition. This scheme introduces deep convolutional neural networks and capsule networks to extract multi-dimensional features of signal time-frequency diagrams to achieve unmanned aerial vehicle identity classification and identification. This is essentially a pattern matching based on signal texture features, focusing on solving target identity attributes rather than high-precision spatial measurement problems. Such methods ignore the nonlinear modulation interference caused by high-frequency vibration and dramatic attitude changes of the monitoring platform on the signal phase, lack fine perception and active compensation mechanisms for platform motion state, and simply relying on neural networks cannot eliminate the Doppler ambiguity and local oscillator frequency drift introduced by platform motion from a physical level. This results in the inability to break through the physical limit of small-aperture platform direction finding accuracy for long-distance long-wavelength signals, making it difficult to meet the high-precision positioning requirements.

[0004] Therefore, it is a technical problem to be solved by the present application to overcome the aperture restriction and motion interference of small unmanned aerial vehicle platforms, achieve high-precision direction finding and rapid positioning of radio frequency countermeasure signals. SUMMARY

[0005] To solve the problems raised in the background art, the technical solution of the present application is as follows: a kind of unmanned aerial vehicle radio frequency countermeasure monitoring system based on multi-source signal state perception, comprising:

[0006] A radio frequency acquisition subsystem includes a first receiving channel and a second receiving channel sharing the same local clock source. The first receiving channel is connected to a monitoring antenna to acquire complex baseband data of target countermeasure signals, and the second receiving channel is connected to a reference antenna to synchronously acquire complex baseband data of external reference signals.

[0007] A state synchronization bus is connected to the UAV flight control system and inertial measurement unit to obtain the UAV motion state vector containing timestamps. The motion state vector includes at least three-dimensional spatial coordinates and three-dimensional velocity vectors.

[0008] The virtual aperture synthesis processor, connected to the RF acquisition subsystem and the status synchronization bus, executes the following data processing logic:

[0009] The instantaneous phase difference between the first receiving channel and the second receiving channel is calculated. The instantaneous phase difference is used to cancel the common-mode frequency drift noise introduced by the local clock source, and a corrected target phase sequence that is time-aligned with the UAV motion state vector is generated.

[0010] Based on the three-dimensional velocity vector, sampling segments with Doppler frequency shift change rate exceeding a preset threshold are selected as effective synthetic aperture intervals, and invalid sampling data that are in uniform linear motion and have constant radial velocity are removed.

[0011] Based on the phase sequence of the corrected target within the effective synthetic aperture range, and according to the spatial trajectory geometric distribution defined by the motion state vector, a virtual antenna array model is constructed.

[0012] Calculate the spatial covariance matrix of the virtual antenna array model, and solve the spatial incident angle of the target countermeasure signal relative to the UAV based on the spatial covariance matrix.

[0013] Preferably, the virtual aperture synthesis processor also performs a geometric autofocus operation, which includes: calculating the virtual spatial spectrum of the external reference signal based on the complex baseband data of the external reference signal and the current UAV motion state vector, and extracting the focus feature index of the virtual spatial spectrum; constructing a trajectory error optimization model with the focus feature index as the objective function and iteratively searching for a trajectory correction amount that makes the focus feature index reach an extreme value; updating the UAV motion state vector using the trajectory correction amount, and constructing a virtual antenna array model for the target countermeasure signal based on the updated UAV motion state vector.

[0014] Preferably, the virtual aperture synthesis processor also performs a multipath consistency screening operation, which includes: dividing the complex baseband data of the first receiving channel into multiple time sub-slices and estimating the instantaneous Doppler frequency of each time sub-slice; constructing a theoretical Doppler frequency response set corresponding to the full-space scanning angle based on the three-dimensional velocity vector; calculating the matching degree between the instantaneous Doppler frequency and the theoretical Doppler frequency response set and removing sub-slice data with a matching degree lower than a preset coherence threshold; and constructing a virtual antenna array model using only the retained sub-slice data.

[0015] Preferably, the virtual aperture synthesis processor also performs a strong interference orthogonal suppression operation, which includes: identifying strong interference signals present in the radio frequency environment and obtaining the interference source azimuth vector of the strong interference signal relative to the UAV trajectory; calculating the theoretical phase response of the strong interference signal at each virtual element of the virtual antenna array model based on the UAV motion state vector to generate an interference steering vector; constructing a projection matrix orthogonal to the interference steering vector and mapping the complex baseband data to the orthogonal complement space of the interference steering vector; and reconstructing the virtual antenna array model using the generated residual data stream.

[0016] Preferably, the state synchronization bus also acquires the high-frequency acceleration components of the UAV body; the virtual aperture synthesis processor calculates the phase error value caused by mechanical vibration using the following phase compensation formula, and uses the phase error value to perform reverse compensation on the correction target phase sequence: ,in, For a moment The phase error value, The carrier wavelength of the target countermeasure signal The line-of-sight vector representing the countermeasure signal from the target. The instantaneous mechanical vibration displacement vector of the antenna phase center relative to the smooth motion trajectory, which is reconstructed by quadratic integration of the high-frequency acceleration components.

[0017] Preferably, the virtual aperture synthesis processor also performs single-station passive ranging operations, which include: performing time unwrapping operations on the phase sequence of the calibration target and calculating the second derivative of the phase sequence with respect to time to extract Doppler frequency modulation features; constructing a radial velocity change rate model containing the distance parameters to be measured based on the three-dimensional velocity vector; and using the Doppler frequency modulation features to perform parameter fitting on the radial velocity change rate model to calculate the radial distance of the target countermeasure signal relative to the UAV.

[0018] Preferably, the virtual aperture synthesis processor combines the calculated radial distance with the spatial incident angle to generate the three-dimensional geographic coordinates of the target countermeasure signal source.

[0019] Preferably, the system also includes a compliance determination module, which extracts the center frequency, bandwidth and modulation type parameters of the target countermeasure signal, and compares the calculated spatial incident angle with a preset database of legitimate transmitter geographical coordinates; when the signal parameters match but the spatial incident angle does not coincide with the coordinates of any legitimate transmitter, an illegal countermeasure signal alarm data is generated.

[0020] Preferably, the external reference signal is one of the following: a broadcast frequency band signal, a cellular network pilot signal, or a downlink signal transmitted by the UAV's own data link; the virtual aperture synthesis processor separates the common-mode frequency error caused by the local clock source by comparing the theoretical Doppler frequency shift and the measured Doppler frequency shift of the external reference signal.

[0021] Preferably, the radio frequency acquisition subsystem adopts a software-defined radio architecture, which includes a direct conversion receiver front-end and an analog-to-digital converter. The analog-to-digital converter directly and synchronously samples the analog baseband signals of the first and second receiving channels.

[0022] Compared with the prior art, the beneficial effects of the present invention are:

[0023] 1. In multi-source signal state sensing, the reference signal is used as the phase anchor point. The local oscillator common-mode frequency drift of the receiver is canceled by dual-channel differential operation, and the phase coherence of long integration time is established. The continuous sampling of the small aperture antenna platform in the time domain is mapped to the virtual large aperture array in the spatial domain. This avoids the limitation of the physical Rayleigh limit on the direction finding resolution and achieves high-precision direction finding of long wavelength signals without relying on a high-stability crystal oscillator.

[0024] 2. Extract the second derivative features of the signal phase to obtain the Doppler frequency modulation features, fit them with the radial velocity change rate model constructed based on the platform motion vector, and utilize the deep phase history information to solve the limitations of traditional direction finding that relies on multi-point intersection or large-scale maneuvering to obtain distance during a single straight flight, thereby improving the positioning response speed of short-term sudden signals.

[0025] 3. Using the three-dimensional velocity vector of the inertial measurement unit as an absolute geometric reference, the consistency of the Doppler history of the radio frequency signal is verified, the matching degree between the measured Doppler trajectory and the theoretical geometric projection is compared, non-line-of-sight multipath signal components are eliminated, and the purity of the virtual aperture data is ensured based on the kinematic constraint screening mechanism. The accuracy of spatial spectrum estimation is maintained in complex electromagnetic environments. The high-frequency acceleration data of the inertial sensor is integrated and reconstructed into the instantaneous small displacement of the antenna phase center. The reverse phase correction factor is generated to clean the baseband signal. The cross-domain compensation mechanism eliminates the interference of platform mechanical vibration on radio frequency phase modulation, prolongs the coherent integration time of the synthetic aperture, and reduces the system's dependence on the mechanical rigidity and vibration reduction structure of the flight platform. Attached Figure Description

[0026] Fig. 1 This is the architecture and processing logic diagram of the UAV radio frequency countermeasure monitoring system with multi-source signal sensing according to the present invention;

[0027] Fig. 2 This is a time-domain comparison curve of the phase error before and after the microelectromechanical vibration decoupling of the present invention;

[0028] Fig. 3 This is a timing flowchart of the dual-channel common-source differential acquisition and phase correction data interaction of the present invention. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0030] This invention discloses a UAV radio frequency countermeasure monitoring system based on multi-source signal state awareness. It consists of three core functional units: a radio frequency acquisition subsystem, a state synchronization bus, and a virtual aperture synthesis processor. The radio frequency acquisition subsystem adopts a software-defined radio architecture, utilizing a direct-conversion receiver front-end and an analog-to-digital converter to synchronously sample analog baseband signals. It includes a first receiving channel and a second receiving channel, both sharing the same local clock source to ensure a strict mathematical correlation between introduced phase noise and frequency drift. The first receiving channel connects to a monitoring antenna to acquire complex baseband data of the target countermeasure signal, while the second receiving channel connects to a reference antenna to synchronously acquire complex baseband data of an external reference signal. The state synchronization bus connects to the UAV flight control system. The system acquires real-time UAV motion state vectors, including timestamps, containing three-dimensional spatial coordinates and three-dimensional velocity vectors. Addressing the high-frequency mechanical vibration issue of the UAV platform, the system employs a phase cleaning mechanism based on microelectromechanical vibration decoupling. The state synchronization bus acquires raw acceleration data from the onboard inertial measurement unit. To distinguish between mechanical noise and effective motion, the system performs frequency domain separation, separating the raw acceleration data into high-frequency acceleration components characterizing body jitter and low-frequency acceleration components characterizing body maneuvering using a second-order high-pass filter with a set cutoff frequency. Subsequently, the virtual aperture synthesis processor performs a time-domain quadratic integration operation on the high-frequency acceleration component to reconstruct the instantaneous mechanical vibration displacement vector of the antenna phase center relative to the smooth motion trajectory. The processor is based on the phase compensation formula Calculate the phase error value caused by mechanical vibration. ,in The carrier wavelength of the target countermeasure signal The phase error value is used to perform reverse compensation on the target phase sequence to eliminate phase noise caused by vibration.

[0031] To address the crystal oscillator frequency drift issue in low-cost receivers, a virtual aperture synthesis processor executes phase denoising differential logic to calculate the instantaneous phase difference between the first and second receiving channels. This counteracts the common-mode frequency drift introduced by the shared local clock source. The external reference signal is selected from broadcast band signals, cellular network pilot signals, or downlink signals transmitted via the UAV's own data link, all exhibiting long-term stability. The processor compares the theoretical and measured Doppler shifts of the external reference signal, separating and eliminating the common-mode frequency error caused by the local clock source. This generates a correction target phase sequence aligned with the UAV's motion state vector time. For non-line-of-sight multipath effects, the virtual aperture synthesis processor performs a multipath consistency screening operation, dividing the complex baseband data of the first receiving channel into time sub-slices ranging from 10 milliseconds to 50 milliseconds. The instantaneous Doppler frequency of each time sub-slice is estimated using short-time Fourier transform. Based on the three-dimensional velocity vector provided by the state synchronization bus, a theoretical Doppler frequency response set corresponding to the full-space scanning angle is constructed. The matching degree between the measured instantaneous Doppler frequency and the theoretical Doppler frequency response set is calculated. Sub-slice data with a matching degree lower than a preset coherence threshold are removed, and only the retained data is used to construct a virtual antenna array model.

[0032] To detect weak signals in a strong interference environment, the system introduces a strong interference orthogonal suppression operation based on a virtual array manifold. The processor identifies strong interference signals in the radio frequency environment, obtains the azimuth vector of the interference source relative to the UAV trajectory, and calculates the theoretical phase response of the strong interference signal at each virtual element of the virtual antenna array model based on the UAV motion state vector. An interference steering vector is generated, and a projection matrix orthogonal to the interference steering vector is constructed. The complex baseband data is mapped to the orthogonal complement space of the interference steering vector, and the virtual antenna array model is reconstructed using the generated residual data stream. The virtual aperture synthesis processor selects sampling segments with Doppler frequency shift change rate exceeding a preset threshold as effective synthetic aperture intervals based on the three-dimensional velocity vector, and removes invalid sampling data with constant radial velocity. The corrected target phase sequence within the effective synthetic aperture interval is used as the data basis, and a virtual antenna array model is constructed based on the spatial trajectory geometric distribution defined by the motion state vector. The spatial covariance matrix of the virtual antenna array model is calculated, and the MUSIC algorithm is used to solve the spatial incident angle of the target countermeasure signal relative to the UAV.

[0033] To eliminate position drift errors, the processor performs geometric autofocus. Based on the complex baseband data of the external reference signal and the current UAV motion state vector, it calculates the virtual spatial spectrum of the external reference signal and extracts the spectral entropy as a focus feature index. A trajectory error optimization model with the focus feature index as the objective function is constructed. The gradient descent method is used to iteratively search for a trajectory correction factor that minimizes the spectral entropy. This trajectory correction factor is used to update the UAV motion state vector, and a virtual antenna array model for the target countermeasure signal is reconstructed based on the updated motion state vector. To achieve single-station positioning, the virtual aperture synthesis processor performs single-station passive ranging, performs time unwrapping operations on the phase sequence of the correction target, and calculates the phase sequence. The system extracts Doppler frequency modulation features by taking the second derivative with respect to time. Based on the three-dimensional velocity vector, it constructs a radial velocity change rate model that includes the distance to be measured. The Doppler frequency modulation features are used to perform nonlinear least squares fitting on the model to calculate the radial distance of the target countermeasure signal relative to the UAV. The radial distance is then combined with the spatial incident angle to generate the three-dimensional geographic coordinates of the target countermeasure signal transmitter. The system also includes a compliance determination module, which extracts the center frequency, bandwidth, and modulation type parameters of the target countermeasure signal. The calculated spatial incident angle is compared with a pre-set database of legitimate transmitter geographic coordinates. When the signal parameters match but the spatial incident angle does not coincide with the coordinates of any legitimate transmitter, an illegal countermeasure signal alarm is generated.

[0034] Example 1: In a scenario where a small rotary-wing UAV performs countermeasure monitoring in a low-altitude urban area, the area is characterized by numerous high-rise buildings and an exceptionally complex electromagnetic environment, filled with high-power broadcast signals and multipath reflections. Limited by its size and payload, the UAV platform carries only a centimeter-scale aperture antenna array and continuously endures high-frequency mechanical vibrations during flight. When the system starts under these conditions, the first receiving channel of the RF acquisition subsystem connects to the monitoring antenna to capture weak countermeasure signals from unknown sources, while the second receiving channel connects to the reference antenna to lock onto the local broadcast signal as a phase anchor point. The state synchronization bus outputs the UAV's three-dimensional spatial coordinates, three-dimensional velocity vector, and high-frequency acceleration components in real time at a frequency higher than the RF sampling rate, including timestamps. To address the phase ambiguity caused by the platform's high-frequency vibrations, the system utilizes the high-frequency acceleration components output by the inertial measurement unit, and performs time-domain quadratic integration via a virtual aperture synthesis processor to reconstruct the instantaneous mechanical vibration displacement vector of the antenna phase center relative to the smooth motion trajectory. The processor is based on the formula Calculate the corresponding phase error value The processor performs reverse compensation on the acquired target phase sequence. This step converts the vibration noise in the mechanical domain into a definite radio frequency phase error and eliminates it, ensuring the consistency of phase data in subsequent synthetic aperture processing. To address the coherence degradation caused by the frequency drift of the inexpensive receiver crystal oscillator, the processor uses the instantaneous phase difference between the first and second receiving channels to perform differential operations and uses the characteristics of a shared clock source to cancel the common-mode frequency drift, thereby establishing a robust coherent integration time of several seconds without the need for an expensive, highly stable crystal oscillator.

[0035] In urban canyon environments with severe multipath effects, reflected signals are often stronger than direct signals, easily leading to direction-finding errors. The virtual aperture synthesis processor performs a multipath consistency filtering operation, dividing continuous received data into multiple short-time sub-slices and calculating the instantaneous Doppler frequency of each sub-slice. The processor compares these measured values ​​with a theoretical Doppler frequency response set constructed based on the UAV's three-dimensional velocity vector, calculating the Pearson correlation coefficient. When the correlation coefficient is lower than a preset coherence threshold, the sub-slice data is determined to be a non-line-of-sight multipath signal and discarded, retaining only direct path signals that conform to kinematic geometric constraints for constructing the virtual antenna array. Addressing the problem of strong interference masking weak targets at the same frequency, the processor identifies the azimuth of the strong interference source, constructs an orthogonal projection matrix, and projects the received data... The orthogonal complement space projected onto the interference steering vector mathematically forms a null trap pointing towards the strong interference source, thus extracting a weak countermeasure signal against a strong interference background. Using the clean phase sequence processed above, the processor combines the UAV's motion trajectory to construct a virtual antenna array model and calculates the spatial covariance matrix. The high-precision spatial incident angle of the target is calculated using the MUSIC algorithm. To achieve single-station positioning, the processor further calculates the second derivative of the corrected target phase sequence with respect to time, extracts the Doppler frequency modulation feature, and fits it with the radial velocity change rate model constructed based on the three-dimensional velocity vector. The radial distance of the target relative to the UAV is directly calculated. Finally, the system combines the radial distance with the spatial incident angle to generate the three-dimensional geographic coordinates of the target.

[0036] Example 2: To verify the direction-finding performance, multipath resistance, and single-station ranging accuracy of the UAV RF countermeasure monitoring system based on multi-source signal state perception of the present invention in a real complex environment, a comprehensive test field was constructed, including a controlled signal source, a multipath simulation environment, and an airborne test platform. The test signal source is a vector signal generator capable of transmitting broadband RF signals covering a frequency range of 30MHz to 6GHz, and supporting simulation of AM, FM, and digital modulation signals. The multipath environment is constructed by setting multiple metal reflectors and absorbing materials along the signal propagation path to simulate a typical non-line-of-sight propagation scenario in an urban canyon. The power ratio (Rice factor K) of the direct path signal and the reflected path signal is adjustable. The airborne test platform is a customized hexacopter UAV equipped with the RF acquisition subsystem, state synchronization bus, and inertial measurement unit of the present invention. The RF acquisition subsystem is configured with two receiving channels sharing the same local clock source, which are connected to the monitoring antenna and the reference antenna, respectively. The state synchronization bus records the three-dimensional position, velocity, and acceleration data of the UAV at a frequency of 100Hz.

[0037] After the experiment was started, the UAV flew along a predetermined trajectory, maintaining an altitude between 50 and 100 meters and a speed between 5 and 15 meters per second. The vector signal generator emitted a simulated countermeasure signal with a center frequency of 1.2 GHz and a bandwidth of 5 MHz. During flight, the radio frequency acquisition subsystem synchronously acquired complex baseband data from the monitoring and reference channels. To verify the system's compensation effect on phase errors caused by mechanical vibration, an additional mechanical vibration with a frequency of 50 Hz and an amplitude of 2 mm was artificially applied to the UAV's arm. The virtual aperture synthesis processor performed a phase cleaning operation based on microelectromechanical vibration decoupling, utilizing inertia... The high-frequency acceleration data collected by the measurement unit is used to reconstruct the instantaneous vibration displacement of the antenna phase center and calculate the corresponding phase error value to compensate for the radio frequency data. The phase denoising differential logic is executed, and the local oscillator frequency drift is offset by the reference channel signal. To intuitively demonstrate the effectiveness of the multipath consistency screening operation, a segment of flight data containing strong multipath interference is selected for analysis and divided into multiple 20-millisecond time sub-slices. The measured instantaneous Doppler frequency of each sub-slice is calculated and correlated with the theoretical Doppler frequency calculated based on the real motion trajectory. Table 1 shows the comparison data of direction finding error and signal quality under different processing stages.

[0038] Table 1: Comparison of Direction Finding Error and Signal Quality at Different Processing Stages

[0039]

[0040] Referring to Table 1, the raw data without any processing has a phase consistency index of only 0.42 due to mechanical vibration and frequency drift, resulting in a direction finding error as high as 12.5°. This indicates that virtual aperture synthesis technology is difficult to apply on small UAV platforms without an effective compensation mechanism. When a vibration compensation mechanism is introduced, the phase consistency is improved to 0.85 and the direction finding error is reduced to 4.8°, verifying the effectiveness of using inertial data to clean the RF phase. After further introducing differential correction of the reference channel, the direction finding error continues to decrease to 3.2°, proving the key role of this mechanism in suppressing local oscillator frequency drift. Most importantly, after performing multipath consistency screening, the direction finding error is significantly reduced to 1.1°, confirming that the strategy of eliminating non-line-of-sight signals based on kinematic constraints can effectively cope with multipath interference in complex urban environments. Finally, after superimposing geometric autofocus operation to correct the trajectory error, the system achieves a direction finding accuracy of 0.6° and a single-station ranging relative error of 1.1%.

[0041] Example 3: This example combines Figs. 1 to 3 The description of the UAV radio frequency countermeasure monitoring system based on multi-source signal state perception is as follows: Fig. 1 As shown, the system receives target countermeasure signals via a monitoring antenna and external reference signals via a reference antenna. Both are input to the radio frequency acquisition subsystem and synchronously acquired using a dual-channel common-source differential mechanism. Simultaneously, a state synchronization bus connects to an inertial measurement unit (IMU) to acquire high-frequency acceleration components and connects to the flight control system to acquire three-dimensional position / velocity, thus completing the motion state vector acquisition. All data is fed into a virtual aperture synthesis processor, where the radio frequency data stream undergoes phase denoising differential logic to cancel local clock frequency drift and multipath consistency filtering to eliminate non-line-of-sight signals. At the same time, the system performs geometric autofocus to optimize trajectory errors. The processor uses the motion state data to perform microelectromechanical vibration decoupling for phase inversion compensation, and then performs strong interference orthogonal suppression to project the signal onto the orthogonal complement space. Based on this, a spatial covariance matrix is ​​constructed using a virtual antenna array model. The data stream then branches to perform single-station passive ranging to complete radial distance calculation and MUSIC algorithm calculation to obtain the spatial incident angle. Finally, the two are combined to generate three-dimensional geographic coordinates as the transmitter source positioning result, and the compliance judgment module outputs an illegal countermeasure signal alarm.

[0042] like Fig. 2 As shown, a comparative coordinate system is established in the figure, with time in seconds as the horizontal axis and phase error in radians as the vertical axis. The horizontal axis scale covers the range from 0 to 10 seconds, and the vertical axis scale covers the range from 0 to 1.2 radians. The figure contains two broken line data. The dashed line represents the phase error before compensation, with the value fluctuating wildly between 0.3 radians and 1.2 radians, showing extremely unstable phase characteristics. The solid line represents the phase error after compensation, with the value consistently remaining below 0.2 radians. Fig. 3As shown, the local clock source simultaneously provides synchronous clock signals to the first and second receiving channels to ensure the introduction of common-mode phase noise. The monitoring antenna transmits the input target countermeasure signal to the first receiving channel, and the reference antenna transmits the input external reference signal, such as a broadcast or cellular pilot, to the second receiving channel. The two channels perform direct frequency conversion and ADC sampling operations respectively. The first receiving channel transmits the target signal complex baseband data, and the second receiving channel transmits the reference signal complex baseband data to the virtual aperture synthesis processor. The processor calculates the instantaneous phase difference between the two channels to cancel the common-mode frequency drift noise, and further compares the theoretical and measured Doppler frequency shift to separate the local oscillator frequency error, finally generating a time-aligned corrected target phase sequence.

[0043] Example 4: This example aims to provide a systematic engineering detail of the core data processing flow of the virtual aperture synthesis processor, particularly addressing the algorithmic black box and parameter calibration issues in multipath consistency screening and geometric autofocus operations. It provides a transparent, verifiable, and logically closed-loop implementation plan. In the multipath consistency screening stage, the physical properties of the input data are defined; the input is a complex baseband signal sequence after vibration compensation and differential correction. and synchronized UAV motion state vectors To quantitatively assess multipath effects, the system does not rely on subjective experience or judgment, but instead executes a standardized correlation analysis procedure. The processor will... A series of sub-signal segments are generated by sliding slices within a 20ms time window. For each sub-signal segment, its instantaneous frequency characteristics are calculated using Short-Time Fourier Transform (STFT). Construct a theoretical Doppler frequency response model, based on The three-dimensional velocity vector in the image is compared with a preset full-space scanning angle grid to calculate the theoretical Doppler frequency shift corresponding to each grid point. The core screening logic is based on the Pearson correlation coefficient. The calculation formula is as follows: ,in, The Pearson correlation coefficient is used. The measured instantaneous Doppler frequency at time t within the time sub-slice; This is the theoretical Doppler frequency corresponding to time t. and These are the arithmetic means of the measured instantaneous Doppler frequency sequence and the theoretical Doppler frequency sequence within the time slice, respectively, with the sign... This indicates that the summation operation is performed on all sampling times t within this time sub-slice, and the system sets a coherence threshold. The determination of this threshold is based on statistical analysis of a large amount of measured data, that is, when The signal is often severely contaminated by non-line-of-sight reflections, causing the direction-finding error to increase exponentially. The time signal is mainly dominated by the direct aperture and satisfies the phase linearity constraint of virtual aperture synthesis. Therefore, only when it satisfies... Only the sub-slices will be retained for subsequent processing.

[0044] In the geometric autofocus stage, to address the inherent position drift error of the inertial navigation system, an optimization strategy based on minimizing image entropy is adopted, defining the objective function as the spectral entropy of the virtual spatial spectrum. The calculation formula is: ,in The objective function represents the normalized spatial spectral intensity distribution. Its physical meaning lies in the fact that when the trajectory error is zero, the energy of the point source target is highly concentrated and the spectral entropy is minimized. As the error increases, the energy diverges, leading to an increase in spectral entropy. The optimization process employs a gradient descent algorithm, using the position offset... To optimize variables, the trajectory data is iteratively updated until... Converging to a local minimum, this process uses an external reference signal as a spatial benchmark to reverse-calibrate the UAV's trajectory, thereby eliminating the impact of sensor errors on direction-finding accuracy. Through the standardized algorithm flow and parameter calibration procedures described above, this embodiment transforms the originally abstract signal processing concept into a set of executable engineering operation instructions. To ensure the engineering reproducibility of the virtual aperture synthesis processor's multipath consistency screening and geometric autofocus operations, the system parameters are determined using standardized calibration procedures. In the multipath consistency screening operation, the Pearson correlation coefficient... coherence threshold Value This value is obtained by offline statistical calibration of urban environmental data. The calibration procedure includes a multipath Rice factor. Actual measurement Root mean square error of direction finding A three-dimensional mapping lookup table, when When the direction finding error begins to deteriorate exponentially, the corresponding Established as This value is in This point represents the critical separation point between the direct trajectory signal and the non-line-of-sight reflected signal in terms of kinematic characteristics. A gradient descent optimization algorithm is used for geometric autofocus operation, with the optimization variable being the UAV's three-dimensional position offset. The initial search boundary is limited to meters, gradient descent learning rate Set as The spectral entropy of the objective function in two consecutive iterations is used. rate of change less than or Euclidean norm less than Metrics are used as a convergence criterion.

[0045] Example 5: To ensure the stable operation of the system of the present invention in complex and ever-changing real-world deployment environments, this example constructs a standardized engineering procedure covering two key stages: offline calibration and pre-deployment calibration. In the offline calibration and data filling stage, the coherence threshold, a core component of the multipath consistency screening operation, is addressed. Instead of using a single empirical value for this crucial parameter, the system employs a controlled hardware-in-the-loop simulation platform for offline data acquisition. This platform comprises an anechoic chamber environment, a programmable multipath channel simulator, and the UAV system under test. It sets up a series of gradient channel scenarios, ranging from pure direct path to extreme non-line-of-sight (corresponding to Rice factor K from...). (Continuous variation up to 0), the system collects and records the correlation coefficient between the measured Doppler frequency and the theoretical value under different multipath intensities. The system collects the corresponding direction-finding error distribution and uses this data to construct a lookup table containing a three-dimensional mapping relationship between Rice factor, correlation coefficient, and direction-finding error. This core prior knowledge is stored in the non-volatile memory of the virtual aperture synthesis processor. In actual operation, the system can dynamically query and call the optimal [data / method] based on the real-time monitored environmental multipath intensity indicators. value.

[0046] During the pre-deployment calibration phase, a power-on self-test and calibration process was defined to address potential installation error angles between the inertial navigation system (IMU) and the RF receiver, as well as initial deviations in the receiver's local oscillator frequency. The RF acquisition subsystem adopts a software-defined radio (SDR) architecture, including a direct-conversion receiver front-end and an analog-to-digital converter (ADC). Common-mode frequency drift between the two channels is effectively canceled. The first and second receiving channels share the same local clock source. A synchronous sampling clock is used to drive the ADC to ensure sampling errors are within nanoseconds. The status synchronization bus operates at a frequency not lower than [missing information]. The Hertz update rate provides the virtual aperture synthesis processor with a motion state vector containing three-dimensional spatial coordinates and three-dimensional velocity vectors. The processor receives the motion state vector and times-aligns it with the complex baseband data provided by the RF acquisition subsystem. By comparing timestamp tags, a linear interpolation algorithm is executed to ensure that the geometric relationship error between the RF phase data and the platform's motion trajectory is less than one-thousandth of the carrier wavelength, meeting the stringent requirements of virtual aperture synthesis for phase coherence. Before the UAV takes off or in the initial stage of entering the mission area, the system automatically locks onto a strong cooperative signal source with known azimuth, such as a ground calibration beacon or a legitimate broadcast tower. The virtual aperture synthesis processor uses this cooperative signal source to perform short-term static direction finding, calculates the deviation between the measured angle and the true angle, and solves for the rotation matrix correction between the IMU coordinate system and the antenna array coordinate system. Simultaneously, using the carrier frequency of the signal source as a reference, the receiver's local oscillator is calibrated to obtain the initial frequency offset value. These two calibration parameters and It is immediately injected into the system's real-time processing flow, serving as the basis for all subsequent dynamic calculations.

[0047] Example 6: To address performance fluctuations caused by hardware batch differences, aging drift, and uncertainties in the electromagnetic background of the deployment environment during actual system deployment, this example provides a standardized pre-deployment calibration procedure. Through a series of controlled self-testing and calibration steps, it ensures that the system is in optimal working condition before each mission. When the system completes physical deployment in an unfamiliar environment and powers on for the first time, the state synchronization bus automatically triggers the zero-bias calibration process of the inertial measurement unit (IMU). The UAV remains stationary, and the virtual aperture synthesis processor continuously acquires 60 seconds of raw IMU acceleration and angular velocity data. The processor uses this 60-second data to calculate the zero-bias deviation of the triaxial accelerometer. Drift of the three-axis gyroscope These two parameters are then written into the IMU's non-volatile memory as real-time compensation terms.

[0048] For the calibration of receiver local oscillator frequency and antenna array installation error, the system uses a known strong signal source in the environment as a spatial reference. The operator specifies a legal broadcast tower with a known azimuth angle within the line of sight as the calibration source via the host computer. The RF acquisition subsystem locks the carrier frequency of the calibration source and continuously collects complex baseband data for 10 seconds. The virtual aperture synthesis processor calculates the average value of the measured Doppler frequency shift during this time period and compares it with the theoretical Doppler frequency shift (i.e., 0Hz) determined based on the relative stationary state of the UAV and the calibration source, thus calculating the absolute frequency deviation of the local clock source. The processor uses the MUSIC algorithm to determine the direction of the calibration source and calculates the deviation angle between the calculated spatial angle and the true geographical location of the calibration source. This deviation angle is considered as a fixed installation error of the antenna array relative to the aircraft coordinate system. After calibration, and The preprocessing module, automatically loaded into the signal processing link, serves as a system-level constant in correcting all subsequent real-time data, thereby eliminating the impact of individual hardware differences on direction-finding accuracy. Addressing the phase error introduced by high-frequency mechanical vibration of the UAV platform, the state synchronization bus acquires raw acceleration components from the inertial measurement unit. In this embodiment, a second-order Butterworth high-pass filter with a cutoff frequency set to 10 Hz is used. This filter rigorously executes the aforementioned frequency domain separation logic, identifying signal components above 10 Hz as mechanical vibration noise and performing phase compensation, while retaining signal components below 10 Hz as effective platform maneuvering acceleration. The Hertz cutoff frequency is set based on the inherent vibration spectrum characteristics of the rotary-wing UAV to ensure it is below [a certain value]. Hertzian motion components are considered as the effective motion required for virtual aperture synthesis, and are higher than... The Hertzian component is used to compensate for mechanical vibration noise and eliminate installation error angles between the inertial navigation system and the antenna array. The system performs a pre-deployment calibration procedure, which requires locking onto at least two cooperative signal sources with known three-dimensional geographic coordinates in the environment before takeoff. The processor uses the true azimuth angles of the signal sources to calculate the complete three-dimensional rotation matrix. , The matrix is ​​used to accurately transform the motion state vector of the IMU coordinate system to the antenna array coordinate system.

[0049] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0050] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A UAV radio frequency countermeasure monitoring system based on multi-source signal state perception, characterized in that, include: The radio frequency acquisition subsystem includes a first receiving channel and a second receiving channel that share the same local clock source. The first receiving channel is connected to a monitoring antenna to acquire complex baseband data of the target countermeasure signal, and the second receiving channel is connected to a reference antenna to synchronously acquire complex baseband data of an external reference signal. A state synchronization bus is connected to the UAV flight control system and inertial measurement unit to obtain the UAV motion state vector containing timestamps. The motion state vector includes at least three-dimensional spatial coordinates and three-dimensional velocity vectors. The virtual aperture synthesis processor, connected to the RF acquisition subsystem and the status synchronization bus, executes the following data processing logic: The instantaneous phase difference between the first receiving channel and the second receiving channel is calculated. The instantaneous phase difference is used to cancel the common-mode frequency drift noise introduced by the local clock source, and a corrected target phase sequence that is time-aligned with the UAV motion state vector is generated. Based on the three-dimensional velocity vector, sampling segments with Doppler frequency shift change rate exceeding a preset threshold are selected as effective synthetic aperture intervals, and invalid sampling data that are in uniform linear motion and have constant radial velocity are removed. Based on the phase sequence of the corrected target within the effective synthetic aperture range, and according to the spatial trajectory geometric distribution defined by the motion state vector, a virtual antenna array model is constructed. Calculate the spatial covariance matrix of the virtual antenna array model, and solve the spatial incident angle of the target countermeasure signal relative to the UAV based on the spatial covariance matrix.

2. The UAV radio frequency countermeasure monitoring system based on multi-source signal state perception according to claim 1, characterized in that, The virtual aperture synthesis processor also performs geometric autofocus operations, which include: calculating the virtual spatial spectrum of the external reference signal based on the complex baseband data of the external reference signal and the current UAV motion state vector, and extracting the focus feature index of the virtual spatial spectrum; constructing a trajectory error optimization model with the focus feature index as the objective function and iteratively searching for the trajectory correction amount that makes the focus feature index reach its extreme value; updating the UAV motion state vector using the trajectory correction amount, and constructing a virtual antenna array model for the target countermeasure signal based on the updated UAV motion state vector.

3. The UAV radio frequency countermeasure monitoring system based on multi-source signal state perception according to claim 1, characterized in that, The virtual aperture synthesis processor also performs a multipath consistency screening operation, which includes: dividing the complex baseband data of the first receiving channel into multiple time sub-slices and estimating the instantaneous Doppler frequency of each time sub-slice; constructing a theoretical Doppler frequency response set corresponding to the full-space scanning angle based on the three-dimensional velocity vector; calculating the matching degree between the instantaneous Doppler frequency and the theoretical Doppler frequency response set and removing sub-slice data with a matching degree lower than a preset coherence threshold; and constructing a virtual antenna array model using only the retained sub-slice data.

4. The UAV radio frequency countermeasure monitoring system based on multi-source signal state perception according to claim 1, characterized in that, The virtual aperture synthesis processor also performs strong interference orthogonal suppression operations, which include: identifying strong interference signals present in the radio frequency environment and obtaining the interference source azimuth vector of the strong interference signal relative to the UAV trajectory; calculating the theoretical phase response of the strong interference signal at each virtual element of the virtual antenna array model based on the UAV motion state vector to generate an interference steering vector; constructing a projection matrix orthogonal to the interference steering vector and mapping the complex baseband data to the orthogonal complement space of the interference steering vector; and reconstructing the virtual antenna array model using the generated residual data stream.

5. The UAV radio frequency countermeasure monitoring system based on multi-source signal state perception according to claim 1, characterized in that, The state synchronization bus also acquires the high-frequency acceleration components of the UAV body; the virtual aperture synthesis processor calculates the phase error value caused by mechanical vibration using the following phase compensation formula, and uses the phase error value to perform reverse compensation on the correction target phase sequence: ,in, For a moment The phase error value, The carrier wavelength of the target countermeasure signal The line-of-sight vector representing the countermeasure signal from the target. The instantaneous mechanical vibration displacement vector of the antenna phase center relative to the smooth motion trajectory, which is reconstructed by quadratic integration of the high-frequency acceleration components.

6. The UAV radio frequency countermeasure monitoring system based on multi-source signal state perception according to claim 1, characterized in that, The virtual aperture synthesis processor also performs single-station passive ranging operations, which include: performing time unwrapping operations on the phase sequence of the calibration target and calculating the second derivative of the phase sequence with respect to time to extract Doppler frequency modulation features; constructing a radial velocity change rate model containing the distance to be measured parameters based on the three-dimensional velocity vector; and using the Doppler frequency modulation features to perform parameter fitting on the radial velocity change rate model to calculate the radial distance of the target countermeasure signal relative to the UAV.

7. The UAV radio frequency countermeasure monitoring system based on multi-source signal state perception according to claim 6, characterized in that, The virtual aperture synthesis processor combines the calculated radial distance with the spatial incident angle to generate the three-dimensional geographic coordinates of the target countermeasure signal source.

8. The UAV radio frequency countermeasure monitoring system based on multi-source signal state perception according to claim 1, characterized in that, The system also includes a compliance determination module, which extracts the center frequency, bandwidth and modulation type parameters of the target countermeasure signal, and compares the calculated spatial incident angle with a preset database of legal transmitter geographical coordinates; when the signal parameters match but the spatial incident angle does not coincide with the coordinates of any legal transmitter, an illegal countermeasure signal alarm data is generated.

9. The UAV radio frequency countermeasure monitoring system based on multi-source signal state perception according to claim 1, characterized in that, The external reference signal is one of the following: a broadcast frequency signal, a cellular network pilot signal, or a downlink signal transmitted by the UAV's own data link; the virtual aperture synthesis processor separates the common-mode frequency error caused by the local clock source by comparing the theoretical Doppler frequency shift and the measured Doppler frequency shift of the external reference signal.

10. The UAV radio frequency countermeasure monitoring system based on multi-source signal state perception according to claim 1, characterized in that, The radio frequency acquisition subsystem adopts a software-defined radio architecture, which includes a direct conversion receiver front-end and an analog-to-digital converter. The analog-to-digital converter directly samples the analog baseband signals of the first and second receiving channels synchronously.

Images