Encrypted signal receiving and guiding system for unmanned aerial vehicle

Abstract

The invention discloses an encrypted signal receiving guide system for an unmanned aerial vehicle. According to the invention, the self-adaptive anti-interference module realizes real-time identification, feature analysis and dynamic suppression of interference signals through a multi-dimensional joint sensing and closed-loop adjustment system. The interference detector is fused with a time-frequency domain depth feature extraction technology, and existence features of complex interference such as frequency hopping and frequency sweeping can be captured in a millisecond-level time window; the interference characteristic analyzer accurately deconstructs a modulation rule and a propagation characteristic of an interference signal based on a nonlinear dynamic modeling and pattern learning algorithm; the self-adaptive filter adopts a multi-modal optimization strategy and combines space-time correlation characteristics of navigation signals, so that the integrity of original signals is reserved to the greatest extent while interference is eliminated. The whole set of mechanism forms a complete anti-interference closed loop from environmental perception to dynamic response, and ensures that the unmanned aerial vehicle can still maintain a stable encrypted communication link and centimeter-level navigation precision in extreme scenes such as strong electromagnetic suppression and dense signal conflicts.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicles, and specifically relates to an encrypted signal receiving and guiding system for unmanned aerial vehicles. Background Art

[0002] The unmanned aerial vehicle signal receiving and guiding system is a high-tech navigation and control device. It realizes precise guidance and control of the unmanned aerial vehicle by receiving signals from the Global Positioning System (GPS), ground control station instructions, and sensors carried by the unmanned aerial vehicle itself. The system mainly consists of an antenna, a signal receiver, a data processing unit, and an instruction transmitter, and can process data such as the position, speed, and attitude of the unmanned aerial vehicle during flight in real time, and optimize and adjust the flight path. In a complex environment, the unmanned aerial vehicle signal receiving and guiding system can ensure the stable flight of the unmanned aerial vehicle, complete tasks such as aerial photography, reconnaissance, and transportation, and at the same time has the characteristics of strong anti-interference ability, fast response speed, and simple operation, providing reliable technical support for the wide application of unmanned aerial vehicles in the military and civilian fields.

[0003] However, the existing technologies mainly have defects such as the static protection system being difficult to cope with dynamic electromagnetic threats, the single-dimensional anti-interference strategy being vulnerable to new types of composite interference suppression, and the encryption and anti-interference modules operating independently, resulting in security protection breaks. Traditional systems face systematic risks such as increased signal distortion, lag in key synchronization, and inaccurate navigation control in a rapidly time-varying interference environment, severely restricting the reliable operation of unmanned aerial vehicles in high-confrontation scenarios. Summary of the Invention

[0004] The purpose of the present invention is to provide an encrypted signal receiving and guiding system for unmanned aerial vehicles to solve the above-mentioned problems.

[0005] The technical solution adopted by the present invention is as follows: An encrypted signal receiving and guiding system for unmanned aerial vehicles, the system includes: a signal encryption module, a signal receiving module, a signal decryption module, a data processing module, a navigation control module, and an adaptive anti-interference module;

[0006] Inside the navigation control module, there are: a flight controller, an attitude resolver, and an actuator module;

[0007] Inside the adaptive anti-interference module, there are: an interference detector module, an interference feature analyzer module, and an adaptive filter module;

[0008] The output port of the signal encryption module and the RF input port of the signal receiving module are physically connected through a wireless channel, and the encrypted data stream enters the RF front end through the antenna.

[0009] The baseband output port of the signal receiving module is directly connected to the input buffer of the signal decryption module through a digital interface to transmit the down-converted encrypted baseband signal;

[0010] The output verification port of the signal decryption module is connected to the input queue of the data processing module through a secure bus to transfer the decrypted original navigation message;

[0011] The attitude data output port of the data processing module is docked with the instruction input port of the navigation control module through a high-speed serial interface to convey the calculated three-dimensional coordinates and velocity vector;

[0012] The detection probe of the adaptive anti-jamming module is embedded in the intermediate frequency processing link of the signal receiving module, and its parameter control port is cross-connected to the gain adjustment node of the RF front end and the filter coefficient loading interface of the baseband processing unit through a parallel bus to form a feedforward-feedback composite control loop.

[0013] In a preferred embodiment, the signal encryption module realizes the dynamic generation and secure distribution of multi-level keys through a key manager. The key manager adopts a key exchange protocol based on elliptic curve cryptography, automatically updates the session key every 30 seconds, and cooperates with the encryption processor to complete the real-time data stream encryption.

[0014] In a preferred embodiment, the signal receiving module adopts a software-defined radio architecture, realizes full-coverage reception in the 1.2 - 1.6 GHz frequency band through a tunable bandpass filter bank. The low-noise amplifier uses temperature compensation technology, and cooperates with a digital control attenuator to construct a dynamic gain control system to ensure that the signal amplitude is stable in the optimal ADC quantization range; the baseband processing unit uses a polyphase filter bank for channelization processing, decomposes the broadband signal into 8 parallel 200 MHz sub-channels, and uses the parallel computing ability of the FPGA to complete 64-order FIR filtering and digital down-conversion in real time.

[0015] In a preferred embodiment, the signal decryption module adopts a heterogeneous computing architecture, where a dedicated cryptographic chip processes the decryption of the underlying data stream, and a general-purpose processor executes the decryption of the protocol layer; for the AES-GCM encrypted data stream, while the hardware decryption engine processes 128-bit ciphertext blocks at line speed, it simultaneously generates an authentication tag for data integrity verification.

[0016] In a preferred embodiment, the data processing module adopts an improved PBCH de-encapsulation technology to separate three types of data streams of positioning, timing, and integrity from the decrypted navigation message; the coordinate converter builds a quaternion Kalman filter, fuses the conversion parameters of the Beidou coordinate system, the WGS-84 coordinate system, and the local grid coordinate system, and compensates the coordinate offset caused by crustal plate movement in real time.

[0017] In a preferred embodiment, the flight controller constructs a three-dimensional trajectory tracking system using a model predictive control framework. A rolling optimization strategy based on Lyapunov stability is designed for the non-linear dynamic characteristics of the UAV in a complex airflow environment. During the path correction process, the controller calculates the lateral and longitudinal deviations between the actual trajectory and the planned trajectory in real time, linearizes the attitude dynamics equation through the Jacobian matrix, and generates a composite control command including thrust allocation and rudder deflection.

[0018] The attitude resolver constructs a deep coupling resolution architecture for the nine-axis MEMS sensor based on an improved extended Kalman filter. A gradient descent method with motion acceleration compensation is designed for the gyroscope integration drift problem caused by high-speed maneuvers, and the accelerometer data is converted into a quaternion correction quantity.

[0019] In a preferred embodiment, the brushless motor drive unit of the actuator module uses space vector pulse width modulation technology to achieve millisecond-level torque response in cooperation with a full carbon fiber reduction gearbox. A three-closed-loop control structure with feedforward compensation is developed for the sudden change characteristics of the aerodynamic load of the rudder surface, and a non-linear friction identification algorithm based on Lissajous figures is embedded in the position loop. The electronic speed control module integrates a temperature-current coupling model, dynamically adjusts the switching frequency by real-time monitoring of the MOSFET junction temperature, and maintains the linearity of the thrust output in the environment range of -40°C to 85°C.

[0020] In a preferred embodiment, the interference detector module fuses cyclic spectrum analysis and wavelet transform to construct a time-frequency joint detection model. By calculating the spectral correlation density function and the multi-scale wavelet coefficient energy distribution of the received signal in real time, an interference discrimination mechanism based on multiple hypothesis testing is established. In the time domain dimension, an energy detector with a dynamic threshold is designed, and the sliding window variance analysis technology is used to identify abnormal power fluctuations. In the frequency domain dimension, an improved cyclic stationary feature extraction algorithm is adopted, and the typical interference signal characteristics are matched by constructing a cyclic frequency spectrum line cluster. For time-varying threats such as frequency hopping interference, an online learning unit is embedded in the module, the interference signature database is dynamically updated based on kernel density estimation, and the fractal dimension calculation is introduced to perform non-linear discrimination on transient interference.

[0021] Where: The calculation formula of the joint detection statistic is:

[0022]

[0023] Where:

[0024] represents the spectral correlation density function of the received signal at the cyclic frequency \(\alpha\) and the spectral frequency \(f\), reflecting the cyclic stationary characteristics of the signal;

[0025] α0 to α1 represent the cyclic frequency interval for key monitoring, which is set to [100 Hz, 10 kHz] based on the typical characteristics of UAV navigation signals;

[0026] W is the detail coefficient matrix obtained by lifting wavelet transform, which represents the time-frequency localization characteristics of the signal;

[0027] λ is an adaptive weighting factor, which is dynamically adjusted according to the ambient noise floor and has a value range of [0.5, 1.2];

[0028] trace(W·W H ) is used to calculate the cumulative energy of wavelet coefficients, which is used to capture the transient characteristics of sudden interference;

[0029] The formula extracts the continuous distribution information of cyclostationary features through integral operation, and uses matrix trace operation to capture energy mutations in the wavelet domain. The two-dimensional features are fused through nonlinear weighted fusion to form a sensitive and robust detection statistic, which will improve the interference recognition rate compared with the traditional single-dimensional detection method.

[0030] In a preferred embodiment, the interference feature analyzer module adopts a hybrid analysis method combining time-frequency ridge tracking and high-order tensor decomposition to perform refined modeling of the interference signal by constructing a multidimensional feature space; at the time-frequency analysis layer, a ridge extraction algorithm based on an adaptive Gaussian chirped wavelet is designed to dynamically match the instantaneous frequency change rate of the interference signal by optimizing the window function bandwidth, accurately capturing the dwell time and frequency switching pattern of the frequency hopping interference; wherein:

[0031] The calculation formula of the adaptive ridge tracking cost function is:

[0032]

[0033] Where: TF(t,f) represents the time-frequency distribution matrix, which is generated by the improved synchro-compressed wavelet transform and has the energy aggregation characteristic;

[0034] Indicates the gradient change of the calculated time-frequency energy along the frequency axis, strengthening the continuity constraint of the instantaneous frequency trajectory;

[0035] TV(f(t)) is the total variation regularization term of the frequency trajectory, which suppresses the generation of false ridge branches;

[0036] β is the adaptive balancing factor, which is dynamically adjusted according to the interference signal bandwidth to meet the intelligent adaptation of narrowband interference to 0.3 and broadband interference to 1.5;

[0037] By fusing the gradient constraint and the variational regularization mechanism, the formula can accurately reconstruct the fast frequency hopping pattern with the interval of frequency mutation points less than 5 μs under the strong noise background, reducing the ridge line positioning error by 62% compared with the traditional dynamic programming method, and providing high-precision feature input for subsequent interference type identification.

[0038] In a preferred embodiment, the adaptive filter module constructs a hybrid processing mechanism of frequency domain subspace projection and time domain recursive filtering, and realizes precise spectrum shaping by establishing a dynamic mapping relationship between interference characteristics and filter parameters; for comb-shaped spectrum interference, a frequency domain stopband generation algorithm based on constrained convex optimization is designed, and a notch response with a depth of up to 60 dB is formed at the specified frequency point by using an improved lattice filter structure; for broadband noise interference, a normalized least mean square algorithm with a time-varying step size is developed, and the spectral feature prior information provided by the interference feature analyzer is combined to dynamically adjust the convergence factor and the filter order; a space-time joint processing scheme is particularly proposed, and the spatial steering vector of the array signal and the time domain weight matrix of adaptive filtering are fused by Kronecker product to form a multi-dimensional filtering space.

[0039] The calculation formula of the hybrid optimization objective function is as follows:

[0040]

[0041] Where:

[0042] H(ω) represents the filter frequency domain response function, and the stopband characteristics are described by a parametric B-spline curve;

[0043] Y(ω) and $X(\omega) represent the spectral components of the filter output and the desired signal respectively

[0044] λ is the time domain correlation weight coefficient, which is dynamically adjusted according to the interference pulse repetition period, and the value range is [0.1, 5]; Φ(τ) is designed as an exponentially decaying window function, which is used to strengthen the extraction of the correlation peak at the key time delay point;

[0045] η controls the sparse constraint intensity, and is evaluated by KL divergence and adjusted in real time to balance the convergence speed and the steady-state error;

[0046] is a time domain gradient morphological operator, which is used to suppress the waveform distortion of the filtered signal; the formula reduces the signal distortion entropy by 45% under the strong interference background through the triple constraints of spectral distortion, time domain correlation and structural sparsity, and especially expands the suppression bandwidth of LFM swept frequency interference to 2.3 times that of the conventional algorithm, significantly improving the fidelity of the navigation signal in the complex electromagnetic environment.

[0047] In summary, due to the adoption of the above technical solutions, the beneficial effects of the present invention are as follows:

[0048] 1. In the present invention, the adaptive anti-interference module realizes real-time identification, feature analysis, and dynamic suppression of interference signals through a multi-dimensional joint perception and closed-loop regulation system. The interference detector integrates time-frequency domain deep feature extraction technology and can capture the existence features of complex interferences such as frequency hopping and frequency scanning within a millisecond-level time window; the interference feature analyzer is based on non-linear dynamics modeling and pattern learning algorithms to accurately deconstruct the modulation rules and propagation characteristics of interference signals; the adaptive filter adopts a multi-modal optimization strategy, combines the spatio-temporal correlation features of navigation signals, and maximally retains the integrity of the original signal while eliminating interference. The entire mechanism forms a complete anti-interference closed-loop from environmental perception to dynamic response, ensuring that the UAV can still maintain a stable encrypted communication link and centimeter-level navigation accuracy in extreme scenarios such as strong electromagnetic suppression and dense signal conflicts.

[0049] 2. In the present invention, the dynamic key management and anti-quantum algorithm design of the signal encryption module block the risks of signal interception and malicious cracking from the source, while the adaptive anti-interference module can actively predict and avoid potential threat areas by intelligently learning the evolution law of interference. This dual protection system enables the UAV to not only ensure the secure transmission of navigation instructions but also independently reconstruct the signal reception path and filtering parameters to form a dynamic defense barrier when encountering collaborative electronic attacks. At the same time, the deep coupling of the anti-interference module and the navigation control module enables the flight attitude correction and signal purification processing to form a linkage response mechanism. Even in extreme cases where some navigation data is lost due to continuous interference, effective control instructions can still be restored through feature compensation algorithms, significantly enhancing the survival ability and mission execution continuity in complex battlefield environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 is the overall system block diagram of the present invention;

[0051] Figure 2 is the system block diagram of the navigation control module in the present invention;

[0052] Figure 3 is the system block diagram of the adaptive anti-interference module in the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] In order to make the objectives, technical solutions, and advantages of the present invention clearer and more understandable, 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 used to limit the present invention.

[0054] Embodiment:

[0055] Refer to Figures 1-3 ,

[0056] An encrypted signal receiving and guiding system for drones, the system includes: a signal encryption module, a signal receiving module, a signal decryption module, a data processing module, a navigation control module, and an adaptive anti-jamming module;

[0057] Inside the navigation control module, there are: a flight controller, an attitude resolver, and an actuator module;

[0058] Inside the adaptive anti-jamming module, there are: a jamming detector module, a jamming feature analyzer module, and an adaptive filter module;

[0059] The output port of the signal encryption module and the RF input port of the signal receiving module establish a physical layer connection through a wireless channel, and the encrypted data stream enters the RF front end through the antenna;

[0060] The baseband output port of the signal receiving module is directly connected to the input buffer of the signal decryption module through a digital interface to transmit the down-converted encrypted baseband signal;

[0061] The output verification port of the signal decryption module is connected to the input queue of the data processing module through a security bus to transfer the decrypted original navigation message;

[0062] The attitude data output port of the data processing module is connected to the instruction input port of the navigation control module through a high-speed serial interface to convey the resolved three-dimensional coordinates and velocity vector;

[0063] The detection probe of the adaptive anti-jamming module is embedded in the intermediate frequency processing link of the signal receiving module, and its parameter control port is connected to the gain adjustment node of the RF front end and the filter coefficient loading interface of the baseband processing unit through a parallel bus to form a feedforward-feedback composite control loop.

[0064] The signal encryption module realizes the dynamic generation and secure distribution of multi-level keys through a key manager to ensure that the encryption system has forward security. The key manager adopts a key exchange protocol based on elliptic curve cryptography, automatically updates the session key every 30 seconds, and cooperates with the encryption processor to complete the real-time data stream encryption. Taking the SM4 national cryptographic algorithm as an example, the processor uses the CTR mode of block cipher under the support of hardware acceleration, divides the navigation data stream into 128-bit blocks for parallel encryption, and at the same time retains the timestamp field as the initial vector, which not only ensures the encryption efficiency but also prevents replay attacks. The key manager is deeply integrated with the encryption algorithm library and supports the emergency switching of encryption standards during wartime. When a quantum computing threat is detected, it can seamlessly transition to a quantum-resistant encryption algorithm.

[0065] The signal receiving module adopts a software-defined radio architecture and realizes full-coverage reception in the 1.2 - 1.6 GHz frequency band through a tunable bandpass filter bank. In a strong electromagnetic interference environment, the low-noise amplifier uses temperature compensation technology and cooperates with a digital control attenuator to construct a dynamic gain control system to ensure that the signal amplitude is stable in the optimal ADC quantization range. The baseband processing unit uses a polyphase filter bank for channelization processing, decomposes the broadband signal into 8 parallel 200 MHz sub-channels, and uses the parallel computing ability of the FPGA to complete 64th-order FIR filtering and digital down-conversion in real time. A specially designed image rejection algorithm can eliminate the false signals caused by local oscillator leakage, making the receiving sensitivity reach the order of -135 dBm.

[0066] The signal decryption module adopts a heterogeneous computing architecture. The dedicated cryptographic chip processes the decryption of the underlying data stream, and the general-purpose processor executes the decryption of the protocol layer. For the AES-GCM encrypted data stream, while the hardware decryption engine processes 128-bit ciphertext blocks at wire speed, it synchronously generates an authentication tag for data integrity verification. The key manager implements a two-factor authentication mechanism, which fuses the preset long-term key and the temporarily obtained key in real time through the hash chain algorithm to generate a working key, which not only meets the requirements of rapid deployment during wartime but also can prevent the risk of key leakage. The decryption process embeds an abnormal traffic monitoring function, and automatically triggers the key rotation process for consecutive data packets with decryption failures.

[0067] The data processing module adopts an improved PBCH de-encapsulation technology to separate three types of data streams, namely positioning, timing, and integrity, from the decrypted navigation message. The coordinate converter builds a quaternion Kalman filter, fuses the conversion parameters of the Beidou coordinate system, WGS-84 coordinate system, and local grid coordinate system, and compensates for the coordinate offset caused by crustal plate movement in real time. The speed calculator introduces a Doppler frequency shift compensation algorithm, smooths the pseudorange observations through carrier phase, and resolves the three-dimensional velocity vector within a 10 ms cycle. A specially designed multipath rejection processing unit uses the time-domain correlation analysis of the direct signal and the reflected signal to improve the positioning accuracy from the meter level to the sub-meter level.

[0068] The flight controller adopts a model predictive control framework to construct a three-dimensional trajectory tracking system. For the non-linear dynamic characteristics of the unmanned aerial vehicle in a complex airflow environment, a rolling optimization strategy based on Lyapunov stability is designed. During the path correction process, the controller resolves the lateral and longitudinal deviations between the actual flight path and the planned flight path in real time, linearizes the attitude dynamics equation through the Jacobian matrix, and generates a composite control instruction including thrust distribution and rudder deflection. A specially developed wind disturbance observer fuses the barometer and inertial navigation data to establish a turbulence intensity estimation model, and injects the wind field disturbance within the prediction time domain into the control law as a feedforward compensation amount, enabling the unmanned aerial vehicle to maintain centimeter-level trajectory tracking accuracy under the condition of a six-level crosswind;

[0069] The attitude resolver constructs a deep coupling resolution architecture for the nine-axis MEMS sensor based on an improved extended Kalman filter. Aiming at the integral drift problem of the gyroscope caused by high-speed maneuvers, a gradient descent method with motion acceleration compensation is designed to convert the accelerometer data into a quaternion correction amount. When it is detected that the magnetometer is interfered by the body current, it automatically switches to the pure inertial resolution mode and constructs a virtual heading reference system using angular velocity differentiation. The vibration suppression unit embedded in the resolver identifies the propeller resonance frequency through frequency domain analysis and uses an adaptive notch filter to eliminate vibration noise of 50 - 200 Hz, enabling the pitch angle resolution accuracy to still reach the 0.1-degree level under a 10g acceleration shock.

[0070] The brushless motor drive unit of the actuator module adopts space vector pulse width modulation technology and cooperates with a full carbon fiber reduction gearbox to achieve millisecond-level torque response. Aiming at the sudden change characteristics of the aerodynamic load of the control surface, a three-closed-loop control structure with feedforward compensation is developed, and a non-linear friction identification algorithm based on Lissajous figures is embedded in the position loop. The electronic speed control module integrates a temperature-current coupling model and dynamically adjusts the switching frequency by real-time monitoring of the MOSFET junction temperature to maintain the linearity of the thrust output in the environmental range of -40°C to 85°C. A specially designed anti-saturation integrator can prevent the actuator stroke from exceeding the limit and ensure that the drone does not experience servo jamming at a control command update frequency of 20 times per second.

[0071] The interference detector module fuses cyclic spectrum analysis and wavelet transform to construct a time-frequency joint detection model. By calculating the spectral correlation density function and multi-scale wavelet coefficient energy distribution of the received signal in real time, an interference discrimination mechanism based on multiple hypothesis testing is established. In the time domain dimension, an energy detector with a dynamic threshold is designed, and the sliding window variance analysis technology is used to identify abnormal power fluctuations; in the frequency domain dimension, an improved cyclic stationary feature extraction algorithm is adopted, and the typical interference signal characteristics are matched by constructing a cyclic frequency spectrum line cluster. For time-varying threats such as frequency hopping interference, an online learning unit is embedded in the module, and the interference feature database is dynamically updated based on kernel density estimation, and the fractal dimension calculation is introduced to perform non-linear discrimination on transient interference;

[0072] Among them: The calculation formula of the joint detection statistic is:

[0073]

[0074] Among them:

[0075] represents the spectral correlation density function of the received signal at the cyclic frequency \(\alpha\) and the spectral frequency \(f\), reflecting the cyclic stationary characteristics of the signal;

[0076] \(\alpha_0\) to \(\alpha_1\) represent the defined cyclic frequency interval for key monitoring, which is set to [100 Hz, 10 kHz] according to the typical characteristics of the UAV navigation signal;

[0077] W is the detail coefficient matrix obtained by the lifted wavelet transform, representing the time-frequency localization characteristics of the signal;

[0078] λ is an adaptive weighting factor, dynamically adjusted according to the environmental noise floor, with a value range of [0.5, 1.2];

[0079] trace(W·W H ) is to calculate the energy accumulation of wavelet coefficients, used to capture the transient characteristics of burst interference;

[0080] The formula extracts the continuous distribution information of cyclostationary characteristics through integral operation, and at the same time uses matrix trace operation to capture the energy mutation in the wavelet domain. The two-dimensional features are non-linearly weighted and fused to form a sensitive and robust detection statistic, which will improve the interference recognition rate compared with traditional single-dimensional detection methods.

[0081] The interference feature analyzer module adopts a hybrid analysis method combining time-frequency ridge tracking and high-order tensor decomposition, and fine-grained models the interference signal by constructing a multi-dimensional feature space. In the time-frequency analysis layer, a ridge extraction algorithm based on adaptive Gaussian chirp wavelet is designed, and the instantaneous frequency change rate of the interference signal is dynamically matched by optimizing the window function bandwidth, accurately capturing the dwell time and frequency point switching law of the frequency hopping interference. Among them:

[0082] The calculation formula of the adaptive ridge tracking cost function is:

[0083]

[0084] Among them: TF(t,f) represents the time-frequency distribution matrix, generated by the improved synchrosqueezed wavelet transform, with energy aggregation characteristics;

[0085] represents calculating the gradient change of time-frequency energy along the frequency axis, strengthening the continuity constraint of the instantaneous frequency trajectory;

[0086] TV(f(t)) is the total variation regularization term of the frequency trajectory, suppressing the generation of false ridge branches;

[0087] β is an adaptive balance factor, dynamically adjusted according to the interference signal bandwidth, satisfying the intelligent adaptation of taking 0.3 for narrowband interference and 1.5 for broadband interference;

[0088] The formula fuses the gradient constraint and variational regularization mechanism, and can still accurately reconstruct the fast frequency hopping pattern with the frequency mutation point interval less than 5 μs under strong noise background. The ridge line positioning error is reduced by 62% compared with the traditional dynamic programming method, providing high-precision feature input for subsequent interference type recognition.

[0089] The adaptive filter module constructs a hybrid processing mechanism of frequency-domain subspace projection and time-domain recursive filtering, and realizes precise spectrum shaping by establishing a dynamic mapping relationship between interference characteristics and filter parameters. For comb-shaped spectrum interference, a frequency-domain stopband generation algorithm based on constrained convex optimization is designed, and an improved lattice filter structure is used to form a notch response with a depth of up to 60 dB at specified frequencies. For wideband noise interference, a normalized least mean square algorithm with a time-varying step size is developed, and the spectral feature prior information provided by the interference feature analyzer is combined to dynamically adjust the convergence factor and filter order. A space-time joint processing scheme is particularly proposed, which performs Kronecker product fusion on the spatial steering vector of the array signal and the time-domain weight matrix of adaptive filtering to form a multi-dimensional filtering space;

[0090] The calculation formula of the hybrid optimization objective function is as follows:

[0091]

[0092] Where:

[0093] H(ω) represents the filter frequency-domain response function, and the stopband characteristics are described by parametric B-spline curves;

[0094] Y(ω) and $X(\omega)$ represent the spectral components of the filter output and the desired signal respectively

[0095] λ is the time-domain correlation weight coefficient, which is dynamically adjusted according to the interference pulse repetition period, and the value range is [0.1, 5]; Φ(τ) is designed as an exponentially decaying window function, which is used to strengthen the extraction of correlation peaks at key time delay points;

[0096] η controls the sparse constraint intensity, and is evaluated by KL divergence and adjusted in real time to balance the convergence speed and steady-state error;

[0097] is a time-domain gradient morphological operator, which is used to suppress the waveform distortion of the filtered signal; the formula reduces the signal distortion entropy by 45% under strong interference background through the triple constraints of spectrum distortion, time-domain correlation and structural sparsity, and especially expands the suppression bandwidth of LFM swept-frequency interference to 2.3 times that of the conventional algorithm, significantly improving the fidelity of navigation signals in complex electromagnetic environments.

[0098] It can be known from the above that:

[0099] In the present invention, the adaptive anti-interference module realizes the real-time identification, feature analysis and dynamic suppression of interference signals through a multi-dimensional joint perception and closed-loop regulation system. The interference detector integrates time-frequency domain deep feature extraction technology and can capture the existence features of complex interferences such as frequency hopping and frequency scanning within a millisecond-level time window; the interference feature analyzer is based on non-linear dynamics modeling and pattern learning algorithms to accurately deconstruct the modulation rules and propagation characteristics of interference signals; the adaptive filter adopts a multi-modal optimization strategy and combines the spatio-temporal correlation features of navigation signals to maximize the integrity of the original signal while eliminating interference. The whole mechanism forms a complete anti-interference closed loop from environmental perception to dynamic response, ensuring that the UAV can still maintain a stable encrypted communication link and centimeter-level navigation accuracy in extreme scenarios such as strong electromagnetic suppression and dense signal conflicts.

[0100] In the present invention, the dynamic key management and anti-quantum algorithm design of the signal encryption module block the risks of signal interception and malicious cracking from the source, while the adaptive anti-interference module can actively predict and avoid potential threat areas by intelligently learning the evolution law of interference. This dual protection system enables the UAV to not only ensure the secure transmission of navigation instructions but also autonomously reconstruct the signal reception path and filtering parameters to form a dynamic defense barrier when encountering collaborative electronic attacks. At the same time, the deep coupling of the anti-interference module and the navigation control module enables the flight attitude correction and signal purification processing to form a linkage response mechanism. Even in extreme cases where some navigation data is lost due to continuous interference, effective control instructions can still be restored through the feature compensation algorithm, significantly enhancing the survival ability and mission execution continuity in complex battlefield environments.

[0101] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the existence of additional identical elements in the process, method, article or device comprising the element.

[0102] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. An encrypted signal receiving and guiding system for an unmanned aerial vehicle, characterized in that: The system includes: a signal encryption module, a signal reception module, a signal decryption module, a data processing module, a navigation control module, and an adaptive anti-jamming module; Inside the navigation control module, there are: a flight controller, an attitude resolver, and an actuator module; Inside the adaptive anti-jamming module, there are: an interference detector module, an interference feature analyzer module, and an adaptive filter module; The output port of the signal encryption module is physically connected to the RF input port of the signal reception module through a wireless channel, and the encrypted data stream enters the RF front-end through the antenna; The baseband output port of the signal reception module is directly connected to the input buffer of the signal decryption module through a digital interface to transmit the down-converted encrypted baseband signal; The output verification port of the signal decryption module is connected to the input queue of the data processing module through a security bus to transfer the decrypted original navigation message; The attitude data output port of the data processing module is connected to the command input port of the navigation control module through a high-speed serial interface to convey the solved three-dimensional coordinates and velocity vectors; The detection probe of the adaptive anti-jamming module is embedded in the intermediate frequency processing link of the signal reception module, and its parameter control port is connected across to the gain adjustment node of the RF front-end and the filter coefficient loading interface of the baseband processing unit through a parallel bus to form a feedforward-feedback composite control loop.

2. The encrypted signal receiving and guiding system for a drone according to claim 1, characterized in that: The signal encryption module realizes the dynamic generation and secure distribution of multi-level keys through a key manager. The key manager adopts a key exchange protocol based on elliptic curve cryptography, automatically updates the session key every 30 seconds, and cooperates with the encryption processor to complete the real-time data stream encryption.

3. The encrypted signal receiving and guiding system for a drone according to claim 1, characterized in that: The signal reception module adopts a software-defined radio architecture, realizes full-coverage reception in the 1.2 - 1.6 GHz frequency band through a tunable bandpass filter bank. The low-noise amplifier uses temperature compensation technology, and cooperates with a digital control attenuator to construct a dynamic gain control system to ensure that the signal amplitude is stable in the optimal ADC quantization range; the baseband processing unit uses a polyphase filter bank for channelization processing, decomposes the broadband signal into 8 parallel 200 MHz sub-channels, and uses the parallel computing ability of the FPGA to complete 64-order FIR filtering and digital down-conversion in real time.

4. An encrypted signal receiving and guiding system for a drone according to claim 1, characterized in that: The signal decryption module adopts a heterogeneous computing architecture, where a dedicated cryptographic chip processes the decryption of the underlying data stream and a general-purpose processor executes the protocol layer decryption; for the AES-GCM encrypted data stream, while the hardware decryption engine processes 128-bit ciphertext blocks at wire speed, it simultaneously generates an authentication tag for data integrity verification.

5. An encrypted signal receiving and guiding system for a drone according to claim 1, characterized in that: The data processing module adopts an improved PBCH de-encapsulation technology to separate three types of data streams of positioning, timing, and integrity from the decrypted navigation message; The coordinate converter builds a quaternion Kalman filter, fuses the conversion parameters of the Beidou coordinate system, WGS-84 coordinate system, and local grid coordinate system, and compensates in real time for the coordinate offset caused by crustal plate movement.

6. The encrypted signal receiving and guiding system for a drone according to claim 1, wherein: The flight controller constructs a three-dimensional trajectory tracking system using a model predictive control framework. For the non-linear dynamic characteristics of the unmanned aerial vehicle in a complex airflow environment, a rolling optimization strategy based on Lyapunov stability is designed; During the path correction process, the controller calculates the lateral and longitudinal deviations between the actual trajectory and the planned trajectory in real time, linearizes the attitude dynamics equations through the Jacobian matrix, and generates composite control instructions including thrust distribution and rudder deflection. The attitude solver is based on an improved extended Kalman filter and constructs a deep coupling solution architecture for a nine-axis MEMS sensor. To address the gyroscope integral drift problem caused by high-speed maneuvers, a gradient descent method with motion acceleration compensation is designed to convert accelerometer data into quaternion corrections.

7. The encrypted signal receiving and guiding system for a drone according to claim 1, characterized in that: The brushless motor drive unit of the actuator module adopts space vector pulse width modulation technology and cooperates with an all-carbon fiber reduction gearbox to achieve millisecond-level torque response; in response to the sudden change characteristics of the aerodynamic load on the rudder, a three-closed-loop control structure with feedforward compensation is developed, and a nonlinear friction identification algorithm based on Lissajous figures is embedded in the position loop; the electronic control module integrates a temperature-current coupling model, and dynamically adjusts the switching frequency by real-time monitoring of the MOSFET junction temperature, maintaining thrust output linearity in an environmental range of -40°C to 85°C.

8. An encrypted signal receiving and guiding system for a drone as described in claim 1, characterized in that: The interference detector module integrates cyclic spectrum analysis and wavelet transform to construct a time-frequency joint detection model. By calculating the spectral correlation density function and multi-scale wavelet coefficient energy distribution of the received signal in real time, it establishes an interference discrimination mechanism based on multiple hypothesis testing. In the time domain, an energy detector with a dynamic threshold is designed, and a sliding window variance analysis technique is used to identify abnormal power fluctuations. In the frequency domain, an improved cyclostationary feature extraction algorithm is used to construct cyclic frequency spectrum line clusters to match typical interference signal characteristics. To address time-varying threats such as frequency-hopping interference, the module has an embedded online learning unit that dynamically updates the interference signature database based on kernel density estimation, and introduces fractal dimension calculation for nonlinear discrimination of transient interference. Among them: The calculation formula of the joint detection statistic is: in: Denote the spectral correlation density function of the received signal at cyclic frequency \(\alpha\) and spectral frequency \(f\), which reflects the cyclostationary characteristics of the signal; α0 to α1 represent the cyclic frequency interval for key monitoring, which is set to [100 Hz, 10 kHz] based on the typical characteristics of UAV navigation signals; W is the detail coefficient matrix obtained by lifting wavelet transform, which represents the time-frequency localization characteristics of the signal; λ is an adaptive weighting factor, which is dynamically adjusted according to the ambient noise floor and has a value range of [0.5, 1.2]; trace(W·W H ) is used to calculate the energy accumulation of wavelet coefficients for capturing the transient characteristics of burst interference; The formula extracts the continuous distribution information of cyclostationary features through integral operation, and uses matrix trace operation to capture energy mutations in the wavelet domain. The two-dimensional features are fused through nonlinear weighted fusion to form a sensitive and robust detection statistic, which will improve the interference recognition rate compared with the traditional single-dimensional detection method.

9. The encrypted signal receiving and guiding system for a drone according to claim 1, characterized in that: The interference feature analyzer module adopts a hybrid analysis method that combines time-frequency ridge tracking with high-order tensor decomposition to finely model the interference signal by constructing a multi-dimensional feature space. At the time-frequency analysis layer, a ridge extraction algorithm based on adaptive Gaussian chirped wavelet is designed to dynamically match the instantaneous frequency change rate of the interference signal by optimizing the window function bandwidth, accurately capturing the dwell time and frequency switching pattern of frequency hopping interference. The calculation formula of the adaptive ridge tracking cost function is: Where: TF(t,f) represents the time-frequency distribution matrix, which is generated by the improved synchro-compressed wavelet transform and has the energy aggregation characteristic; It represents the gradient change of the calculated time-frequency energy along the frequency axis, strengthening the continuity constraint of the instantaneous frequency trajectory; TV(f(t)) is the total variation regularization term of the frequency trajectory, which suppresses the generation of false ridge line branches; β is an adaptive balance factor, which is dynamically adjusted according to the interference signal bandwidth, and satisfies the intelligent adaptation of taking 0.3 for narrowband interference and 1.5 for broadband interference; By fusing the gradient constraint and the variational regularization mechanism, the formula can still accurately reconstruct the fast frequency hopping pattern with the interval of frequency mutation points less than 5 μs under the strong noise background. Compared with the traditional dynamic programming method, the ridge line positioning error is reduced by 62%, providing high-precision feature input for subsequent interference type recognition.

10. The encrypted signal receiving and guiding system for a drone according to claim 1, wherein: The adaptive filter module constructs a hybrid processing mechanism of frequency domain subspace projection and time domain recursive filtering, and realizes precise spectrum shaping by establishing a dynamic mapping relationship between interference features and filter parameters; for comb-shaped spectrum interference, a frequency domain stopband generation algorithm based on constrained convex optimization is designed, and an improved lattice filter structure is used to form a notch response with a depth of up to 60 dB at the specified frequency point; for broadband noise interference, a normalized least mean square algorithm with a time-varying step size is developed, and combined with the prior information of the spectrum features provided by the interference feature analyzer, the convergence factor and the filter order are dynamically adjusted; a space-time joint processing scheme is particularly proposed, and the spatial steering vector of the array signal and the time domain weight matrix of the adaptive filter are fused by Kronecker product to form a multi-dimensional filtering space; The calculation formula of the hybrid optimization objective function is as follows: Where: H(ω) represents the filter frequency domain response function, and the stopband characteristics are described by a parametric B-spline curve; Y(ω) and $X(\omega) represent the spectral components of the filter output and the desired signal respectively. λ is the time domain correlation weight coefficient, which is dynamically adjusted according to the interference pulse repetition period, and the value range is [0.1, 5]; Φ(τ) is designed as an exponentially decaying window function, which is used to strengthen the extraction of the correlation peak at the key time delay point; η controls the sparse constraint intensity, and is evaluated by KL divergence and adjusted in real time to balance the convergence speed and the steady-state error; It is a time-domain gradient morphological operator used to suppress the waveform distortion of the filtered signal. Through the triple constraints of spectral distortion, time-domain correlation, and structural sparsity, the formula reduces the signal distortion entropy by 45% under strong interference background. Especially for the suppression bandwidth of LFM swept-frequency interference, it is extended to 2.3 times that of the conventional algorithm, significantly improving the fidelity of navigation signals in complex electromagnetic environments.

Images