Low-power-consumption star flash control module applied to unmanned aerial vehicle and dummy pilot control method

Abstract

The invention relates to the technical field of unmanned aerial vehicles, in particular to a low-power-consumption star flash control module applied to an unmanned aerial vehicle and a dummy pilot control method. The method comprises the following steps: acquiring an original control instruction of a real flying hand, and converting the original control instruction into an encrypted pulse signal containing a dynamic key; fusing a frequency spectrum camouflage signal simulating the environment background of a dummy pilot into the encrypted pulse signal, and then transmitting the encrypted pulse signal to the dummy pilot in a self-adaptive frequency hopping mode; synchronously analyzing the dynamic key at a dummy flying hand end and stripping the frequency spectrum camouflage signal to generate a virtual control track matched with the real operation behavior characteristics; and reversely modulating an actual response signal of the unmanned aerial vehicle based on the virtual control track, so that the unmanned aerial vehicle executes an action and forms dynamic error compensation with the virtual control track. According to the invention, a real flying hand can be effectively protected from electronic reconnaissance positioning and attack of an enemy, and the concealment is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles, in particular to a low-power XingShan control module applied to unmanned aerial vehicles and a control method for a dummy pilot. Background Art

[0002] In modern warfare, unmanned aerial vehicles have become key equipment due to their efficient and flexible combat capabilities. However, the control mode of traditional pilots directly connecting to unmanned aerial vehicles via radio has serious security flaws. The pilot's control signals are easily intercepted by the enemy's electronic reconnaissance system during transmission. Through spectrum analysis or thermal imaging positioning technology, the risk of the pilot being exposed is extremely high, and thus the pilot becomes the primary target for attack. To solve this problem, the dummy pilot technology has emerged. Its core lies in constructing a command link that combines virtual and real: by deploying a dummy pilot with bionic characteristics as an instruction transfer node, it is difficult for the enemy to distinguish the real control source. The dummy pilot adopts a bionic thermal radiation structure (such as a high-temperature area at the head simulating the human body temperature distribution and a constant-temperature area in the torso) and dynamic spectrum camouflage technology to effectively counter infrared detection and radio positioning. At the same time, through the low-power encrypted transmission of the XingShan protocol, after the real pilot's instructions are encrypted with a dynamic key, they are forwarded to the dummy pilot in an adaptive frequency hopping manner, and then the precise control of the unmanned aerial vehicle is achieved through a virtual trajectory generation and error compensation mechanism. This three-layer architecture of "real person - dummy person - unmanned aerial vehicle" not only retains the real-time nature of direct control but also greatly improves the battlefield survival ability through physical separation and electronic deception means. However, the existing dummy pilot control technology still has technical difficulties such as insufficient dynamic key synchronization accuracy leading to instruction delay and the spectrum camouflage signal being easily detected by advanced electronic reconnaissance equipment, making it difficult to effectively and reliably protect the safety and concealment of real pilots in complex battlefield environments. Summary of the Invention

[0003] The present invention overcomes the deficiencies of the prior art and provides a low-power XingShan control module applied to unmanned aerial vehicles and a control method for a dummy pilot.

[0004] The technical solution adopted by the present invention to achieve the above object is as follows: The first aspect of the present invention discloses a control method for a dummy pilot applied to an unmanned aerial vehicle, including the following steps: Obtain the original control instructions of the real pilot and convert the original control instructions into encrypted pulse signals containing dynamic keys; Integrate a spectrum camouflage signal simulating the environmental background of the dummy pilot into the encrypted pulse signal and transmit it to the dummy pilot in an adaptive frequency hopping mode; Synchronously analyze the dynamic key at the dummy pilot end and strip the spectrum camouflage signal to generate a virtual control trajectory matching the characteristics of real operation behaviors; Based on the virtual control trajectory, the actual response signal of the UAV is reversely modulated to enable the UAV to perform actions and form dynamic error compensation with the virtual control trajectory.

[0005] Preferably, obtain the original control instructions of a real pilot, and convert the original control instructions into an encrypted pulse signal containing a dynamic key, specifically: Real-time capture the hand myoelectric signals, nerve impulse frequencies, and eye movement tracking vectors of a real pilot through a wearable sensing array to form a multi-dimensional biological control feature stream; Perform denoising processing on the multi-dimensional biological control feature stream through a normalization convolution kernel, and segment and align the multi-dimensional biological control feature stream according to timestamps to generate a control instruction feature matrix with spatio-temporal correlation; Based on the entropy value distribution of the control instruction feature matrix, use a chaotic mapping algorithm to generate a dynamic key seed bound to the current control behavior, and perform key chaotic diffusion in combination with the real-time flight attitude parameters of the UAV to output a segmented time-varying encryption key; Input the dynamic key seed and the segmented time-varying encryption key into a StarFlash protocol converter, and perform key cross-bit interleaving encoding according to a preset protocol structure to generate a pulse baseband sequence containing a key fingerprint identifier; Use a quantum random number generator to generate an aperiodic noise masking factor, and perform phase-amplitude dual perturbation on the pulse baseband sequence to form an encrypted pulse signal with statistical characteristics similar to environmental noise.

[0006] Preferably, after integrating a spectrum camouflage signal simulating the environmental background of a dummy pilot into the encrypted pulse signal, it is transmitted to the dummy pilot in an adaptive frequency hopping mode, specifically: Obtain the environmental electromagnetic background noise within the preset range of the dummy pilot, extract the energy ridge line distribution characteristics in the signal time-frequency domain, and construct a dynamic camouflage spectrum fingerprint library; Based on the statistical characteristics of the dynamic camouflage spectrum fingerprint library, use the orthogonal subspace projection theorem to generate a camouflage noise base with the same time-frequency ridge line characteristics, and perform multi-dimensional phase synchronous perturbation superposition on the camouflage noise base and the encrypted pulse signal; According to the instantaneous bandwidth characteristics of the superimposed signal, extract the zero-crossing distribution of the power spectral density envelope to generate a pseudo-random frequency hopping sequence template matching the time-frequency characteristics of the current environmental electromagnetic background noise; Input the camouflage noise base and the pseudo-random frequency hopping sequence template into a signal synthesizer, perform time slot dynamic allocation based on phase continuity, and reconstruct a pseudo-carrier waveform with environmental adaptability while retaining the phase jump characteristics of the original encrypted pulse signal; Obtain the real-time motion state of the dummy pilot, and determine the real-time Doppler frequency shift parameter according to the real-time motion state; Perform time slot compression compensation on the pseudo-random frequency hopping sequence template according to the real-time Doppler frequency shift parameter, and transmit it to the dummy pilot after making the frequency deviation characteristics of the pseudo-carrier waveform consistent with the dynamic change of the background noise in terms of spectrum.

[0007] Preferably, synchronously analyze the dynamic key at the dummy pilot end and strip the spectrum camouflage signal to generate a virtual control trajectory matching the characteristics of real operation behavior. Specifically: Perform dynamic phase correction on the signal received at the dummy pilot end, generate a phase synchronization calibration factor based on the time slot compression parameter of the pseudo-random frequency hopping sequence template, and output a time-frequency aligned baseband signal after eliminating the transmission path distortion; Combine the energy ridge line distribution characteristics in the dynamic camouflage spectrum fingerprint library to construct a time-varying notch filter bank matching the camouflage noise base, and strip the superimposed spectrum camouflage components in the baseband signal frame by frame through the ridge line matching degree threshold to generate a spectrum purification signal; Extract the key fingerprint identification of the cross-bit interleaved coding structure from the spectrum purification signal, reverse calculate the bit interleaving pattern in combination with the preset satellite flash protocol, and separate the dynamic key seed and the chaotic residual vector of the segmented time-varying encryption key; Input the chaotic residual vector into the chaotic inverse diffusion model constrained by the real-time flight attitude parameters of the unmanned aerial vehicle to reconstruct the spatio-temporal correlation topology structure of the original control instruction feature matrix; Perform biometric behavior simulation on the original control instruction feature matrix based on the biometric coupling weights of the electromyogram signal and the eye movement tracking vector to generate virtual trajectory primitives containing microscopic operation jitters; Perform dynamic compensation for the dynamics of the unmanned aerial vehicle on the virtual trajectory primitives through a time-frequency domain sliding window fitter, and dynamically adjust the trajectory curvature and phase continuity constraint boundaries according to the attitude parameters to generate a virtual control trajectory.

[0008] Preferably, reverse modulate the actual response signal of the unmanned aerial vehicle based on the virtual control trajectory to make the actions performed by the unmanned aerial vehicle form a dynamic error compensation with the virtual control trajectory. Specifically: Extract the virtual trajectory primitives of the virtual control trajectory, and construct a time-delay coupling matrix between the trajectory primitives and the current pose of the unmanned aerial vehicle according to the real-time flight attitude parameters of the unmanned aerial vehicle and in combination with the virtual trajectory primitives; Use the time-delay coupling matrix to perform spatio-temporal interpolation on the phase jump characteristics of the actual response signal, introduce an inertial compensation amount based on the dynamic weight of the trajectory curvature in the attitude control loop, and generate an attitude correction gradient containing non-linear perturbations; Perform backpropagation iteration on the attitude correction gradient according to the Jacobian matrix of the unmanned aerial vehicle dynamics model, and combine the angular velocity deviation vector output by the inertial measurement unit of the aircraft to determine the frequency domain energy entropy of the trajectory tracking error; When the frequency-domain energy entropy exceeds the preset stealth control threshold, perform double-phase-amplitude constraints on the attitude correction gradient, and generate a compensation amount chaos injection sequence based on the chaotic residual vector of the dynamic key seed; Perform time-frequency domain convolution fusion on the compensation amount chaos injection sequence and the virtual trajectory primitive to obtain the trajectory-noise coupling entropy value, and adjust the inertial weight coefficient of the convolution kernel based on the trajectory-noise coupling entropy value to generate an environment-adaptive inverse modulation signal; Use the phase synchronization calibration factor to perform Doppler frequency shift compensation on the inverse modulation signal, so that it is time slot synchronized with the pseudo-random frequency hopping sequence template of the encrypted pulse signal, and output the actual response control instruction distributed in the same frequency band as the environmental background noise.

[0009] The dummy pilot control method further includes the following steps: Extract the phase entropy change index and spectral distortion residual of the encrypted pulse signal at the receiving end of the star flash feedback link, generate a signal integrity check code based on the chaotic residual vector of the dynamic key seed, and construct a time-frequency domain distortion feature matrix; Determine the spectral confusion degree between the environmental noise and the camouflage signal in the transmission path according to the time-frequency ridge line distribution of the distortion feature matrix; Based on the gradient descent direction of the spectral confusion degree, use the orthogonal subspace projection theorem to generate a time-frequency confusion factor, and inject the time-frequency confusion factor into the time slot allocator of the pseudo-random frequency hopping sequence template to generate an adaptive frequency hopping parameter with the characteristic of phase mutation concealment; Synchronously extract the Doppler frequency shift covariance matrix in the real-time flight attitude parameters of the UAV, and dynamically weight the time-frequency ridge line energy of the camouflage noise base combined with the chaotic mapping residual to generate a camouflage signal strength adjustment coefficient that is phase synchronized with the frequency hopping parameter; Input the strength adjustment coefficient into the time-varying notch filter bank, and feedback and close-loop control the spectral overlap degree between the camouflage noise base and the background noise through the ridge line matching degree. When the spectral overlap degree is not within the dynamic stealth interval, update the camouflage energy weight of the spectral fingerprint library based on the inverse operation of the Jacobian matrix; Use the phase entropy change index to perform secondary encryption of the reconstructed frequency hopping parameter and camouflage signal strength according to the star flash protocol, and generate an environment-adaptive stealth instruction set including a synchronization calibration factor.

[0010] The second aspect of the present invention discloses a low-power star flash control module, which is applied to any one of the dummy pilot control methods, including: A star flash control unit, integrated with the Hi3861V100 main control chip, supporting dual-mode communication of the star flash protocol stack and BLE5.3; A power management unit, supporting 0.8-3.3V dynamic voltage regulation, and the sleep power consumption ≤5μA; The StarFlash instruction generation unit is used to encrypt the original control instruction and generate an encrypted pulse signal containing a dynamic key; The adaptive frequency hopping transmission unit integrates a narrowband RF chip and a spectrum camouflage engine, enabling the encrypted pulse signal to pseudo-randomly hop within a preset frequency band; The dummy pilot unit is designed with a bionic thermal radiation structure, and a multi-region independent temperature control unit is arranged inside the bionic shell. Each unit is divided into a head high-temperature area, a torso constant-temperature area, and limb gradient temperature areas according to human anatomical characteristics; The StarFlash feedback verification unit real-time monitors the signal transmission quality through a closed-loop link and triggers dynamic parameter adjustment.

[0011] The present invention solves the technical defects in the background art and has the following beneficial effects: obtaining the original control instruction of a real pilot, converting the original control instruction into an encrypted pulse signal containing a dynamic key; integrating a spectrum camouflage signal simulating the background environment of a dummy pilot into the encrypted pulse signal and transmitting it to the dummy pilot in an adaptive frequency hopping mode; synchronously parsing the dynamic key at the dummy pilot end and stripping the spectrum camouflage signal to generate a virtual control trajectory matching the characteristics of real operation behaviors; based on the virtual control trajectory, reversely modulating the actual response signal of the unmanned aerial vehicle, so that the actions executed by the unmanned aerial vehicle form a dynamic error compensation with the virtual control trajectory. The present invention can effectively protect real pilots from enemy electronic reconnaissance, positioning, and attacks, and improve concealment. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0013] Figure 1 It is the first method flow chart of this dummy pilot control method; Figure 2 It is the second method flow chart of this dummy pilot control method; Figure 3 It is a schematic diagram of the structure of this low-power StarFlash control module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0014] In order to be able to more clearly understand the above-mentioned objects, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the drawings and specific embodiments. It should be noted that, without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other.

[0015] In the following description, many specific details are set forth in order to provide a thorough understanding of the present invention. However, the present invention may be practiced in other ways than those specifically described herein. Therefore, the scope of protection of the present invention is not limited by the specific embodiments disclosed below.

[0016] As Figure 1 shown, a dummy pilot control method applied to an unmanned aerial vehicle according to a first aspect of the present invention includes the following steps: S102. Obtain the original control instructions of a real pilot, and convert the original control instructions into an encrypted pulse signal including a dynamic key; S104. Incorporate a spectrum camouflage signal simulating the background environment of a dummy pilot into the encrypted pulse signal, and transmit it to the dummy pilot in an adaptive frequency hopping mode; S106. Synchronously analyze the dynamic key at the dummy pilot end and strip the spectrum camouflage signal to generate a virtual control trajectory matching the characteristics of real operation behaviors; S108. Based on the virtual control trajectory, reversely modulate the actual response signal of the unmanned aerial vehicle, so that the actions executed by the unmanned aerial vehicle form a dynamic error compensation with the virtual control trajectory.

[0017] It should be noted that the present invention solves the problem that the control signals of real pilots are easily intercepted and located by the enemy. Through dynamic key encryption, spectrum camouflage and adaptive frequency hopping transmission, the concealment of the signals in a complex electromagnetic environment is ensured; the virtual control trajectory generated at the dummy pilot end matches the characteristics of real behaviors, enabling the actions of the unmanned aerial vehicle to perform dynamic error compensation therewith, thereby effectively protecting real pilots from enemy electronic reconnaissance, positioning and attacks, and improving concealment.

[0018] Preferably, obtaining the original control instructions of a real pilot and converting the original control instructions into an encrypted pulse signal including a dynamic key, as Figure 2 shown, specifically: S202. Real-time capture the hand myoelectric signals, nerve impulse frequencies and eye movement tracking vectors of a real pilot through a wearable sensor array to form a multi-dimensional biological control feature stream; S204. Denoise the multi-dimensional biological control feature stream through a normalized convolutional kernel, and segment and align the multi-dimensional biological control feature stream according to time stamps to generate a control instruction feature matrix with spatio-temporal correlation; S206. Based on the entropy value distribution of the control instruction feature matrix, adopt a chaotic mapping algorithm to generate a dynamic key seed bound to the current control behavior, and perform key chaotic diffusion in combination with the real-time flight attitude parameters of the unmanned aerial vehicle to output a segmented time-varying encryption key; Wherein, the real-time flight attitude parameters include the pitch angle, roll angle, yaw angle, three-dimensional acceleration and angular velocity data of the unmanned aerial vehicle.

[0019] It should be noted that the chaotic mapping algorithm utilizes the dynamic change characteristics of the entropy value of the biometric signals of the drone operator (such as electromyogram, nerve impulse, eye movement data), and combines the real-time flight attitude parameters of the drone to generate an encryption key with high randomness and unpredictability. Specifically, the algorithm maps the Shannon entropy value of the biometric matrix to the initial parameters of the chaotic system (such as the μ value of the Logistic mapping), and generates a key seed through chaotic iterative operations; at the same time, the drone attitude data is introduced as a perturbation factor, and the extreme sensitivity of the chaotic system to the initial conditions is utilized to make the generated key strongly correlated with the real-time operation behavior of the drone operator and dynamically bound to the drone motion state. This dual chaotic modulation mechanism based on biometric entropy change and flight parameters ensures the time-varying and unique nature of the key, so that even if the enemy intercepts fragmentary signals, they cannot obtain the complete key sequence through reverse derivation.

[0020] S208. Input the dynamic key seed and the segmented time-varying encryption key into the SparkLink protocol converter, and perform key cross-bit interleaving encoding according to the preset protocol structure to generate a pulse baseband sequence containing a key fingerprint identifier; Among them, the preset protocol structure refers to the key data encapsulation rule predefined in the SparkLink communication protocol. Its core function is to embed the dynamically generated encryption key into the communication frame structure in a specific format to achieve secure transmission and efficient decoding. The protocol adopts a hierarchical frame structure design: the physical layer includes a preamble (8-byte synchronization header) and a frame start delimiter (2 bytes); the data link layer sets a key fingerprint area (8-byte MAC address identifier), a dynamic key payload area (256-bit segmented key), and a check field (16-bit CRC). Among them, the key data is scattered in multiple symbol periods through bit interleaving encoding (interleaving depth 16 bit).

[0021] S210. Use a quantum random number generator to generate an aperiodic noise masking factor, and perform dual phase-amplitude perturbation on the pulse baseband sequence to form an encrypted pulse signal with statistical characteristics similar to environmental noise.

[0022] In a specific embodiment of the present invention, the wearable sensing array includes an 8-channel electromyogram sensor arranged on the forearm of the drone pilot, a neural signal acquisition module, and a head-mounted eye tracker. The three are synchronously transmitted to the embedded processing unit via Bluetooth 5.2, and the three signals are aligned with a time window of 20 ms to generate a 12-dimensional feature vector stream including electromyogram amplitude (range ±5 mV), neural pulse interval (resolution 10 μs), and eyeball deflection angle (range ±60°). The normalized convolution kernel uses a 3×3 Gaussian kernel (σ = 0.8) to perform sliding filtering on the feature stream to eliminate the 50 Hz power frequency interference in the electromyogram signal (attenuation above 40 dB). Then, the signals of each dimension are aligned according to the starting point of the operation action through the dynamic time warping algorithm to form a 128×12 instruction feature matrix (time axis accuracy ±5 ms). The chaotic mapping algorithm maps the entropy value interval [0.3, 1.2] to the initial parameters of the Logistic chaotic system (μ ∈ [3.6, 4.0]) according to the Shannon entropy value of the feature matrix (calculation window width 50 ms), and combines the pitch angle (±90°) and roll angle (±180°) data transmitted back by the drone. After three iterations, a 32-byte dynamic key seed is output, and then a 256-bit segmented key updated every 50 ms is generated through the Lorenz attractor equation. The SparkLink protocol converter performs bitwise cross-recombination on the key seed and the segmented key at a ratio of 4:1 (interleaving depth 16 bit), and adds an 8-byte device MAC address as a fingerprint identifier to generate a GMSK modulation sequence with a baseband symbol rate of 10 Mbps. The quantum random number generator generates a noise mask obeying the Rayleigh distribution (amplitude standard deviation 0.15 V) based on the detection of laser phase noise by a silicon photomultiplier (sampling rate 100 MS / s), and performs joint modulation on the baseband sequence for carrier phase (perturbation range ±π / 6) and amplitude (fluctuation ratio ±12%), and finally outputs an encrypted pulse signal with an equivalent noise factor > 0.95.

[0023] Overall, this method uses multi-dimensional biometric features to generate dynamic keys and combines chaotic encryption technology to make the encrypted signal have high randomness and environmental noise characteristics, effectively resisting enemy electronic reconnaissance and signal cracking; at the same time, by binding the key with the flight attitude in real time, the reliability and anti-interference ability of the instruction transmission are ensured, thereby improving the concealment and survival ability of the real drone pilot.

[0024] Preferably, after integrating a spectrum camouflage signal simulating the environmental background of a dummy drone pilot into the encrypted pulse signal, it is transmitted to the dummy drone pilot in an adaptive frequency hopping mode, specifically: Obtain the environmental electromagnetic background noise within the preset range of the dummy drone pilot, extract the energy ridge line distribution characteristics in the time-frequency domain of the signal, and construct a dynamic camouflage spectrum fingerprint library; Based on the statistical characteristics of the dynamic camouflage spectrum fingerprint library, the orthogonal subspace projection theorem is used to generate a camouflage noise base with the same time-frequency ridge line characteristics, and the camouflage noise base is superimposed on the encrypted pulse signal through multi-dimensional phase synchronous perturbation; According to the instantaneous bandwidth characteristics of the superimposed signal, the zero-crossing distribution of the power spectral density envelope is extracted to generate a pseudo-random frequency hopping sequence template that matches the time-frequency characteristics of the current environmental electromagnetic background noise; The camouflage noise base and the pseudo-random frequency hopping sequence template are input into the signal synthesizer to perform time-slot dynamic allocation based on phase continuity, and while retaining the phase jump characteristics of the original encrypted pulse signal, a pseudo-carrier wave waveform with environmental self-adaptability is reconstructed; Obtain the real-time motion state of the dummy pilot, and determine the real-time Doppler frequency shift parameter according to the real-time motion state; Among them, the real-time motion state of the dummy pilot includes three-dimensional space coordinates, motion speed, acceleration and attitude angle data; the real-time Doppler frequency shift parameter includes the carrier frequency offset, the frequency offset change rate, and the cosine value of the real-time included angle between the motion direction and the signal transmission direction.

[0025] According to the real-time Doppler frequency shift parameter, the pseudo-random frequency hopping sequence template is subjected to time-slot compression compensation, so that the frequency offset characteristics of the pseudo-carrier wave waveform are spectrally consistent with the dynamic change of the background noise and then transmitted to the dummy pilot.

[0026] It should be noted that time-slot compression compensation is a signal processing technology that dynamically adjusts the frequency hopping communication time-slot parameters according to the Doppler effect generated by the real-time motion of the dummy pilot. Its core principle is to determine the real-time change rate of the signal transmission path through the motion state data (speed, acceleration, attitude angle), and accordingly perform dynamic compression or expansion of the frequency hopping time-slot: when the dummy pilot moves towards the signal source, the time-slot length is shortened (compensating for the positive Doppler frequency shift), and when moving away, the time-slot is extended (compensating for the negative Doppler frequency shift), and at the same time, the cosine value of the included angle between the motion direction and the signal transmission direction is combined for three-dimensional space vector correction. The carrier phase continuity is maintained through a digital phase-locked loop, so that the frequency hopping signal can still maintain a frequency offset accuracy of ±50 Hz when experiencing a maximum relative speed of ±50 m / s, ensuring that the spectrum characteristics of the pseudo-carrier wave and the environmental noise continue to match, and effectively counteracting the signal detection means based on Doppler feature analysis.

[0027] In a specific embodiment of the present invention, the ambient electromagnetic background noise is collected in real time by a broadband receiver (frequency range: 20 MHz - 6 GHz, sampling rate: 50 MS / s) built into the dummy drone pilot. The energy ridge line features (resolution: 1 MHz × 10 ms) are extracted through short-time Fourier transform (window length: 256 points, overlap rate: 75%), and a dynamic fingerprint library containing 100 groups of typical spectrum templates (update period: 1 s) is formed. The camouflage noise floor is generated by a 16-channel parallel filter bank (bandwidth adjustable range: ±15%). The center frequency of each channel is aligned with the energy peak value of the fingerprint library (error: ±50 kHz). The encrypted pulse signal (symbol rate: 10 Mbps) and the noise floor are weighted and superimposed in the I / Q two channels through an orthogonal mixer (carrier phase perturbation range: ±π / 4). The hopping sequence template generates a pseudo-random sequence according to the zero-crossing interval of the power spectrum (statistical window: 200 ms) (hopping rate: 500 hops / s, frequency point interval: 2 MHz), and the phase-continuous carrier hopping is realized by a direct digital frequency synthesizer (phase noise: -110 dBc / Hz @ 1 kHz). The Doppler compensation module obtains the three-dimensional velocity (range: ±50 m / s) of the dummy drone pilot through an inertial measurement unit (update rate: 100 Hz), and uses a digital phase-locked loop (loop bandwidth: 10 kHz) to dynamically adjust the time slots of the hopping template (compensation accuracy: ±0.1 μs), and finally outputs a camouflage signal with a frequency offset error less than 50 Hz.

[0028] Overall, by deeply integrating the encrypted pulse signal with the dynamic ambient electromagnetic noise characteristics and adopting the adaptive frequency hopping transmission technology, the control signal fully matches the characteristics of the surrounding electromagnetic environment in the time-frequency domain. The generated pseudo-carrier waveform has the same spectrum characteristics and dynamic hopping rules as the ambient noise, effectively avoiding the signal separation and recognition of conventional spectrum detection equipment; at the same time, through real-time Doppler frequency shift compensation, the spectrum consistency of signal transmission is ensured, and the probability of being detected by the electronic reconnaissance system is reduced.

[0029] Preferably, the dynamic key is synchronously analyzed at the dummy drone pilot end and the spectrum camouflage signal is stripped to generate a virtual control trajectory matching the characteristics of the real operation behavior, specifically: Perform dynamic phase correction on the signal received at the dummy drone pilot end, generate a phase synchronization calibration factor based on the time slot compression parameters of the pseudo-random frequency hopping sequence template, and output a baseband signal with time-frequency alignment after eliminating the transmission path distortion; Combined with the energy ridge line distribution characteristics in the dynamic camouflage spectrum fingerprint library, construct a time-varying notch filter bank matching the camouflage noise floor, and strip the superimposed spectrum camouflage components in the baseband signal frame by frame through the ridge line matching degree threshold to generate a spectrum purification signal; Extract the key fingerprint identification of the cross-bit interleaved coding structure from the spectrum purification signal, combine it with the preset SparkLink protocol to reverse calculate the bit interleaving pattern, and separate the dynamic key seed and the chaotic residual vector of the segmented time-varying encryption key; Among them, the preset SparkLink protocol refers to a standardized data transmission specification for secure communication between a dummy pilot and a drone. Its core function is to ensure the reliable transmission and accurate parsing of encrypted control instructions. The protocol realizes the efficient organization of signals through a predefined frame structure (including a synchronization header, a key fingerprint area, an encrypted data payload, and a check field). The key fingerprint identification is encoded using the unique MAC address of the device. The cross-bit interleaved coding structure realizes the dispersion embedding and recombination of key data through a 16-stage deep shift register. The protocol stipulates strict timing relationships (time slot accuracy of 1 μs) and modulation methods (GMSK, BT = 0.3), enabling the receiving end to reverse calculate the bit interleaving pattern based on the known protocol structure, accurately separate the dynamic key components, and at the same time maintain hardware compatibility with commercial SparkLink devices.

[0030] Input the chaotic residual vector into the chaotic inverse diffusion model constrained by the real-time flight attitude parameters of the drone to reconstruct the spatio-temporal correlation topological structure of the original control instruction feature matrix; Among them, the chaotic inverse diffusion model is the decryption core module based on the dynamic characteristics of the Lorenz attractor. Its function is to reverse deduce the chaotic encryption process through the real-time attitude parameters of the drone and accurately reconstruct the original control instructions. The model uses the received chaotic residual vector as the initial condition of the system, combines the attitude data (sampling rate of 100 Hz) such as the pitch angle and roll angle transmitted back by the drone as parameter constraints, reversely solves the chaotic differential equations, and restores the instruction feature matrix before encryption through three iterative operations. Taking the motion state of the aircraft as the dynamic parameter boundary of the chaotic system makes the key reconstruction process have spatio-temporal uniqueness: only when the receiving party obtains the drone attitude parameters that are exactly the same as those at the encryption moment can the original instructions be accurately restored, thus realizing a dynamic security mechanism of one state, one key.

[0031] Perform biometric behavior simulation on the original control instruction feature matrix based on the biometric coupling weights of the electromyogram signal and the eye movement tracking vector to generate virtual trajectory primitives containing microscopic operation jitters; Perform inverse compensation of the drone dynamics on the virtual trajectory primitives through a time-frequency domain sliding window fitter, and dynamically adjust the trajectory curvature and phase continuity constraint boundaries according to the attitude parameters to generate a virtual control trajectory.

[0032] In specific implementation, the time-frequency domain sliding window fitter adopts a sliding window structure with a dual buffer (window width: 200 ms, sliding step: 10 ms), and receives virtual trajectory primitives (update rate: 100 Hz) and UAV attitude parameters (such as pitch angle ±90°, roll angle ±180°) in real time through an embedded processor (Cortex-M7 core). In the frequency domain compensation link, a 32nd-order FIR filter (cutoff frequency: 15 Hz) is used to perform phase linearization processing on the high-frequency flutter components of the trajectory (group delay compensation ±5 ms); in the time domain adjustment link, based on the current attitude angular velocity of the UAV (range ±300° / s), the trajectory curvature constraint boundary (lower limit of curvature radius: 2 m) is dynamically calculated through the cubic spline interpolation algorithm, and model predictive control (prediction time domain: 80 ms) is used to generate a smooth trajectory that meets the maximum roll rate (100° / s) and maximum overload (5g) limits of the UAV. The finally output virtual control trajectory maintains C2 continuity (curvature change rate < 50° / m²), and the correlation coefficient with the real pilot operation characteristics is > 0.85, and is sent to the flight control system for execution through the CAN bus (transmission rate: 1 Mbps).

[0033] In a specific implementation of the present invention, the dynamic phase correction module uses a digital phase-locked loop (loop bandwidth: 20 kHz) to perform carrier synchronization on the received signal, generates a phase compensation factor (adjustment range ±15°) according to the time slot parameters (time resolution: 1 μs) of the frequency hopping sequence template, and outputs a baseband signal through an orthogonal down-converter (I / Q path amplitude imbalance ≤ 0.5 dB). The time-varying notch filter bank consists of 32 adaptive FIR filters (order: 128, coefficient update rate: 1 kHz), dynamically adjusts the notch center frequency (step accuracy: 10 kHz) according to the ridge line characteristics of the spectrum fingerprint library (matching threshold: 0.85), and suppresses the camouflage noise by more than 40 dB. The key extraction module identifies the fingerprint identifier through a correlation detector (correlation window: 64 bit), uses a shift register array (depth: 16 levels) to inversely calculate the bit interleaving pattern, and outputs a 256-bit chaotic residual vector (bit error rate < 10 -6 . The chaotic inverse diffusion module uses the attitude data (update rate: 100 Hz) transmitted back by the UAV to constrain the Lorenz attractor parameters and reconstructs a 12-dimensional instruction feature matrix. The biological behavior simulation module simulates the electromyogram-eyemovement coupling characteristics (jitter frequency: 5 - 15 Hz) through a pre-trained LSTM network (hidden layer: 128 nodes) and generates trajectory primitives with artificial operation flutter. The dynamics compensation module uses model predictive control (prediction time domain: 50 ms) to perform real-time smoothing processing on the trajectory and outputs a virtual control trajectory (curvature continuous order C²) according to the UAV inertial navigation data (delay compensation: 10 ms).

[0034] Overall, this method effectively eliminates signal distortion and camouflage noise during the transmission process, achieves precise extraction of encrypted signals, and restores virtual trajectories with real operation behavior characteristics, including natural operation jitter and dynamic compensation characteristics. The finally generated virtual control trajectory not only maintains biometric consistency with the real person's operation but also conforms to the actual flight constraints of the UAV, making it difficult for the enemy to identify true and false pilots through trajectory analysis.

[0035] Preferably, based on the virtual control trajectory, the actual response signal of the UAV is reversely modulated to enable the UAV to perform actions and form dynamic error compensation with the virtual control trajectory. Specifically: Extract the virtual trajectory primitives of the virtual control trajectory, and construct a time-delay coupling matrix between the trajectory primitives and the current pose of the UAV according to the real-time flight attitude parameters of the UAV and in combination with the virtual trajectory primitives; Use the time-delay coupling matrix to perform spatio-temporal interpolation on the phase jump characteristics of the actual response signal, introduce an inertial compensation amount based on the dynamic weight of the trajectory curvature in the attitude control loop, and generate an attitude correction gradient containing non-linear perturbations; Among them, the actual response signal includes various signals reflecting the actual action execution of the UAV, such as attitude change signals (such as changes in pitch angle, roll angle, and yaw angle), motion speed, and acceleration change signals generated by the UAV after receiving control commands.

[0036] Perform backpropagation iteration on the attitude correction gradient according to the Jacobian matrix of the UAV dynamics model, and combine it with the angular velocity deviation vector output by the aircraft inertial measurement unit to determine the frequency-domain energy entropy of the trajectory tracking error; Among them, the aircraft inertial measurement unit continuously outputs an angular velocity deviation vector, which reflects the difference between the actual angular velocity of the UAV and the expected angular velocity of the virtual control trajectory. Combine this angular velocity deviation vector with the attitude correction gradient obtained by backpropagation iteration based on the Jacobian matrix of the UAV dynamics model. Through mathematical operations on the two, convert them to the frequency domain. In the frequency domain, perform mean statistical analysis on the frequency-domain energy distribution, and then determine the frequency-domain energy entropy of the trajectory tracking error.

[0037] It should be noted that the UAV dynamics model is a mathematical description constructed based on Newton's second law and the principles of rigid body kinematics, used to describe the motion state of the UAV under the action of various forces and torques. It comprehensively considers factors such as the mass, moment of inertia, aerodynamic force, and propeller thrust of the UAV, and describes the relationship between its position, velocity, acceleration, and attitude angle changing with time through a series of differential equations. This model is presented in matrix form, such as the state-space equation, where the state vector contains information such as position, velocity, and attitude angle, and the input vector covers control inputs (such as throttle, rudder deflection).

[0038] When the frequency-domain energy entropy exceeds the preset stealth control threshold, perform double-phase-amplitude constraints on the attitude correction gradient, and generate a compensation amount chaos injection sequence based on the chaotic residual vector of the dynamic key seed; Perform time-frequency domain convolution fusion on the compensation amount chaos injection sequence and the virtual trajectory primitive to obtain the trajectory-noise coupling entropy value, and adjust the inertial weight coefficient of the convolution kernel based on the trajectory-noise coupling entropy value to generate an inverse modulation signal with environmental adaptability; Use the phase synchronization calibration factor to compensate the Doppler frequency shift of the inverse modulation signal, so that it maintains time slot synchronization with the pseudo-random frequency hopping sequence template of the encrypted pulse signal, and output the actual response control instruction distributed in the same frequency band as the environmental background noise.

[0039] In a specific implementation of the present invention, the virtual trajectory primitive is updated once every 100 ms, and the position, speed, and acceleration information it contains and the real-time attitude parameters of the UAV (pitch angle, roll angle, yaw angle, update rate 100 Hz) jointly construct a time-delay coupling matrix. Use this matrix to interpolate the phase jump characteristics of the actual response signal in the spatio-temporal domain (time resolution 1 ms, space resolution 0.1°), introduce an inertial compensation amount based on the dynamic weight of the trajectory curvature (range 0-0.5 / m), and generate an attitude correction gradient. Through the backpropagation iteration of the Jacobian matrix, combine the angular velocity deviation vector to determine the frequency-domain energy entropy. When the entropy exceeds the threshold, perform double-phase-amplitude constraints on the attitude correction gradient, generate a compensation amount chaos injection sequence based on the dynamic key seed, perform convolution fusion with the virtual trajectory primitive, adjust the weight coefficient of the convolution kernel to generate an inverse modulation signal, and output the actual response control instruction in the same frequency band as the environmental background noise after compensating the Doppler frequency shift through the phase synchronization calibration factor.

[0040] Overall, this method constructs a time-delay coupling matrix and introduces non-linear perturbation compensation, so that the UAV's execution actions and the virtual trajectory maintain a dynamic error balance, and generate an inverse modulation signal with noise characteristics. The finally output control instruction is perfectly fused with the spectral characteristics of the background noise, which not only ensures that the UAV accurately executes the real instruction, but also makes it impossible for external observers to identify the real control intention through trajectory analysis, effectively realizing the dual protection effects of stealth control and electronic deception.

[0041] The dummy pilot control method further includes the following steps: Extract the phase entropy change index and spectral distortion residual of the encrypted pulse signal at the receiving end of the StarFlash feedback link, generate a signal integrity check code based on the chaotic residual vector of the dynamic key seed, and construct a time-frequency domain distortion feature matrix; According to the time-frequency ridge line distribution of the distortion feature matrix, determine the spectral confusion degree of the environmental noise and the camouflage signal in the transmission path; It should be noted that, first, analyze the distribution of time-frequency ridge lines in the distortion feature matrix. These ridge lines represent the energy concentration regions of the signal in the time-frequency domain. Then, compare these ridge lines with the spectrum of the environmental noise in the transmission path and calculate the degree of overlap between the two, that is, the spectrum confusion degree. This confusion degree reflects the similarity between the environmental noise and the camouflaged signal in the spectrum. The higher the confusion degree, the more difficult it is to distinguish the camouflaged signal from the environmental noise. In this way, the security and concealment of the signal in the transmission path can be evaluated, and corresponding measures can be taken to optimize the signal transmission to ensure that the signal is not easily detected and interfered by the enemy during the transmission process.

[0042] Based on the gradient descent direction of the spectrum confusion degree, use the orthogonal subspace projection theorem to generate a time-frequency confusion factor, and inject the time-frequency confusion factor into the time slot allocator of the pseudo-random frequency hopping sequence template to generate an adaptive frequency hopping parameter with the characteristic of phase mutation concealment; For example, when a drone formation deployed in a complex electromagnetic environment performs a reconnaissance mission, the spectrum confusion degree of the encrypted pulse signal extracted by the receiving end is 0.82 (calculation window 200 ms). Based on the gradient descent direction of this confusion degree, use the orthogonal subspace projection theorem to generate a set of 16-dimensional time-frequency confusion factors, and the values of each dimension are [0.05, -0.12, 0.08,..., 0.03]. Inject these confusion factors into the time slot allocator of the pseudo-random frequency hopping sequence template, and the time slot length is dynamically adjusted to 1.2 ms (the original time slot length is 1 ms) to generate a new frequency hopping parameter sequence. The phase jump amplitude of the new sequence is ±π / 6 (originally ±π / 8), and the frequency hopping rate is increased to 600 hops / s (originally 500 hops / s), and the phase synchronization mechanism of the pseudo-random frequency hopping sequence template is used to ensure consistency with the camouflage noise base. The finally generated adaptive frequency hopping parameter exhibits obvious phase mutation concealment characteristics in the spectrum characteristics and successfully avoids the enemy's spectrum analysis based on the continuous signal characteristics.

[0043] Synchronously extract the Doppler frequency shift covariance matrix in the real-time flight attitude parameters of the drone, and dynamically weight the time-frequency ridge line energy of the camouflage noise base in combination with the chaos mapping residual to generate a camouflaged signal strength adjustment coefficient that is phase-synchronized with the frequency hopping parameter; Input the strength adjustment coefficient into the time-varying notch filter bank, and feedback and close-loop control the spectrum overlap degree between the camouflage noise base and the background noise through the ridge line matching degree. When the spectrum overlap degree is not within the dynamic concealment interval, update the camouflage energy weight of the spectrum fingerprint library based on the inverse operation of the Jacobian matrix; Use the phase entropy change index to perform secondary encryption on the reconstructed frequency hopping parameter and the camouflaged signal strength according to the star flash protocol to generate an environment-adaptive concealment instruction set containing a synchronization calibration factor.

[0044] For example, the Doppler frequency shift covariance matrix in the attitude parameters of the UAV is extracted in real time, and its element values are [[0.02, -0.01], [-0.01, 0.03]] (unit: Hz / s). Combining the chaotic mapping residuals [0.08, -0.12, 0.05,...], the time-frequency ridge energy of the camouflage noise base is dynamically weighted to generate the camouflage signal strength adjustment coefficient [0.95, 0.85, 1.02,...]. This coefficient is input into the time-varying notch filter bank (filter order 8), and through the feedback closed-loop control of the ridge matching degree, the spectral overlap degree between the camouflage noise base and the background noise reaches 0.75 (the dynamic concealment interval is [0.7, 0.8]). When the spectral overlap degree exceeds the interval, the camouflage energy weight of the spectral fingerprint library is updated based on the inverse operation of the Jacobi matrix, and the energy weight of some frequency bands is adjusted from 0.8 to 1.1. Finally, the phase entropy change index is used to perform secondary encryption on the reconstructed frequency hopping parameters and camouflage signal strength according to the star flash protocol, generating an environment-adaptive concealment instruction set containing a synchronization calibration factor to ensure the concealment and robustness of the instructions in a complex electromagnetic environment.

[0045] Overall, this method generates a check code and a distortion feature matrix by extracting the phase entropy change index, etc., realizes the full-dimensional optimization of the transmitted signal, and finally generates an environment-adaptive concealment instruction set, which can effectively improve the concealment and anti-interference ability of the signal in a complex environment and further ensure communication security.

[0046] In this embodiment, the dummy pilot control method may further include the following steps: Obtain the blood oxygen fluctuation phase vector of the real pilot and the gradient distribution tensor of the body surface temperature field, and construct a biometric fusion matrix with spatio-temporal correlation; Based on the singular value decay rate of the biometric fusion matrix, use the tensor orthogonal decomposition algorithm to extract the blood oxygen-temperature coupled oscillation mode and generate a dynamic entropy weight vector containing the chaotic characteristics of the biological rhythm; Input the dynamic entropy weight vector into a preset entropy value perturbation channel, and perform non-linear phase modulation on the chaotic mapping parameters of the dynamic key seed through the entropy weight fusion operator to generate a key chaotic residual with biometric binding characteristics; When the key transmission is interrupted, intercept the time-frequency distortion residual segment of the encrypted pulse signal, combine the key chaotic residual with the spatio-temporal distortion covariance matrix of the biometric fusion matrix, and generate multi-dimensional compensation parameters using the phase synchronization calibration factor; Among them, the multi-dimensional compensation parameters include an amplitude compensation factor, a time delay calibration coefficient, a frequency offset correction amount, and a phase synchronization offset amount, which are used to jointly correct the time-frequency domain distortion generated by the key chaotic residual after the transmission interruption; Based on the multi-dimensional compensation parameters, perform reverse diffusion reconstruction on the chaotic residuals, and use the entropy value iterator constrained by the chaotic characteristics of biological rhythms to restore the phase synchronization reconstruction parameters of the original dynamic key seed, and output a reconstructed encrypted pulse signal that is synchronized with the environment adaptive frequency hopping pattern.

[0047] Among them, the preset entropy value perturbation channel refers to a non-linear modulation circuit or module based on the dynamic entropy weight vector of biological characteristics. By coupling the chaotic characteristics of biological rhythms with the chaotic parameters of the key in the phase domain, it realizes the biometric binding encryption perturbation of the dynamic key seed.

[0048] For example, through wearable biosensors (such as near-infrared blood oxygen meters and distributed temperature sensor arrays), the blood oxygen saturation fluctuation signal (sampling rate 100Hz) at the tiger's mouth of the left hand of the drone pilot and the 5×5 temperature gradient distribution data (accuracy 0.1°C) on the inner side of the forearm are collected in real time. Align the phase offset of the blood oxygen signal with the temperature gradient tensor according to a 10ms time window to construct a 16-dimensional biometric fusion matrix. Using the singular value decay curve of this matrix (such as the time delay of 20ms when the main component decays to 0.5 times the threshold), extract the coupled oscillation components (frequency band 0.1 - 0.3Hz) of blood oxygen fluctuation and temperature change, and generate a dynamic entropy weight vector through non-linear weighting (amplitude range ±0.7V). Input this vector into the preset entropy value perturbation module (bandwidth 50kHz), perform phase interleaving modulation with the dynamic key seed, and output the key chaotic residuals carrying biological characteristics. When it is detected that the encrypted pulse signal is interrupted (such as when the packet loss rate > 15%), intercept the time-frequency distortion section (frequency offset ±2kHz) of the last 200ms signal, combine the covariance of the key chaotic residuals and the biological matrix (correlation coefficient above 0.85), generate multi-dimensional parameters including amplitude compensation (±3dB) and time delay calibration (±0.5ms), and finally reconstruct an encrypted pulse signal that is synchronized with the frequency hopping sequence (frequency hopping interval 50ms) (center frequency 2.4GHz ± 100kHz).

[0049] In summary, considering that the traditional UAV key transmission system is difficult to quickly self-recover when the signal is interrupted, and the key generation process lacks biometric dynamic binding, resulting in insufficient security and weak anti-interference ability, in this embodiment, by fusing the blood oxygen and body temperature biometric characteristics of the drone pilot in real time to generate a dynamic entropy weight vector, the key chaotic residuals have the uniqueness of biological rhythms, enhancing the anti-cracking ability; and when the transmission is interrupted, using the spatio-temporal correlation between the biometric fusion matrix and the key chaotic residuals, quickly reconstruct the synchronous key to ensure that the encrypted pulse signal can still adaptively recover in a complex electromagnetic environment, improving the robustness and continuity of covert communication.

[0050] In this embodiment, the dummy drone pilot control method may further include the following steps: Establish a quantum entanglement channel among distributed dummy flying hand nodes to generate a quantum entanglement state feature matrix carrying node positions and environmental noise characteristics; Each node collects local electromagnetic environment spectrum observation data in real time, and performs cross-node correlation compression coding in combination with the quantum entanglement state feature matrix to generate a group of collaborative observation vectors; Input the group of collaborative observation vectors into a Bayesian game equalizer, and generate a set of differential camouflage spectrum fingerprint primitives with asymmetric correlation characteristics through strategy space mapping; Based on the frequency-domain mutual information of the fingerprint primitive set, construct a phase entanglement mapping relationship among nodes, and extract the spatio-temporal-frequency three-dimensional correlation weight distribution tensor of the group signal; Perform phase-domain entanglement superposition on the camouflage spectrum fingerprint primitives according to the weight distribution tensor to generate a group camouflage signal synthesis instruction carrying collaborative optimization parameters; Use the quantum teleportation feedback link to distribute the synthesis instruction to each node, and drive the camouflage signal transmitter to output a collaborative interference waveform that maintains spectral entanglement consistency with the environmental background noise.

[0051] Among them, the Bayesian game equalizer in this embodiment refers to a distributed decision-making module based on the game theory of incomplete information. By analyzing the statistical characteristics of the collaborative observation vector group of each node and the quantum entanglement correlation degree, it constructs a strategy space including prior probabilities and conditional payoff functions, enabling each dummy node to autonomously generate a camouflage spectrum decision with optimal asymmetric correlation characteristics even when the complete strategies of other nodes are unknown. This equalizer uses a posterior probability dynamic update mechanism to quantify the entanglement state characteristics transmitted by the quantum channel and the uncertainty of environmental observation data as a game payoff matrix, and finally outputs a set of differential spectrum primitives that make the group camouflage signal reach Nash equilibrium in the spatio-temporal-frequency dimension.

[0052] For example, deploy a polarization-coded quantum key distribution system (operating wavelength 1550 nm) among three distributed dummy drone nodes (spacing 50 - 100 meters). Generate a quantum entanglement state feature matrix (dimension 4×4) containing the node GPS coordinates (accuracy ±1 m) and the environmental noise spectrum characteristics (sampling bandwidth 20 MHz) through the quantum state preparation module. Each node uses software-defined radio (SDR, sampling rate 100 MS / s) to collect the electromagnetic environment IQ data in the 2.4 GHz band in real time, and combines it with the quantum entanglement state matrix for singular value threshold compression (retaining the first 3 principal components) to generate a set of collaborative observation vectors containing time delay differences and frequency offsets (length 128 dimensions). Input this vector set into a preset Bayesian strategy engine (strategy space dimension 8), and output a set of camouflage spectrum fingerprint primitives with a correlation probability of 0.7 - 0.9 (including 3 different hopping frequency templates). Based on the mutual information of the primitive set (mutual information value ≥ 0.6 bit), construct a phase entanglement relationship matrix between nodes (complex weight, modulus value 0.5 - 1.2), and then generate a three-dimensional space-time-frequency weight tensor (size 3×3×64). Use this tensor to perform QPSK phase modulation on the primitives (symbol rate 10 kSymbol / s) to generate a synthetic command containing collaborative time slot allocation (time slot width 20 ms) and power control (dynamic range ±3 dB), and finally drive the transmitters of each node to output a collaborative interference waveform with a center frequency of 2.41 GHz ± 50 kHz and a bandwidth of 5 MHz through a quantum teleportation link (transmission rate 1 Mbps), and the correlation coefficient between its power spectral envelope and the measured environmental noise is above 0.85.

[0053] It should be noted that to solve the problems of poor coordination of camouflage signals in the multi-dummy drone system and easy recognition of spectrum characteristics, and to overcome the defect that it is difficult to achieve high-synchronization spectrum camouflage between distributed nodes, this embodiment realizes the real-time sharing of environmental noise characteristics between nodes through a quantum entanglement channel, combines Bayesian game optimization to generate differentiated camouflage spectra, enables the interference signals of multiple dummy nodes to maintain dynamic coordination in the three dimensions of space-time-frequency, forms a group camouflage effect highly integrated with the real environmental noise, and improves the anti-detection ability of the system.

[0054] As Figure 3 shown, the second aspect of the present invention discloses a low-power StarFlash control module 8, which is applied to any one of the described dummy drone control methods, and includes: A StarFlash control unit 1, integrated with a Hi3861V100 main control chip, supporting dual-mode communication of the StarFlash protocol stack and BLE5.3; A power management unit 2, supporting dynamic voltage regulation from 0.8 - 3.3 V, with a sleep power consumption ≤ 5 μA; A StarFlash instruction generation unit 3, used to encrypt the original control instruction and generate an encrypted pulse signal containing a dynamic key; The adaptive frequency hopping transmission unit 4 integrates a narrowband radio frequency chip and a spectrum camouflage engine, enabling the encrypted pulse signal to pseudo-randomly hop within a preset frequency band; The dummy pilot unit 5 adopts a bionic thermal radiation structure design, and a multi-region independent temperature control unit is arranged inside the bionic shell. Each unit is divided into a head high-temperature region, a torso constant-temperature region, and a limb gradient temperature region according to human anatomical characteristics; The StarFlash feedback verification unit 6 monitors the signal transmission quality in real time through a closed-loop link and triggers dynamic parameter adjustment.

[0055] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should be covered within the protection scope of the present invention.

Claims

1. A dummy pilot control method for a drone, characterized in that: The following steps are involved: Obtain the original control instructions of the real pilot, and convert the original control instructions into an encrypted pulse signal containing a dynamic key; A spectrum camouflage signal simulating the dummy pilot's environmental background is integrated into the encrypted pulse signal and then transmitted to the dummy pilot in an adaptive frequency hopping mode; The dynamic key is synchronously parsed on the dummy pilot side and the spectrum camouflage signal is stripped to generate a virtual control trajectory that matches the real operation behavior characteristics; The actual response signal of the UAV is reversely modulated based on the virtual control trajectory, so that the UAV performs an action and forms a dynamic error compensation with the virtual control trajectory.

2. A method for controlling a dummy pilot applied to a UAV according to claim 1, characterized in that: Get the original control instructions of the real pilot, and convert the original control instructions into an encrypted pulse signal containing a dynamic key, specifically: The wearable sensor array captures the hand electromyography signals, nerve pulse frequency and eye tracking vector of the real pilot in real time, forming a multi-dimensional biological control feature flow; De-noising the multi-dimensional biological manipulation feature stream by using a normalized convolution kernel, and aligning the multi-dimensional biological manipulation feature stream by time stamp segments to generate a manipulation instruction feature matrix with spatiotemporal correlation; Based on the entropy value distribution of the control instruction feature matrix, a chaotic mapping algorithm is used to generate a dynamic key seed bound to the current control behavior, and the key chaotic diffusion is performed in combination with the real-time flight attitude parameters of the UAV to output a segmented time-varying encryption key; Input the dynamic key seed and the segmented time-varying encryption key into the Star Flash protocol converter, perform key cross-bit interleaving encoding according to the preset protocol structure, and generate a pulse baseband sequence containing a key fingerprint identifier; A quantum random number generator is used to generate a non-periodic noise mask factor, and the pulse baseband sequence is subjected to phase-amplitude dual perturbations to form an encrypted pulse signal with environmental noise-like statistical characteristics.

3. The method for controlling a dummy pilot applied to a UAV according to claim 1, characterized in that: The encrypted pulse signal is integrated with a spectrum camouflage signal simulating the background of the dummy pilot's environment and then transmitted to the dummy pilot in an adaptive frequency hopping mode, specifically: Obtain the environmental electromagnetic background noise within the preset range of the dummy pilot, extract the energy ridge distribution characteristics in the signal time-frequency domain, and build a dynamic camouflage spectrum fingerprint library; Based on the statistical characteristics of the dynamic camouflage spectrum fingerprint library, the orthogonal subspace projection theorem is used to generate a camouflage noise floor with the same time-frequency ridge characteristics, and the camouflage noise floor is superimposed with the encrypted pulse signal through multi-dimensional phase synchronization disturbance. According to the instantaneous bandwidth characteristics of the superimposed signal, the zero-crossing point distribution of its power spectrum density envelope is extracted to generate a pseudo-random frequency hopping sequence template that matches the time-frequency characteristics of the current environmental electromagnetic background noise; The camouflaged noise floor and the pseudo-random frequency hopping sequence template are input into the signal synthesizer to perform dynamic time slot allocation based on phase continuity, thereby reconstructing a camouflaged carrier waveform with environmental adaptability while retaining the phase jump characteristics of the original encrypted pulse signal; Acquire the real-time motion state of the dummy pilot, and determine the real-time Doppler frequency shift parameter according to the real-time motion state; The pseudo-random frequency hopping sequence template is time-slot compressed and compensated according to the real-time Doppler frequency shift parameter, so that the frequency deviation characteristics of the camouflaged carrier waveform keep the spectrum consistent with the dynamic changes of the background noise, and then transmitted to the dummy pilot.

4. The method for controlling a dummy pilot applied to a UAV according to claim 1, characterized in that: The dynamic key is parsed synchronously on the dummy pilot side and the spectrum camouflage signal is stripped to generate a virtual control trajectory that matches the real operation behavior characteristics, specifically: Dynamic phase correction is performed on the signal received by the dummy pilot, and a phase synchronization calibration factor is generated based on the time slot compression parameter of the pseudo-random frequency hopping sequence template. After eliminating the transmission path distortion, a time-frequency aligned baseband signal is output; Combined with the energy ridge distribution characteristics in the dynamic camouflage spectrum fingerprint library, a time-varying notch filter bank matching the camouflage noise floor is constructed. The spectrum camouflage component superimposed in the baseband signal is stripped frame by frame through the ridge matching threshold control to generate a spectrum purification signal. Extract the key fingerprint of the cross-bit interleaved coding structure from the spectrum purification signal, combine the preset star flash protocol to reversely solve the bit interleaving pattern, and separate the dynamic key seed and the chaotic residual vector of the segmented time-varying encryption key; The chaotic residual vector is input into the chaotic inverse diffusion model constrained by the real-time flight attitude parameter of the UAV to reconstruct the spatiotemporal correlation topological structure of the original control instruction feature matrix; Based on the biological feature coupling weights of electromyographic signals and eye tracking vectors, the original control instruction feature matrix is ​​simulated for biological behavior, generating virtual trajectory primitives containing micro-operation jitters. The virtual trajectory primitive is subjected to reverse compensation of UAV dynamics through a time-frequency domain sliding window fitter, and the trajectory curvature and phase continuity constraint boundary are dynamically adjusted according to attitude parameters to generate a virtual control trajectory.

5. The method for controlling a dummy pilot applied to a UAV according to claim 1, characterized in that: Based on the virtual control trajectory, the actual response signal of the UAV is reversely modulated so that the UAV performs an action and forms a dynamic error compensation with the virtual control trajectory, specifically: Extracting a virtual trajectory primitive of the virtual control trajectory, and constructing a time delay coupling matrix between the trajectory primitive and the current posture of the UAV according to the real-time flight attitude parameters of the UAV and in combination with the virtual trajectory primitive; The time-delay coupling matrix is ​​used to perform spatiotemporal interpolation on the phase jump characteristics of the actual response signal, an inertia compensation amount based on the dynamic weight of the trajectory curvature is introduced into the attitude control loop, and an attitude correction gradient containing nonlinear disturbance is generated; The attitude correction gradient is back-propagated iteratively according to the Jacobian matrix of the UAV dynamics model, and the frequency domain energy entropy of the trajectory tracking error is determined by combining the angular velocity deviation vector output by the aircraft inertial measurement unit. When the frequency domain energy entropy exceeds the preset hidden control threshold, the attitude correction gradient is subjected to phase-amplitude dual constraints, and a compensation chaotic injection sequence is generated based on the chaotic residual vector of the dynamic key seed. The compensation chaotic injection sequence is convolved with the virtual trajectory primitive in the time-frequency domain to obtain a trajectory-noise coupling entropy value, and the inertia weight coefficient of the convolution kernel is adjusted based on the trajectory-noise coupling entropy value to generate a reverse modulation signal with environmental self-adaptation; The phase synchronization calibration factor is used to compensate the Doppler frequency shift of the reverse modulated signal so that it can maintain time slot synchronization with the pseudo-random frequency hopping sequence template of the encrypted pulse signal and output actual response control instructions distributed in the same frequency band as the environmental background noise.

6. The method for controlling a dummy pilot applied to a UAV according to claim 1, characterized in that: The dummy flying hand control method also includes the following steps: At the receiving end of the star flash feedback link, the phase entropy change index and spectrum distortion residual of the encrypted pulse signal are extracted, the signal integrity check code is generated based on the chaotic residual vector of the dynamic key seed, and the time-frequency domain distortion feature matrix is ​​constructed. Determining the frequency spectrum confusion degree of the ambient noise and the camouflage signal in the transmission path according to the time-frequency ridge distribution of the distortion feature matrix; Based on the gradient descent direction of the spectrum confusion degree, the orthogonal subspace projection theorem is used to generate a time-frequency confusion factor, and the time-frequency confusion factor is injected into the time slot allocator of the pseudo-random frequency hopping sequence template to generate an adaptive frequency hopping parameter with a phase mutation concealment characteristic; The Doppler frequency shift covariance matrix in the real-time flight attitude parameters of the UAV is synchronously extracted, and the time-frequency ridge energy of the camouflage noise floor is dynamically weighted in combination with the chaotic mapping residual to generate a camouflage signal strength adjustment coefficient that is phase-synchronized with the frequency hopping parameters. The intensity adjustment coefficient is input into a time-varying notch filter bank, and the spectrum overlap between the camouflage noise floor and the background noise is controlled through a ridge matching feedback closed loop. When the spectrum overlap is not in a dynamic concealment interval, the camouflage energy weight of the spectrum fingerprint library is updated based on the inverse operation of the Jacobian matrix; The reconstructed frequency hopping parameters and camouflage signal strength are re-encrypted by the Star Flash protocol using the phase entropy variation index to generate an environment-adaptive covert instruction set containing a synchronization calibration factor.

7. A low-power star flash control module, applied to the dummy pilot control method according to any one of claims 1 to 6, characterized in that: include: The Star Flash control unit integrates the Hi3861V100 main control chip and supports the Star Flash protocol stack and BLE5.3 dual-mode communication; Power management unit, supports 0.8-3.3V dynamic voltage regulation, sleep power consumption ≤ 5μA; The star flash command generation unit is used to encrypt the original control command and generate an encrypted pulse signal containing a dynamic key; Adaptive frequency hopping transmission unit, integrating narrowband RF chip and spectrum camouflage engine, makes the encrypted pulse signal jump pseudo-randomly within the preset frequency band; A dummy pilot unit, wherein the dummy pilot unit adopts a bionic thermal radiation structure design, and a multi-zone independent temperature control unit is arranged inside the bionic shell. Each unit is divided into a head high temperature zone, a torso constant temperature zone, and a limb gradient temperature zone according to the anatomical characteristics of the human body; The Star Flash feedback verification unit monitors the signal transmission quality in real time through a closed-loop link and triggers dynamic parameter adjustment.

Images