Detection, avoidance and traceability method for unmanned aerial vehicle navigation spoofing attack

Abstract

The invention discloses an unmanned aerial vehicle navigation spoofing attack-oriented detection, avoidance and traceability method, which comprises the following steps: S10, establishing a satellite navigation spoofing attack model which comprises a model of influence of signals on noise intensity and a model of attenuation of signal intensity along with distance; s20, establishing a navigation spoofing attack detection mechanism according to the influence model of the signal on the noise intensity and the model that the signal intensity is attenuated along with the distance, including comparison based on satellite navigation result data and detection based on the signal noise intensity, so as to realize the detection of the navigation spoofing attack; s30, according to the model that the signal intensity is attenuated along with the distance, attack source avoidance of signal intensity gradient decline is carried out; and S40, on the basis of avoiding the attack source, realizing positioning and tracing of the attack source. According to the navigation signal features and information content, detection, avoidance and traceability of the navigation spoofing attack of the unmanned aerial vehicle are realized, and the threat of the current navigation spoofing attack can be effectively dealt with.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of unmanned aerial vehicle navigation signal processing, and particularly relates to a detection, avoidance and tracing method for unmanned aerial vehicle navigation deception attack. BACKGROUND

[0002] As a key equipment in modern military operations and low-altitude economic development, unmanned aerial vehicles have shown great application potential in fields such as battlefield reconnaissance, precision strike, emergency rescue, line inspection, and logistics distribution. Their efficient, flexible, and low-cost characteristics make them an important technical support for multiple industries. However, as the application scenarios of unmanned aerial vehicles continue to expand, the information security problems they face have become increasingly prominent, especially in terms of navigation and positioning.

[0003] Currently, most civilian and some military unmanned aerial vehicles rely on global satellite navigation systems for positioning and navigation. Such systems provide position and time information to receiving devices through wireless signals, with the advantages of wide coverage and high precision. However, due to the openness and vulnerability of navigation signals during transmission, they are extremely susceptible to malicious interference and deception attacks.

[0004] Navigation deception attack is a typical information attack method. Attackers emit radio signals that are consistent in format with real navigation signals but have false content, inducing the navigation receiver of the unmanned aerial vehicle to capture and track these deception signals, thereby causing it to calculate incorrect position, speed, or time information. At the same time, attackers may also use signal suppression or shielding techniques to block the unmanned aerial vehicle from receiving real navigation signals, further enhancing the deception effect. Such attacks have the following notable characteristics: strong concealment: the deception signal is highly similar to the real signal in terms of code type, frequency, and modulation method, and conventional receiving equipment has difficulty in identifying its authenticity within a short period of time; great threat: once the deception is successful, the attacker can control the flight path of the unmanned aerial vehicle without arousing its alarm, and even guide it into a dangerous area or cause it to fail its mission; non-physical damage: the attack process does not rely on firepower strikes or physical destruction, and is difficult to trace. How to detect and avoid navigation deception attacks has become a pressing problem.

[0005] Currently, the detection and protection technology against unmanned aerial vehicle navigation deception attacks is not mature. Existing methods have certain limitations in terms of real-time performance, applicable environment, system complexity, and cost. SUMMARY

[0006] To address the aforementioned problems, this invention proposes a method for detecting, avoiding, and tracing the source of navigation deception attacks on unmanned aerial vehicles (UAVs). By detecting the characteristics of navigation signals and the content of navigation information, deception attacks are detected. By direction finding the attack source, deception attacks are avoided, and the attack source is located. A method for detecting, avoiding, and tracing the source of navigation deception attacks based on signal-noise strength is also proposed.

[0007] To achieve the above objectives, the technical solution adopted by this invention is: a method for detecting, avoiding, and tracing the source of navigation deception attacks on unmanned aerial vehicles (UAVs), comprising the following steps:

[0008] S10: Establish a satellite navigation spoofing attack model, including a model of the impact of signal on noise intensity and a model of signal strength attenuation with distance;

[0009] S20: Establish a navigation spoofing attack detection mechanism based on the signal-to-noise intensity influence model and the signal intensity-to-distance attenuation model, including data comparison based on satellite navigation results and detection based on signal-to-noise intensity, to detect navigation spoofing attacks;

[0010] S30: Based on the signal strength attenuation model with distance, perform attack source avoidance by reducing the signal strength gradient;

[0011] S40: Based on the avoidance of the attack source, the attack source is located and traced.

[0012] Furthermore, in step S10, establishing a model of the influence of the signal on noise intensity includes the following steps:

[0013] S101, In a real attack scenario, the attacker uses multiple signals to launch an attack. The pseudo-random codes of each signal will interfere with each other, resulting in N... Auth Road real signal and N Spoof The deceptive signal is defined as follows: the autocorrelation amplitude y of the l-th real signal during the K-th pseudo-random code period. l [K] is represented as:

[0014]

[0015] In the formula, F il [K] represents the cross-correlation interference between the i-th real signal and the l-th real signal, F kl [k] represents the cross-correlation interference between the k-th deceptive signal and the l-th real signal; and Let _i_th channel's true signal power and _k_th channel's spoofed signal power be respectively, and let exp(·) denote the exponential function. Let represent the carrier phase of the l-th signal, j represent the imaginary unit, used to describe the signal components with angular frequency and phase, and η[K] be the ambient noise that follows a normal distribution;

[0016] S102, a hypothetical pseudo-random code f, which is uncorrelated with all PRN codes (Pseudorandom Noise Codes) in the real system, is used to perform correlation operations with the signal to estimate the noise floor. Due to the hypothetical nature of f, its cross-correlation value with other PRN codes approaches zero, thus satisfying:

[0017]

[0018] Where, ω l This represents the carrier angular frequency of the l-th signal;

[0019] Therefore, the correlation results of noise basis estimation for:

[0020]

[0021] In the formula, F if [K] and F kf [K] represents the parameters introduced into the noise basis estimation of the i-th real signal and the k-th spoof signal, respectively, and var[·] represents the calculation of the variance of the random variable;

[0022] S103, the above parameter F if [K] is a complex signal containing branches I and Q, whose values ​​follow a two-dimensional normal distribution with a mean of 0, and is represented as:

[0023]

[0024] in, and F if The variances of the same-direction I branch and the orthogonal Q branch of [K], N c Represents a complex normal distribution;

[0025] The integral value of the environmental noise η[K] also follows a zero-mean Gaussian distribution:

[0026]

[0027] in, The power of the environmental noise η[K] is represented by N, and the period length of the pseudo-random code is represented by N.

[0028] After incorporating both white noise and cross-correlation interference, the noise floor estimate is as follows:

[0029]

[0030] Where N0 represents the ambient noise power, T s Indicates the sampling period. The noise basis is obtained from the noise basis estimation. It increases as the total power of the spoofing signal increases.

[0031] Furthermore, in step S10, a signal strength attenuation model with distance is established:

[0032]

[0033] Where E is the electromagnetic noise intensity at a distance r from the point source, and E0 is the initial electromagnetic noise intensity at a distance r0 from the point source;

[0034] The farther away from the point source, the greater the attenuation of electromagnetic noise intensity.

[0035] Furthermore, in step S20, a detection mechanism based on satellite navigation result data comparison is established to detect navigation spoofing attacks in terms of information content, including the following steps:

[0036] S211, in the drone position controller, set the speed limit v. max ;

[0037] S212, the UAV coordinates (X) are obtained from GNSS (Global Navigation Satellite System). U1 ,Y U1 Z U1 ), and the coordinates (X) of the previous moment. U0 ,Y U0 Z U0 ), and calculate the straight-line distance Δρ;

[0038] At the same time, based on the timestamps t of two adjacent moments U0 and t U1 Calculate the time difference Δt;

[0039] The velocity during that time interval Δt is:

[0040] S213, if the velocity v 01 >2v max If the signal is positive, it is considered to be a navigation spoofing signal; otherwise, the signal is considered reliable.

[0041] Furthermore, in step S20, a detection mechanism based on signal-noise intensity is established to detect navigation spoofing attacks on the physical channel, including the following steps:

[0042] S221, Obtain the carrier noise density ratio C / N0 from the received navigation message, where C represents the carrier power. Calculate the signal-to-noise ratio (SNR) based on the equivalent noise bandwidth B and the carrier noise density ratio C / N0:

[0043]

[0044] S222, obtain the signal power P from the power meter. S And the signal-to-noise ratio (SNR), calculate the noise power P. n :

[0045]

[0046] S223, for noise power P n A maximum threshold P is set based on the intensity distribution of signal noise when not subjected to navigation spoofing attacks. n,max If P n >P n,max If the signal is positive, it is considered to be a navigation spoofing signal; otherwise, the signal is considered reliable.

[0047] Furthermore, in step S30, a fast attack source avoidance method based on signal strength gradient descent is designed, including:

[0048] S301, According to the signal strength attenuation model with distance, the electromagnetic noise intensity is inversely proportional to the square of the distance to the point source. The farther away from the point source, the greater the attenuation of the electromagnetic noise intensity. A threshold E0 is set, and it is assumed that when the electromagnetic noise intensity drops below E0, the attack will not be effective. At this time, the distance between the drone and the point source is r0.

[0049] S302, Establish a repulsive field centered on the point source;

[0050] The repulsive field function is:

[0051]

[0052] Where U(r) represents the repulsive potential energy when the distance between the UAV and the point source is r;

[0053] S303, by the definition of an artificial potential field, experiences a force in a force field that is the negative gradient of the artificial potential field:

[0054]

[0055] F(r) represents the force experienced by the UAV in the artificial potential field. In the point source model, the direction of the force at any location is from the point source to that location, and this direction is obtained by the maximum signal method. The gradient operator represents the rate of change and direction of change of a scalar function.

[0056] The drone flies away from the attack source to reduce the impact of the deceptive attack signal until the measured maximum noise signal strength drops below the threshold E0.

[0057] Furthermore, a directional antenna is used to receive signals, and the direction in which the strongest signal is received is measured by horizontal rotation as the direction of the attack source.

[0058] Furthermore, in step S40, the location and tracing of the attack source includes the following steps: after completing the avoidance of the attack source, the direction of the interference source is determined by the maximum signal method, the direction of the interference source is selected to fly in a direction perpendicular to the interference source, and the direction of the interference source is measured again after flying a certain distance.

[0059] In a two-dimensional plane, the target interference source is located at (x, y), and the two measurement positions are (x1, y1) and (x2, y2). The arrival angles of the measured noise signals are θ1 and θ2. The estimated location of the interference source is calculated as follows:

[0060]

[0061] The beneficial effects of adopting this technical solution are:

[0062] This invention establishes a satellite navigation spoofing attack model and analyzes the relationship between signal noise intensity and spoofing signal intensity. It collects carrier noise density ratio and navigation positioning data, and calculates environmental noise intensity based on equivalent noise bandwidth and carrier noise density ratio. Based on environmental noise intensity and satellite navigation data, it establishes a detection mechanism against navigation spoofing attacks. It establishes a signal intensity attenuation model in space with distance and designs a direction-finding method for the attack source. Based on the direction-finding method, it designs a rapid attack source avoidance method based on signal intensity gradient and a method for measuring the physical location of the attack source. This method, designed based on electromagnetic signal characteristics and information content correlation, can achieve timely and effective detection of navigation spoofing attacks. Simultaneously, after detecting an attack, it can quickly implement targeted avoidance and reverse positioning, thus solving the limitations and passivity of existing navigation spoofing detection methods.

[0063] This invention proposes a detection, avoidance, and source tracing method for UAV navigation deception attacks, achieving effective detection from both information content and channel characteristics perspectives. By quantitatively analyzing the environmental noise changes caused by navigation deception attacks, a noise-change-based detection method is designed, achieving effective detection at the channel level and providing a priori warning capabilities. By analyzing the characteristics of navigation attack data, a navigation data comparison-based detection method is designed, achieving information content-level detection and providing a reactive remediation capability. Due to the directional and intensity-attenuating characteristics of signal propagation, a method for rapidly avoiding the attack source after detection is designed, and reverse location of the attack source is performed. Compared to traditional single detection methods, this scheme achieves more comprehensive navigation deception attack detection, rapidly reduces impact while maintaining performance, reverse-locates the target, and provides crucial information for subsequent responses. Attached Figure Description

[0064] Figure 1 This is a schematic diagram of a method for detecting, avoiding, and tracing the source of navigation deception attacks for unmanned aerial vehicles (UAVs) according to the present invention.

[0065] Figure 2 This is a schematic diagram illustrating the principle of attack source avoidance in an embodiment of the present invention;

[0066] Figure 3 This is a schematic diagram illustrating the principle of attack source location and tracing in an embodiment of the present invention. Detailed Implementation

[0067] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described below with reference to the accompanying drawings.

[0068] In this embodiment, see Figure 1 As shown, this invention proposes a method for detecting, avoiding, and tracing the source of navigation deception attacks on unmanned aerial vehicles (UAVs), including the following steps:

[0069] S10: Establish a satellite navigation spoofing attack model, including a model of the impact of signal on noise intensity and a model of signal strength attenuation with distance;

[0070] S20: Establish a navigation spoofing attack detection mechanism based on the signal-to-noise intensity influence model and the signal intensity-to-distance attenuation model, including data comparison based on satellite navigation results and detection based on signal-to-noise intensity, to detect navigation spoofing attacks;

[0071] S30: Based on the signal strength attenuation model with distance, perform attack source avoidance by reducing the signal strength gradient;

[0072] S40: Based on the avoidance of the attack source, the attack source is located and traced.

[0073] As an optimization of the above embodiment, step S10, establishing a model of the influence of signal on noise intensity, includes the following steps:

[0074] S101, In a real attack scenario, the attacker uses multiple signals to launch an attack. The pseudo-random codes of each signal will interfere with each other, resulting in N... Auth Road real signal and N Spoof The deceptive signal is defined as follows: the autocorrelation amplitude y of the l-th real signal during the K-th pseudo-random code period. l [K] is represented as:

[0075]

[0076] In the formula, F il [K] represents the cross-correlation interference between the i-th real signal and the 1-th real signal, F kl [K] represents the cross-correlation interference between the k-th deceptive signal and the l-th real signal; and Let _o_th channel's true signal power and _k_th channel's spoofed signal power be respectively, and let exp(·) denote the exponential function. Let represent the carrier phase of the l-th signal, j represent the imaginary unit, used to describe the signal components with angular frequency and phase, and η[K] be the ambient noise that follows a normal distribution;

[0077] Regarding F il [K], we have:

[0078]

[0079] Where, Δω ilk τ ilK and c represents the Doppler frequency shift difference, time delay, and carrier phase difference between the i-th local synchronization signal and the l-th satellite signal, respectively. i (n) represents the pseudo-random code sequence of the i-th signal at discrete time n, and N represents the period length of the pseudo-random code.

[0080] S102, a hypothetical pseudo-random code f, which is uncorrelated with all PRN codes (Pseudorandom Noise Codes) in the real system, is used to perform correlation operations with the signal to estimate the noise floor. Due to the hypothetical nature of f, its cross-correlation value with other PRN codes approaches zero, thus satisfying:

[0081]

[0082] Where, ω l This represents the carrier angular frequency of the l-th signal;

[0083] Therefore, the correlation results of noise basis estimation for:

[0084]

[0085] In the formula, F if [K] and F kf [K] represents the parameters introduced into the noise basis estimation of the o-th real signal and the k-th spoof signal, respectively, and var[·] represents the calculation of the variance of the random variable;

[0086] S103, the above parameter F if [K] is a complex signal containing branches I and Q, whose values ​​follow a two-dimensional normal distribution with a mean of 0, and is represented as:

[0087]

[0088] in, and F if The variances of the same-direction I branch and the orthogonal Q branch of [K], N c Represents a complex normal distribution;

[0089] The integral value of the environmental noise η[K] also follows a zero-mean Gaussian distribution:

[0090]

[0091] in, The power of the environmental noise η[K] is represented by N, and the period length of the pseudo-random code is represented by N.

[0092] After incorporating both white noise and cross-correlation interference, the noise floor estimate is as follows:

[0093]

[0094] Where N0 represents the ambient noise power, T s Indicates the sampling period. The noise basis is obtained from the noise basis estimation. It increases as the total power of the spoofing signal increases.

[0095] In step S10, a signal strength attenuation model with distance is established:

[0096]

[0097] Where E is the electromagnetic noise intensity at a distance r from the point source, and E0 is the initial electromagnetic noise intensity at a distance r0 from the point source;

[0098] The farther away from the point source, the greater the attenuation of electromagnetic noise intensity.

[0099] As an optimization of the above embodiment, in step S20, a detection mechanism based on satellite navigation result data comparison is established to detect navigation deception attacks in terms of information content, including the following steps:

[0100] S211, in the drone position controller, set the speed limit v. max ;

[0101] S212, the UAV coordinates (X) are obtained from GNSS (Global Navigation Satellite System). U1 ,Y U1 Z U1 ), and the coordinates (X) of the previous moment. U0 ,Y U0 Z U0 ), and calculate the straight-line distance Δρ;

[0102]

[0103] At the same time, based on the timestamps t of two adjacent moments U0 and t U1 Calculate the time difference Δt;

[0104] Δt=t U1 -t U0

[0105] The velocity during that time interval Δt is:

[0106] S213, if the velocity v 01 >2v max If the signal is positive, it is considered to be a navigation spoofing signal; otherwise, the signal is considered reliable.

[0107] In step S20, a detection mechanism based on signal-noise intensity is established to detect navigation spoofing attacks on the physical channel, including the following steps:

[0108] S221, Obtain the carrier noise density ratio C / N0 from the received navigation message, where C represents the carrier power. Calculate the signal-to-noise ratio (SNR) based on the equivalent noise bandwidth B and the carrier noise density ratio C / N0:

[0109]

[0110] S222, obtain the signal power P from the power meter. s And the signal-to-noise ratio (SNR), calculate the noise power P. n :

[0111]

[0112] S223, for noise power P n A maximum threshold P is set based on the intensity distribution of signal noise when not subjected to navigation spoofing attacks. n,max If P n >P n,max If the signal is positive, it is considered to be a navigation spoofing signal; otherwise, the signal is considered reliable.

[0113] As an optimization of the above embodiment, in step S30, a fast attack source avoidance based on signal strength gradient descent is designed, including:

[0114] like Figure 2 As shown, a threshold E0 is set, assuming that the attack will not be effective when the electromagnetic noise intensity drops below E0. P is set based on the noise intensity distribution under normal environmental conditions. n,max And E0. First, data samples were collected to record the distribution of environmental noise at different intensities over a period of time. Based on the Gaussian distribution fitting results, the noise intensity corresponding to a cumulative probability of 99.75% was selected as P. n,max Let E0 be the noise intensity corresponding to a cumulative probability of 98%.

[0115] S301, According to the signal strength attenuation model with distance, the electromagnetic noise intensity is inversely proportional to the square of the distance to the point source. The farther away from the point source, the greater the attenuation of the electromagnetic noise intensity. Set a threshold E0, and assume that when the electromagnetic noise intensity drops below E0, the attack will not be effective. At this time, the distance between the drone and the point source is r0.

[0116] S302, Establish a repulsive field centered on the point source;

[0117] The repulsive field function is:

[0118]

[0119] Where U(r) represents the repulsive potential energy when the distance between the UAV and the point source is r;

[0120] S303, by the definition of an artificial potential field, experiences a force in a force field that is the negative gradient of the artificial potential field:

[0121]

[0122] F(r) represents the force experienced by the UAV in the artificial potential field. In the point source model, the direction of the force at any location is from the point source to that location, and this direction is obtained by the maximum signal method. The gradient operator represents the rate of change and direction of change of a scalar function.

[0123] The drone flies away from the attack source to reduce the impact of the deceptive attack signal until the measured maximum noise signal strength drops below the threshold E0.

[0124] A directional antenna is used to receive signals, and the direction in which the strongest signal is received is measured by horizontal rotation as the direction of the attack source.

[0125] As an optimization of the above embodiment, in step S40, the location and tracing of the attack source includes the following steps: after completing the avoidance of the attack source, the direction of the interference source is determined by the maximum signal method, the direction of the interference source is selected to fly in a direction perpendicular to the interference source, and the direction of the interference source is measured again after flying a certain distance.

[0126] like Figure 3 As shown, on a two-dimensional plane, the target interference source is located at (x, y), and the two measurement positions are (x1, y1) and (x2, y2). The arrival angles of the measured noise signals are θ1 and θ2. The coordinates of the measurement points are obtained according to the definition of angle:

[0127]

[0128] Written in matrix form:

[0129] AX = Z

[0130] In the formula,

[0131] Expanding the matrix, we obtain the estimated location of the interference source:

[0132]

[0133] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A method for detection, avoidance and tracing of navigation spoofing attacks for unmanned aerial vehicles, characterized in that, The method comprises the steps of: S10: establishing a satellite navigation spoofing attack model, including a signal-to-noise intensity influence model and a signal intensity distance attenuation model; S20: establishing a navigation spoofing attack detection mechanism according to the signal-to-noise intensity influence model and the signal intensity distance attenuation model, including satellite navigation result data comparison and signal noise intensity detection, to detect the navigation spoofing attack; S30: performing signal intensity gradient descent attack source avoidance according to the signal intensity distance attenuation model; S40: on the basis of the attack source avoidance, realizing the positioning and tracing of the attack source.

2. The method of claim 1, wherein, In the step S10, the signal-to-noise intensity influence model is established, comprising the steps of: S101, in a real attack scene, the attacker uses multiple signals to implement the attack, the pseudo-random codes of each signal will interfere with each other, and there are N Auth real signals and N Spoof fake signals, wherein the autocorrelation amplitude y1[K] of the first real signal in the Kth pseudo-random code period is represented as: where F il [K] represents the cross-correlation interference between the ith real signal and the 1th real signal, F kl [K] represents the cross-correlation interference between the kth spoof signal and the 1th real signal; and respectively represent the ith real signal power and the kth spoof signal power, exp(·) represents the exponential function, represents the carrier phase of the 1th signal, j represents the imaginary unit, is used to describe the angular frequency and the phase of the signal component, and η[K] is the environmental noise subject to the normal distribution; S102: a pseudo-random code f unrelated to all PRN codes in the real system is fictitiously created, and correlation operation is performed on the signal to estimate the noise base; since the fictitious attribute of f, the cross-correlation value with other PRN codes tends to be zero, and thus the following condition is met: where μ l represents the carrier angular frequency of the 1st signal; Thus, the relevant results for the noise floor estimation are: In the formula, F if [K] and F kf [K] is the parameter introduced by the ith real signal and the kth spoof signal in the noise floor estimation, respectively, and var[·] represents the variance of a random variable. S103, the above-mentioned parameter F if [K] complex signal containing the I and Q branches, whose values are distributed according to a two-dimensional normal distribution with mean 0, and is represented by: wherein and respectively represent F if variance of the co-phased I-branch and quadrature Q-branch of [K], N c denotes a complex normal distribution; The integral value generated by the environmental noise η[K] also obeys the zero-mean Gaussian distribution: wherein represents the power of the ambient noise η[K], N represents the period length of the pseudo-random code; after the white noise and the cross-correlation interference factor are integrated, the noise base is estimated as: where N0represents the ambient noise power, T s denotes the sampling period; From the noise floor estimate: noise floor Rises with increasing total power of the spoofing signal.

3. The method of claim 1, wherein, In the step S10, the signal intensity distance attenuation model is established: Wherein, E is the electromagnetic noise intensity at a distance of r from the point source, E0 is the initial electromagnetic noise intensity at a distance of r0 from the point source; The farther the distance from the point source, the greater the attenuation of the electromagnetic noise intensity.

4. The method of claim 1, wherein, In the step S20, the detection mechanism based on satellite navigation result data comparison is established to detect the navigation spoofing attack in the information content, comprising the steps of: S211, in the UAV position controller, set the speed upper limit v max ; S212, get the UAV coordinates (X U1 ,Y U1 ,Z U1 ) from GNSS, and the coordinates (X U0 ,Y U0 ,Z U0 ) at the previous time, and calculate the straight-line distance Δρ; meanwhile, calculate the time difference Δt according to the timestamps t U0 and t U1 of the adjacent two time points; then the speed in the time Δt is: S213, if the speed v 01 >2v max a navigation spoofing signal is considered to exist, otherwise the signal is considered to be authentic.

5. The method of claim 1 or 4, wherein, In the step S20, the detection mechanism based on signal noise intensity is established to detect the navigation spoofing attack in the physical channel, comprising the steps of: S221: obtaining the carrier-to-noise density ratio C / N0 from the received navigation message, wherein C represents the carrier power; calculating the signal-to-noise ratio SNR of the signal according to the equivalent noise bandwidth B and the carrier-to-noise density ratio C / N0: S222, calculate the signal power P from the power meter s , and the signal-to-noise ratio SNR, calculate the noise power P n : S223, for the noise power P n , a maximum threshold P n,max is set based on the intensity distribution of the signal noise when not subject to a navigation spoofing attack; if P n >P n,max , a navigation spoofing signal is considered to exist, otherwise the signal is considered to be reliable.

6. The method of claim 1 or 3, wherein, In the step S30, the attack source rapid avoidance based on the signal intensity gradient descent is designed, comprising: S301: according to the signal intensity distance attenuation model, the electromagnetic noise intensity is inversely proportional to the square of the distance from the point source, the farther the distance from the point source, the greater the attenuation of the electromagnetic noise intensity, a threshold E0 is set, and it is considered that when the electromagnetic noise intensity decreases below E0, the attack will not take effect, at this time the distance between the unmanned aerial vehicle and the point source is r0; S302: a repulsive force field is established with the point source as the center; The repulsive force field function is: Wherein, U(r) represents the repulsive potential energy of the unmanned aerial vehicle when the distance from the point source is r; S303: according to the definition of artificial potential field, the force in the force field is the negative gradient of the artificial potential field: F(r) is the force received by the UAV in the artificial potential field; in the point source model, the force direction at any position is from the point source to the position, and the direction is obtained by the maximum signal method; is the gradient operator, representing the rate and direction of change of the scalar function; The unmanned aerial vehicle flies away from the attack source to reduce the influence of the spoofing attack signal until the maximum noise signal intensity measured decreases below the threshold E0.

7. The method of claim 6, wherein, The signal is received using a directional antenna, and the direction of the strongest signal received is measured as the direction of the attack source by horizontal rotation.

8. The method of claim 1, wherein, In the step S40, the positioning and tracing of the attack source, comprising the steps of: after completing the avoidance of the attack source, the direction of the interference source is determined by the maximum signal method, the direction perpendicular to the interference source is selected for flight, and the interference source is measured again after flying a certain distance; In two-dimensional plane, the target interferer is located at (x, y), the two measurement positions are (x1, y1) and (x2, y2), the measured noise signal angles of arrival are θ1 and θ2, and the position estimation of the target interferer is calculated as follows:

Images