A radar pulse signal concealment method based on artificial noise to resist detection and sorting

Abstract

The present invention proposes a new radar pulse signal concealment method based on artificial noise that is resistant to detection and sorting. Before transmitting the signal, by making full use of the directional characteristics of the channel, artificial noise designed based on the directional characteristics of the channel is cleverly superimposed on the radar pulse waveform to be transmitted, and the superimposed artificial noise does not cause any disturbance to the radar pulse waveform transmitted by the own party. Finally, by controlling the power ratio of the artificial noise and the own radar pulse signal, a compromise optimization between the own detection performance and the third-party pulse detection probability can be achieved, effectively reducing the detection probability of the radar signal by the third-party receiver and achieving anti-detection performance; by designing and optimizing the parameters of the random overlap of the time domain artificial noise and the radar signal, the PRI, pulse width, TOA and other parameters of the radar signal can be accurately concealed, thereby disrupting the reconnaissance receiver from forming the correct pulse description word PDW, thereby achieving the anti-sorting and waveform concealment capabilities of the radar pulse signal with the lowest cost-effectiveness.

Description

Technical Field

[0001] The present invention relates to radar signal processing technology, in particular to radar pulse waveform concealment, anti-detection and anti-sorting technology. Background Art

[0002] The main task of electronic reconnaissance is to obtain the carrier frequency, arrival direction, pulse width, arrival time, and pulse amplitude of the target radar pulse signal. It is very important to ensure the concealment of the key parameters of the radar pulse signal. At present, the existing technology mainly achieves the anti-interception and concealment of the radar pulse signal by controlling the energy and other parameters of the transmitted pulse waveform, which will lead to a decrease in the detection performance of the party. Artificial noise is a secure transmission method used in the field of wireless communications. It mixes pre-coded noise signals that are orthogonal to the receiving channel into the transmitted signal. While ensuring that the receiving performance of the receiver in the target direction is not affected, it interferes with eavesdropping users in the non-target direction. In electronic reconnaissance and counter-reconnaissance, the radar pulse waveform also needs to be concealed and counter-reconnaissance. However, there is currently no technology that fully considers the directional characteristics of the channel to fully utilize the radar pulse waveform to conceal, resist detection, and resist sorting.

[0003] Currently, research on anti-interception technology for modulated radar signals focuses on carrier frequency energy, while baseband anti-detection technology remains a niche topic. Current research on anti-sorting radar signals involves simply adding a jamming pulse to the radar pulse signal for splicing. This approach does not conceal the radar pulse signal's TOA and PRI parameters, but only the PW parameter. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to propose a radar pulse signal anti-detection sorting waveform concealment scheme which utilizes the directional characteristics of the channel to reduce the detection probability of the radar pulse signal by a third-party receiver.

[0005] The technical solution adopted by the present invention to solve the above technical problems is a method for transmitting radar pulse signals based on artificial noise and resistant to detection and sorting, comprising the steps of:

[0006] The radar pulse signal transmitter determines the power ratio of the signal α1 and the power ratio of the artificial noise α2;

[0007] The radar pulse signal transmitter generates a first noise n which is sent synchronously with the radar pulse signal. AN1 (t), the second noise n is generated by pseudo-random code control AN2 (t) and the third noise n AN3 (t); second noise n AN2 (t) is used to randomly mask the leading edge of the radar pulse waveform, the third noise n AN3(t) is used to randomly mask the trailing edge of the radar pulse waveform; the first noise n AN1 (t), the second noise n AN2 (t) and the third noise n AN3 (t) obeys Gaussian distribution, and the second noise n AN2 (t) and the third noise n AN3 (t) is equal to the power of the first noise n AN1 The power of (t) is less than the second noise n AN2 (t) and the third noise n AN3 (t) power;

[0008] The radar pulse signal transmitter is based on the first noise n AN1 (t), the second noise n AN2 (t) and the third noise n AN3 (t) Complete artificial noise n AN Generation of (t): n AN (t) = b r [n AN1 (t)+n AN2 (t)+n AN3 (t)]; where t is the time variable, b r is the AN precoding matrix, and the legal channel H between the radar pulse signal transmitter and the radar pulse signal receiver B Orthogonal, satisfying H B b r =0;

[0009] The radar pulse signal transmitter adds artificial noise n to the radar pulse signal to be sent according to the power ratio α1 and α2 of the signal and artificial noise. AN (t) and then sent to the radar pulse signal receiver.

[0010] The present invention makes full use of the directional characteristics of the channel. While transmitting the radar pulse waveform, it cleverly superimposes artificial noise designed based on the directional characteristics of the channel on the third-party radar pulse signal. The superimposed artificial noise does not cause any disturbance to the radar pulse waveform transmitted by the own party. By controlling the power ratio of the artificial noise and the own radar pulse signal, the detection probability of the radar pulse signal by the third-party receiver is controlled, thereby achieving anti-detection performance. By randomly overlapping the artificial noise with the radar pulse, the present invention can conceal the arrival time TOA, pulse repetition period PRI and pulse width PW of the radar pulse signal, thereby disrupting the formation of the correct pulse descriptor PDW by the reconnaissance receiver. By further adjusting the power ratio of the artificial noise and the radar signal, the radar signal anti-detection technology at the baseband end can be realized.

[0011] The present invention has the beneficial effect of superimposing artificial noise on the transmitted radar pulse waveform without disrupting the transmitted radar pulse waveform, but instead causing a third-party receiver to generate an erroneous pulse descriptor word (PDW), thereby achieving radar pulse signal anti-sorting capabilities with minimal cost-effectiveness. Furthermore, by controlling the power ratio α2 / α1 between the artificial noise and the radar pulse signal, a compromise between detection performance and third-party detection probability can be achieved, achieving anti-detection performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 System model: Third-party Eve detects and sorts radar pulse signals;

[0013] Figure 2 Receive waveform for Eve: anti-sorting type based on artificial noise;

[0014] Figure 3 Bob receives the waveform: unaffected;

[0015] Figure 4 Schematic diagram of the effect of the power ratio of artificial noise and radar signal on the detection probability;

[0016] Figure 5 This is the 1-4 order CDIF radar sorting result diagram;

[0017] Figure 6 This is a schematic diagram of the pulse width error rate at the Eve end. DETAILED DESCRIPTION

[0018] System model such as Figure 1 As shown in the figure, Alice is a radar pulse signal transmitter, and Bob is a legitimate radar receiver. Alice and Bob are transmitting and receiving radar pulses. Bob receives and extracts Alice's transmitted pulse signal, performs parameter estimation and detection, and in some scenarios, further extracts internal information of the pulse, such as the interrogation / response signal of a secondary radar. At this time, a third party, Eve, is eavesdropping on this radar pulse process from a non-target direction. Eve intercepts the radar pulse signal, then detects and sorts it. Assume that the legitimate channel between Alice and Bob is H B , the reconnaissance channel between Alice and Eve is H E In order to achieve anti-detection and anti-sorting of radar pulse signals, Alice adds the artificial noise AN in the form of precoding designed by the present invention to the radar pulse signal and ensures that Bob is not interfered by the artificial noise AN. B With the eavesdropping channel H ETo ensure that the difference does not affect Bob and interferes with the detection and sorting of radar pulse signals by third parties, the precoding of artificial noise is designed to be orthogonal to Bob's channel.

[0019] The radar waveform x(t) sent by Alice is expressed as:

[0020]

[0021] Among them, P t is the total transmitted power, α1, α2 are the power ratios of the signal and artificial noise, defined as the ratio of the total transmitted power, α1 + α2 = 1. s(t) is the radar pulse signal, n AN (t) is the artificial noise designed by the present invention and obeys Gaussian distribution, b t is the beamforming matrix of the radar transmitter, and t is the time variable.

[0022] Bob channel matrix H B and Eve channel matrix H E The expression is:

[0023]

[0024] in, T is the transpose of the matrix, h bt is θ bt The channel vector in the directional far-field space; h br is θ br The channel vector in the directional far-field space; h et is θ et The channel vector in the directional far-field space; h er is θ er Channel vector in directional far-field space;

[0025] N A ,N B ,N E The total number of array elements of Alice, Bob, and Eve, n A ,n B ,n E They are the array element number variables of Alice, Bob, and Eve, and their values ​​range from 1 to N. A ,1~N B ,1~N E ;

[0026] θ bt is the outgoing signal beam angle from the transmitter Alice to the receiver Bob, θ br is the incident signal beam angle from the transmitter Alice to the receiver Bob, θ etThe outgoing signal beam angle from the transmitter Alice to the third party Eve, θ er The beam angle of the incident signal from the transmitter Alice to the third party Eve;

[0027] Δφ bt Alice phased array antenna elements at θ bt Phase difference in direction, Δφ br is the position of each element of Bob phased array antenna at θ bt Phase difference in direction, Δφ et Alice phased array antenna elements at θ et Phase difference in direction, Δφ er is the position of each element of Bob phased array antenna at θ er Phase difference in direction. Δφ bt The expression is:

[0028]

[0029] Where d is the phased array element spacing, λ is the RF signal wavelength, Δφ br ,Δφ et ,Δφ er Similarly, the expression of the artificial noise to be added is:

[0030] n AN (t) = b r [n AN1 (t)+n AN2 (t)+n AN3 (t)];

[0031] Among them, b r is the AN precoding matrix, and the target channel H B Orthogonal, satisfying H B b r =0.n AN1 (t) is artificial noise 1, which is sent synchronously with the radar pulse signal, n AN2 (t),n AN3 (t) is artificial noise 2 and artificial noise 3, controlled by pseudo-random code; n AN2 (t) is used to randomly mask the leading edge of the radar pulse waveform, n AN3 (t) is used to randomly mask the trailing edge of the radar pulse waveform, n AN1 (t), n AN2 (t), n AN3 (t) all obey Gaussian distribution, and the power relationship of the three parts of noise is P AN2 =P AN3 >P AN1 , and P AN2 +P AN3 +PAN1 =P AN .

[0032] After transmission through the channel, the target receiver Bob receives the signal r b (t) is:

[0033]

[0034] Among them, b br is the array steering vector at Bob’s receiver, and n(t) is the additive white Gaussian noise (AWGN) in the channel.

[0035] Because H B b r =0, therefore, And b t The radar transmitter beamforming matrix can be combined with the channel vector h et Offset, that is The Bob signal is not disturbed by injected artificial noise.

[0036] The signal Eve receives is:

[0037]

[0038] Among them, b er is the array guide vector of the third-party Eve, h et b r ≠0, the third party cannot eliminate the influence of artificial noise AN, and the radar pulse signal received by Eve is seriously interfered with by the artificial noise, which will interfere with the third party Eve's detection and sorting of the radar pulse signal, thereby achieving signal waveform concealment and anti-detection sorting of the radar pulse signal.

[0039] The detection method for radar signals is mainly based on a binary detection model, which is as follows:

[0040]

[0041] Where, 0≤t≤T, T is the detection time length, s(t) is the radar signal, n(t) is the artificial noise, and n(t)~N(0,σ 2 ), N(0,σ 2 ) means the mean is 0 and the variance is σ 2 Gaussian distribution. Eve samples x(t) in time T and can obtain N observations x1(t),x2(t),...,x N (t), when Eve receiver receives the signal, according to the Neyman-Pearson criterion, that is, maintaining the false alarm probability P FWhen the noise power is known, the optimal detection threshold V is obtained by the following integration method. T , greater than the detection threshold V T The signal is recognized as a valid signal, which is less than the detection threshold V T Identification of as noise:

[0042]

[0043] V T For the decision-making mechanism in the radar system according to P F The best decision threshold is obtained, p(G|H0) is the probability density function of the statistic G under the original hypothesis of noise-free condition H0, N0 is the noise power spectrum density, E S is the signal energy, and Φ is the complementary cumulative distribution function of the standard normal distribution.

[0044] By adjusting the power ratio of artificial noise and radar signal α2 / α1, E is affected by α2 / α1. S The ratio of N0 to the detection probability P of the radar signal after it is intercepted by a third-party receiver can be accurately controlled. D . The relationship is as follows:

[0045]

[0046] Eve's radar sorting primarily separates overlapping radar emitter signals. The identified arrival time TOA further derives parameters such as the pulse repetition period (PRI) and pulse width (PW). This requires the use of sorting algorithms such as the cumulative difference histogram (CDIF) and the sequence difference histogram (SDIF).

[0047] Without using this method, the TOA of the radar pulse signal sorted by Eve is periodic, and TOA is easy to obtain, so the radar sorting success rate is high. This patent uses artificial noise and radar signals to randomly overlap in the time domain, and the TOA, PW and other parameters of the radar signal are concealed and randomized. In this way, it is difficult for traditional radar sorting algorithms to obtain the parameters of the radar signal because the parameters of the radar signal after random aliasing in the time domain are already randomized, such as Figure 2 shown.

[0048] The artificial noise of the present invention can be designed in the above two ways. Among them, for Eve, artificial noise 1 is sent synchronously with the radar pulse signal. Artificial noise 2 and artificial noise 3 are randomly overlapped with the radar pulse signal to mask the parameters such as TOA, PW and PRI of the radar pulse signal. The degree of random overlap is controlled by the pseudo-random code PN. The power P of artificial noise 2 is 1 / 4. AN2 and the power P of artificial noise 3 AN3 Equal, artificial noise 1 is used to cover the radar pulse waveform, and the power of artificial noise 1 PAN1 Greater than the power of the radar pulse signal. At the same time, the peak power of artificial noise 2 and artificial noise 3 should be equal to the peak power of artificial noise 1 superimposed with the radar pulse waveform, so that the PRI and PW information can be masked. For Bob, since the artificial noise is orthogonal to Bob's channel, Bob is not affected by the artificial noise, so Bob can receive the radar pulse signal completely. Figure 3 shown.

[0049] Example

[0050] S1: The radar pulse signal is superimposed with a synchronous artificial noise 1 sent at the same time, and then a pseudo-random code PN is generated to control the overlap degree δ1, δ2 of artificial noise 2 and artificial noise 3 with the radar pulse signal. Then artificial noise 2 and artificial noise 3 are superimposed on the radar pulse signal. At this time, each radar pulse is randomly overlapped with artificial noise 2 and artificial noise 3, and the peak power of artificial noise 2 is consistent with that of artificial noise 3, both of which are greater than the power of the radar pulse signal. Then the artificial noise AN is multiplied by the precoding matrix b that is orthogonal to the target channel. r , used to interfere with the third-party Eve while leaving the legitimate end Bob unaffected. Finally, the radar pulse signal with artificial noise superimposed on the baseband end is sent out. The specific relationship is shown below.

[0051]

[0052] Among them, δ1 and δ2 are the frontier masking noise n AN2 (t) and the trailing edge masking noise n AN3 (t) is the degree of overlap with the radar pulse signal, and δ1,δ2∈(0,N], N is the length of the radar pulse waveform, and δ1 and δ2 are independent of each other.

[0053] S2: Eve performs matched filtering on the received radar signal and then uses the Neyman-Pearson criterion to calculate the false alarm probability P according to the preset false alarm probability P. F Determine the detection threshold V T , through V T Radar signal detection and judgment. Assuming that Eve receiver knows the power of artificial noise, the optimal detection threshold V is obtained in the worst case. T :

[0054]

[0055] At this time, the Eve receiver has a certain probability of detecting the radar pulse signal, and then by controlling the power ratio α2 / α1 when the artificial noise and radar signal are sent, the detection probability P of the third-party receiver for the intercepted radar pulse signal can be accurately controlled. DBelow the set safety limit:

[0056]

[0057] S3: The Eve radar receiver uses the CDIF sorting algorithm to sort the received signal. It obtains the arrival time of the received signal by detecting the rising edge and further calculates the TOA difference of each level:

[0058] TOA'=TOA k+c -TOA k ,k=1,2,...,n;

[0059] Among them, TOA' is the difference of TOA at each level, TOA k is the kth TOA value, c is the level, and n is the total number of TOAs.

[0060] By counting the TOA difference frequency of each order and comparing it with the threshold T threshold , take the value exceeding the threshold as PRI, thereby obtaining the PRI parameter of the radar signal, and then further obtain the pulse width PW by detecting the rising and falling edges. threshold Settings:

[0061] T threshold =α×T / τ;

[0062] Where α is a constant related to the pulse loss rate and has a value range of (0,1]. T represents the observation time, and τ represents the arrival time difference of the pulses, which is the horizontal axis variable of the cumulative difference histogram.

[0063] By injecting random overlap of artificial noise and radar pulse signal in the time domain, since artificial noise randomly overlaps with each radar pulse, the degree of overlap is expressed as δ, δ k is the degree of overlap at the kth TOA, and the TOA difference at each level is:

[0064] TOA'=(TOA k+c -δ k+c )-(TOA k -δ k ),k=1,2,...,n;

[0065] Therefore, the above equation shows that the injected AN can disrupt the Eve sorting system's estimation of TOA, randomizing Eve's TOA estimate and thus achieving anti-sorting of radar signals. This makes it difficult for the sorting system to obtain correct parameters such as PRI and PW.

[0066] like Figure 4As shown in the figure, full-time AN ​​transmits AN signals throughout the entire time period. Except during pulse signal transmission periods, the entire transmitter power is used to transmit AN signals. Transmit-period AN transmits AN signals only when radar pulse signals are being transmitted, and does not transmit AN signals at other times. Therefore, transmit-period AN is more power-efficient than full-time AN. It can be seen that as the noise-to-signal power ratio increases, the detection probability of both methods decreases. However, because transmit-period AN is more power-efficient, its detection probability is lower than that of full-time AN ​​for the same power ratio.

[0067] like Figure 5 As shown in the figure, the yellow line represents the detection threshold. The PRI of the actual radar pulse signal received by Bob is 1000us, and the corresponding 1st-order to 4th-order TOA differences are 1000us, 2000us, 3000us, and 1000us, respectively. It can be seen that the 1st- to 4th-order differences detected by the CDIF sorting system on Bob's side are all correct and exceed the threshold, corresponding to the blue line in the figure, thus extracting the correct PRI. On Eve's side, however, due to the influence of artificial noise, the TOA differences received by Eve have become randomized. Because each radar pulse randomly overlaps with the artificial noise, Eve receives multiple TOA differences for each CDIF order. The corresponding CDIF histograms for each order are multiple random values, corresponding to the red line. Furthermore, the 2nd-order CDIF histogram does not exceed the threshold, so the sorting system cannot obtain the true PRI.

[0068] from Figure 6 It can be seen from the figure that as the power ratio of random artificial noise to radar signal α2 / α1 increases, the pulse width error rate gradually increases, which interferes with the sorting system's recognition of PW parameters.

Claims

1. A radar pulse signal concealment method based on artificial noise emission and resistant to detection and sorting, characterized in that: Including steps: The radar pulse signal transmitter determines the power ratio of the signal α1 and the power ratio of the artificial noise α2; The radar pulse signal transmitter generates a first noise n which is sent synchronously with the radar pulse signal. AN1 (t), the second noise n is generated by pseudo-random code control AN2 (t) and the third noise n AN3 (t); second noise n AN2 (t) is used to randomly mask the leading edge of the radar pulse waveform, the third noise n AN3 (t) is used to randomly mask the trailing edge of the radar pulse waveform; the first noise n AN1 (t), the second noise n AN2 (t) and the third noise n AN3 (t) obeys Gaussian distribution, and the second noise n AN2 (t) and the third noise n AN3 (t) is equal to the power of the first noise n AN1 The power of (t) is less than the second noise n AN2 (t) and the third noise n AN3 (t) power; The radar pulse signal transmitter is based on the first noise n AN1 (t), the second noise n AN2 (t) and the third noise n AN3 (t) Complete artificial noise n AN Generation of (t): n AN (t) = b r [n AN1 (t)+n AN2 (t)+n AN3 (t)]; where t is the time variable, b r is the AN precoding matrix, and the legal channel H between the radar pulse signal transmitter and the radar pulse signal receiver B Orthogonal, satisfying H B b r =0; The radar pulse signal transmitter adds artificial noise n to the radar pulse signal to be sent according to the power ratio α1 and α2 of the signal and artificial noise. AN (t) and obtain the final transmission signal, which is then transmitted to the radar pulse signal receiver.

2. The method according to claim 1, wherein: The radar pulse signal transmitter adds artificial noise n to the radar pulse signal to be sent according to the power ratio α1 and α2 of the signal and artificial noise. AN The final transmitted signal obtained after (t) is: Among them, P t is the total transmission power, α1 is the ratio of the signal to the total transmission power, α2 is the ratio of the artificial noise to the total transmission power, α1+α2=1; s(t) is the radar pulse signal to be sent, b t is the beamforming matrix of the radar transmitter.

3. The method according to claim 2, wherein: The radar pulse signal transmitter controls the detection probability P after the radar signal is intercepted by a third-party receiver by adjusting the power ratio α2 / α1 between the artificial noise and the radar signal. D .

4. The method according to claim 3, wherein: The detection probability P after the radar signal is intercepted by a third-party receiver D The calculation method is: Where Φ is the complementary cumulative distribution function of the standard normal distribution, N0 is the noise power spectral density, E S is the signal energy, V T is the radar system decision threshold; E S The ratio of N0 to N1 is controlled by α2 / α1.

5. The method according to claim 4, wherein: Radar system decision threshold V T According to the Neyman-Pearson criterion, when determining the false alarm probability P F Then we calculate:

6. The method according to claim 5, wherein: By adding artificial noise n AN (t) is randomly overlapped with the radar pulse signal to be sent in the time domain to hide the arrival time TOA, pulse repetition period PRI and pulse width PW of the radar pulse signal to be sent.

7. The method according to claim 6, wherein: The arrival time difference TOA' of each level obtained by the third party receiving the transmitted signal is: TOA'=(TOA k+c -δ k+c )-(TOA k -δ k ),k=1,2,...,n; Among them, TOA' is the TOA difference of level c, TOA k is the kth TOA value, c is the level, n is the total number of TOAs, δ k is the degree of overlap at the kth TOA.

Images