Self-interference signal suppression method based on transfer function fitting and phase matching

Abstract

The present invention proposes a self-interference signal suppression method based on transfer function fitting and phase matching. When there is no target, the transfer function is fitted to generate a cancellation signal. When a target may be present, multi-stage time delay estimation is used to perform phase matching between the self-interference signal and the cancellation signal to achieve a better self-interference suppression effect. When suppressing self-interference of underwater sonar, the present invention uses the self-interference signal in the target-free state to perform transfer function fitting, which has a small amount of calculation and can meet the self-interference suppression requirements. Multi-stage time delay estimation is used to accurately match the phase of the cancellation signal and the self-interference signal, resulting in a good self-interference suppression effect, thereby improving the range of the underwater detection and communication integrated system. For multi-channel signals, transfer function fitting and phase matching are performed once in each channel to complete the self-interference suppression of the multi-channel signal.

Description

Technical Field

[0001] The present invention belongs to the fields of detection technology and communication technology, and in particular to the field of self-interference signal suppression and signal processing at an integrated underwater detection and communication receiving end. The present invention is a self-interference signal suppression method based on transfer function fitting and phase matching. Background Art

[0002] In recent years, the integrated detection and communication system has been widely used in the radar field because it can share hardware equipment, reduce system size, share transmission signals, reduce energy consumption and improve frequency band utilization.

[0003] Since the integrated system needs to adopt a fully shared working system, the system receiver will receive the leakage signal from the transmitter, and its signal strength is much greater than the echo signal, causing the echo signal to be submerged in the transmitted leakage signal, making it difficult to extract the target information in the echo. How to suppress the self-interference signal is the key to ensuring the stability of the integrated detection and communication system.

[0004] There is already a relatively complete theory for eliminating radar self-interference signals. Pursula, Lin and others use a variety of methods such as adaptive elimination, reflected power compensation elimination and digital elimination to suppress transmitted carrier leakage. The self-interference suppression effect for radar can be as high as 60dB or more.

[0005] However, due to the significant differences between acoustic and electromagnetic waves, current radar self-interference suppression methods are not applicable to integrated underwater detection and communication systems, leaving a significant gap in self-interference suppression for these systems. In 2018, Lu Jun et al. employed an analog domain method based on error factor correction to achieve a 22-25dB suppression effect on sonar self-interference signals. However, this method struggles to meet the detection range requirements of integrated underwater detection and communication systems and is computationally intensive. Summary of the Invention

[0006] In order to solve the problems existing in the prior art, the present invention proposes an underwater multi-channel self-interference signal suppression method based on transfer function fitting and phase matching, including: fitting the transfer function to generate a cancellation signal when there is no target, and using multi-stage time delay estimation to perform phase matching between the self-interference signal and the cancellation signal when a target may be present, so as to achieve better self-interference suppression effect.

[0007] The technical solution of the present invention is:

[0008] A self-interference signal suppression method based on transfer function fitting and phase matching comprises the following steps:

[0009] Step 1: In a targetless scenario, collect the self-interference signal x1(t) received by the receiving end of the underwater sounding integrated system. The self-interference signal x1(t) is transmitted by the underwater sounding integrated system transmitting signal s(t) to the receiving array of the underwater sounding integrated system;

[0010] Step 2: According to the formula

[0011]

[0012] Solve the transmit leakage transfer function H1(f) from x1(t), where X1(f) and S(f) are the frequency domain forms of x1(t) and s(t), respectively;

[0013] Step 3: Construct the cancellation signal spectrum Y1(f) based on the transmit leakage transfer function H1(f) obtained in step 2:

[0014] Y1(f)=S(f)*H1(f)

[0015] Then perform IFFT on Y1(f) to obtain the time domain cancellation signal y1(t);

[0016] Step 4: Collect the next frame signal x2(t) received by the receiving end of the underwater sounding integrated system, use the time domain cancellation signal y1(t), use the basic cross-correlation method to make a rough estimate of the time delay, and obtain a rough estimate of the time delay τ0

[0017] Step 5: Use the rough estimate of the delay τ0 obtained in step 4 Delay the received signal x2(t) accordingly to achieve coarse alignment:

[0018]

[0019] Where x'2(t) is x2(t) through The result obtained after the delay; further delay estimation is performed by the cross-spectral phase method to obtain the estimated value of the delay τ1 between x'2(t) and x1(t)

[0020] Step 6: Based on the estimated value Perform time delay compensation on Y1(f) and reconstruct the cancellation signal:

[0021]

[0022] Self-interference suppression is achieved, and the residual signal after self-interference cancellation is obtained as:

[0023] e(t)=x'2(t)-y2(t)

[0024] Where y2(t) is the time domain signal obtained by IFFT of Y2(f).

[0025] Furthermore, the process of roughly estimating the time delay using the basic cross-correlation method in step 4 is as follows:

[0026] Calculate the cross-correlation function R between y1(t) and x2(t) y1x2 (τ).

[0027]

[0028] Where X2(f) is the frequency domain form of x2(t), is the conjugate of X2(f); by calculating R y1x2 The maximum value of (τ) gives a rough estimate of the delay τ0

[0029] Furthermore, in step 5, the process of further delay estimation using the cross-spectral phase method is as follows:

[0030] According to the formula

[0031]

[0032] calculate Among them, Y1 * (f) is the conjugate of Y1(f), X'2(f) is the frequency domain form of x'2(t), and calculate The phase spectrum θ(f) is given by the formula

[0033]

[0034] The estimated value of the delay τ1 is obtained by using the CSP method where f i is the i-th frequency point of θ(f), and N is the number of frequency points of θ(f).

[0035] Furthermore, for multi-channel signals, transfer function construction and multi-stage delay estimation are performed on each channel to achieve self-interference suppression in each channel.

[0036] Beneficial effects

[0037] The underwater multi-channel self-interference signal suppression method based on transfer function fitting and phase matching proposed in the present invention uses the self-interference signal in the targetless state to perform transfer function fitting when suppressing self-interference of underwater sonar. The method has low computational complexity and can meet the self-interference suppression requirements. Multi-stage time delay estimation is used to accurately match the phase of the cancellation signal and the self-interference signal, resulting in a good self-interference suppression effect, thereby improving the effective range of the underwater detection and communication integrated system. For multi-channel signals, transfer function fitting and phase matching are performed once in each channel to complete the self-interference suppression of the multi-channel signals.

[0038] Additional aspects and advantages of the present invention will be set forth in part in the description which follows and, in part, will be obvious from the description which follows, or may be learned by practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments with reference to the accompanying drawings, in which:

[0040] Figure 1 : Single-channel self-interference suppression flow chart;

[0041] Figure 2 : Suppression effects of various methods in different directions. (a) Suppression ratio after NCC delay estimation, (b) Suppression ratio after CSP delay estimation, (c) Suppression ratio after multi-stage delay estimation. DETAILED DESCRIPTION

[0042] Since the stability of the underwater sounding integrated system depends on the suppression effect of the self-interference signal, the present invention proposes the following suppression ideas based on the characteristics of the multi-channel self-interference signal of underwater sonar:

[0043] 1. In the absence of a target, the self-interference signal leaked from the transmitter is collected at the receiving end as the output of the transmission leakage system, and the transmission signal is used as the input to estimate the transfer function of the transmission leakage system through frequency domain deconvolution.

[0044] 2. Use the transfer function and the current transmitted signal to convolve to obtain the basic cancellation signal.

[0045] 3. The receiving end collects the subsequent frame signal, which is the superposition of the self-interference signal and the target echo signal.

[0046] 4. Use the cancellation signal estimated in the targetless scenario to offset the transmission leakage in subsequent communication frames. To maximize the suppression of self-interference signals, it is necessary to align the cancellation signal with the superimposed signal waveform to achieve phase matching between the two signals. This reduces the self-interference signal after suppression and meets the requirements of subsequent work. The key issue here is whether the delay estimation accuracy between the two signals meets the requirements for self-interference suppression.

[0047] Commonly used delay estimation methods include: (a) basic cross-correlation; (b) cross-spectral phase; and (c) adaptive delay estimation. Each method has its own advantages and disadvantages. The basic cross-correlation method is simple to use, but its estimation accuracy is limited by the sampling frequency. The cross-spectral phase method has better estimation accuracy than the basic cross-correlation method, but its estimation performance decreases sharply with decreasing signal-to-noise ratio and increasing reverberation. The adaptive delay estimation method has high estimation accuracy and good adaptability to noise, and can accurately estimate delay under low signal-to-noise ratio conditions. However, its computational complexity is too high, making it unsuitable for real-time engineering applications.

[0048] To this end, the present invention uses a multi-stage delay estimation algorithm that combines the Normalized Cross Correlation (NCC) method and the Cross Spectral Phase (CSP) method. First, a coarse delay estimate is performed using the NCC method, with an estimation accuracy equal to the sampling interval, but without phase wrapping. The CSP method then performs a more refined delay estimation, with an estimation error significantly smaller than the sampling interval. This multi-stage delay estimation method estimates the precise delay between the cancellation signal and the received signal.

[0049] 5. Compensate the cancellation signal based on the time delay, and subtract the delay-compensated cancellation signal from the received signal to achieve a self-interference suppression effect.

[0050] 6. For multi-channel signals, since the self-interference signals of each channel vary significantly, it is necessary to collect the self-interference signals of multiple channels as a reference in an untargeted environment and estimate the transmit leakage system transfer function on each channel. After receiving subsequent frames, a cancellation signal is constructed and multi-stage delay compensation is performed on each channel. The cancellation signal is then subtracted from the received signal of the channel to achieve multi-channel self-interference cancellation.

[0051] At this point, the suppression of the self-interference signal is completed, and on this basis, the subsequent echo signal detection and related parameter estimation can be performed on the obtained multi-channel residual signal.

[0052] The following is a description of a specific embodiment, which includes the following specific steps:

[0053] Step 1: In an untargeted scenario, first collect the self-interference signal x1(t) received by the receiving end, which is the transmitted signal s(t) reaching the receiving array through a certain propagation path.

[0054] Step 2: Based on the relationship between convolution and Fourier transform, the transmit leakage transfer function H1(f) is calculated from x1(t):

[0055]

[0056] In formula (1), X1(f) and S(f) are the frequency domain forms of x1(t) and s(t), respectively.

[0057] Step 3: Construct the cancellation signal spectrum Y1(f):

[0058] Y1(f)=S(f)*H1(f) (2)

[0059] Performing IFFT on Y1(f) yields the time-domain cancellation signal y1(t), which can be used to cancel the self-interference signal in the next frame of received signal x2(t).

[0060] Step 4: Use the basic cross-correlation method to make a rough estimate of the delay:

[0061] Directly calculate the cross-correlation function of the signal and estimate the signal delay based on the peak position of the cross-correlation function. This method has low computational complexity and is easy to implement, but the estimation accuracy is not high. As the peak value is the sampling interval of the signal, it cannot meet the delay estimation accuracy requirements for self-interference suppression.

[0062] Perform NCC delay estimation and calculate the cross-correlation function R between y1(t) and x2(t) y1x2 (τ).

[0063]

[0064] In formula (3), X2(f) is the frequency domain form of x2(t), is the conjugate of X2(f); theoretically, when τ=τ0, R y1x2 (τ) takes the maximum value, so by calculating R y1x2 The maximum value of (τ) gives the estimated value of the delay τ0

[0065] Step 5: Use the cross-spectral phase method to perform more accurate delay estimation.

[0066] By calculating the cross-power spectrum of the signal and solving for the cross-spectral phase slope, a refined delay estimate is obtained. When solving for the slope, since the cross-spectral phase varies with a period of 2π, phase variations exceeding 2π cause phase inward rolloff, making it difficult to determine the slope. A common solution is phase unwrapping, which compensates for 2π at phase transition points and performs splicing, constructing a line with a slope of 2πτ within the signal band.

[0067] First, extract the delay estimation result obtained by the NCC method and delay the received signal accordingly to achieve coarse alignment:

[0068]

[0069] In formula (4), is the estimated value of τ0 after NCC delay estimation, and x'2(t) is the estimated value of x2(t) after The result obtained after delay.

[0070] The time delay is further estimated by the cross-spectral phase method.

[0071]

[0072] θ(f)=2πfτ1 (6)

[0073] Where Y1 * (f) is the conjugate of Y1(f), X'2(f) is the frequency domain form of x'2(t), and θ(f) is The phase spectrum, f is The frequency of , τ1 is the time delay between x'2(t) and x1(t).

[0074] The delay estimation formula of the CSP method is:

[0075]

[0076] In formula (7), is the estimated value of the time delay τ1 using the CSP method, f i is the i-th frequency point of θ(f), and N is the number of frequency points of θ(f).

[0077] Perform time delay compensation on Y1(f) and reconstruct the cancellation signal:

[0078]

[0079] Self-interference suppression is achieved, and the residual signal after self-interference cancellation is obtained as:

[0080] e(t)=x'2(t)-y2(t) (9)

[0081] Where y2(t) is the time domain signal obtained by IFFT of Y2(f).

[0082] For multi-channel signals, transfer function construction and multi-stage delay estimation can be performed once in each channel to ensure the self-interference suppression effect of each channel.

[0083] The following is an analysis of the effectiveness of the self-interference suppression method:

[0084] The NCC method, CSP method, and multi-stage delay estimation method are used to suppress self-interference. The multi-channel signals before and after suppression are subjected to beam scanning, and then the output of each beam is cross-correlated with the transmitted signal to obtain the cross-correlation peak value at each angle. By comparing the decrease in the cross-correlation peak before and after suppression, the suppression ratio at each angle can be obtained. The results are as follows: Figure 2 shown.

[0085] contrast Figure 2 In (a), (b), and (c), when only NCC estimation is used, the suppression ratio is basically around -10dB. It also becomes positive at certain angles, indicating that the self-interference signal is actually enhanced after self-interference suppression. When only CSP estimation is used, the suppression ratio is positive at most angles, and near 0dB at other angles, indicating poor suppression effect. When multi-stage delay estimation is used, the suppression ratio is negative at all angles, and the maximum suppression ratio can reach around -23dB. By comparing the suppression ratio effects obtained by the above three methods, it can be seen that multi-stage delay estimation can effectively suppress self-interference signals.

[0086] Perform echo detection and echo parameter estimation on the suppressed residual signal:

[0087] The detection method of first time domain and then space domain is adopted. First, the eight-channel data after self-interference suppression is cross-correlated with the transmitted signal. Since the target distance should generally be greater than 50m, 50m is set as the distance threshold, and the data after the distance threshold is extracted in the correlation domain. The eight-channel cross-correlation data after the distance threshold is filtered is incoherently superimposed to obtain the peak value and its corresponding time delay. Assuming that the transmitted signal is an LFM signal, the corresponding point of the time delay and its left and right sides are intercepted in the eight-channel cross-correlation data. Beamforming is performed at points (B is the frequency variation range) to obtain a relatively pure echo signal spatial spectrum.

[0088] To eliminate possible noise influences, a peak-to-average ratio threshold is set to further filter the resulting echo spatial spectrum. In the absence of a target, 100 segments of 0.5-second noise data are collected, with each segment overlapping by 10%. The noise spatial spectrum is calculated to obtain its peak-to-average ratio (PAR). The PAR sequence is sorted from highest to lowest, and the fifth PAR value is selected as the PAR threshold. This PAR represents the detection threshold for a 5% false alarm probability. The PAR threshold is then used to determine whether the resulting echo spatial spectrum exhibits strong directivity.

[0089] The superimposed signal without self-interference suppression was subjected to the same echo detection processing as above, with the signal-to-interference ratio decreasing by 5 dB each time. The results after self-interference suppression are compared, as shown in Table 1. This demonstrates that the self-interference suppression method proposed in this invention reduces the signal-to-interference ratio of the detectable echo from -30 dB to -55 dB, equivalent to more than doubling the range of the integrated detection and communication system.

[0090] Table 1 Comparison of echo detection results

[0091] Signal-to-noise ratio Before self-interference suppression Temporal domain first, then spatial domain detection method -30dB √ √ -35dB × √ -40dB × √ -45dB × √ -50dB × √ -55dB × √

[0092] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and modify the above embodiments within the scope of the present invention without departing from the principles and purpose of the present invention.

Claims

1. A self-interference signal suppression method based on transfer function fitting and phase matching, characterized by: The following steps are involved: Step 1: In a targetless scenario, collect the self-interference signal x1(t) received by the receiving end of the underwater sounding integrated system. The self-interference signal x1(t) is transmitted by the underwater sounding integrated system transmitting signal s(t) to the receiving array of the underwater sounding integrated system; Step 2: According to the formula Solve the transmit leakage transfer function H1(f) from x1(t), where X1(f) and S(f) are the frequency domain forms of x1(t) and s(t), respectively; Step 3: Construct the cancellation signal spectrum Y1(f) based on the transmit leakage transfer function H1(f) obtained in step 2: Y1(f)=S(f)*H1(f) Then perform IFFT on Y1(f) to obtain the time domain cancellation signal y1(t); Step 4: Collect the next frame signal x2(t) received by the receiving end of the underwater sounding integrated system, use the time domain cancellation signal y1(t), use the basic cross-correlation method to make a rough estimate of the time delay, and obtain a rough estimate of the time delay τ0 Step 5: Use the rough estimate of the delay τ0 obtained in step 4 Delay the received signal x2(t) accordingly to achieve coarse alignment: Where x'2(t) is x2(t) through The result obtained after the delay; further delay estimation is performed by the cross-spectral phase method to obtain the estimated value of the delay τ1 between x'2(t) and x1(t) Step 6: Based on the estimated value Perform time delay compensation on Y1(f) and reconstruct the cancellation signal: Self-interference suppression is achieved, and the residual signal after self-interference cancellation is obtained as: e(t)=x'2(t)-y2(t) Where y2(t) is the time domain signal obtained by IFFT of Y2(f).

2. The method for suppressing self-interference signals based on transfer function fitting and phase matching according to claim 1, characterized in that: The process of rough estimation of time delay using basic cross-correlation method in step 4 is as follows: Calculate the cross-correlation function R between y1(t) and x2(t) y1x2 (τ): Where X2(f) is the frequency domain form of x2(t), is the conjugate of X2(f); by calculating R y1x2 The maximum value of (τ) gives a rough estimate of the delay τ0 3. The self-interference signal suppression method based on transfer function fitting and phase matching according to claim 1, Its characteristics are: In step 5, the process of further delay estimation using the cross-spectral phase method is as follows: According to the formula calculate in is the conjugate of Y1(f), X'2(f) is the frequency domain form of x'2(t), and calculate The phase spectrum θ(f) is given by the formula The estimated value of the delay τ1 is obtained by using the CSP method where f i is the i-th frequency point of θ(f), and N is the number of frequency points of θ(f).

4. The method for suppressing self-interference signals based on transfer function fitting and phase matching according to claim 1, characterized in that: For multi-channel signals, transfer function construction and multi-stage delay estimation are performed on each channel to achieve self-interference suppression of each channel.

Images