A method for interference suppression in a strong interference background on a small-scale underwater acoustic array

Abstract

The application discloses a kind of interference suppression methods under strong interference background on small scale underwater acoustic array, comprising the following steps: (1) initialization array parameter and data processing parameter (including the scan angle and scanning interval of extended virtual array element number, DOA estimation);(2) based on the received data of actual array element, first virtual array element extension is carried out, and strong interference signal DOA is estimated;(3) based on the estimated value of strong interference signal DOA, strong interference signal suppression is carried out using beam forming;(4) based on the actual array element data after strong interference suppression obtained in the third step, second virtual array element extension is carried out, and target signal DOA is estimated;(5) based on target signal DOA estimation value, the time domain waveform signal of target is obtained.

Description

Technical Field

[0001] This invention relates to an interference suppression method for small-scale underwater acoustic arrays under strong interference backgrounds, belonging to the field of underwater acoustic target recognition technology. Background Technology

[0002] Underwater acoustic target identification or detection refers to the technology of classifying and identifying or detecting targets by processing the echo signals of active sonar or the target radiated noise of passive sonar. It is generally divided into two types: passive sonar noise identification, where passive sonar receives the noise radiated by the target and identifies and detects it based on the mapping relationship between the noise characteristics and the target type; and active sonar echo identification, where active sonar transmits a signal and identifies and detects it based on the mapping relationship between the characteristics of the received echo signal and the target type. Passive sonar is widely used due to its strong concealment and high security.

[0003] Currently, target detection and identification are considered as a single problem. The detection problem is essentially a question of presence or absence, i.e., detecting whether a target exists within a specified range, which can be viewed as a binary classification problem. The target identification problem is identifying the category or type of a target within a specified range, which is a multi-target classification problem.

[0004] In practice, the spatial characteristics of array signals can be used to estimate the azimuth of target signals, thereby enabling the identification of multiple targets. When processing a specific target signal, target signals from other directions are considered interference signals, and a low signal-to-interference ratio (SIR) can lead to a significant deviation in the DOA estimation result for that target. When using beamforming to suppress interference, the low estimation accuracy of small-scale arrays can significantly reduce the interference suppression effect. To address these issues, this invention proposes an interference suppression method based on two virtual array element expansions. The first expansion is used to estimate the DOA of strong interference signals to achieve strong interference suppression, while the second expansion is used to estimate the DOA of the target. Summary of the Invention

[0005] The purpose of this invention is to address the interference suppression problem in underwater acoustic target identification by providing an interference suppression method for small-scale underwater acoustic arrays under strong interference backgrounds. This method first uses a linear prediction algorithm to expand virtual array elements at both ends of the actual array and performs DOA estimation on the received signal of the expanded array to obtain an estimated DOA value for the strong interference signal. Based on beamforming estimation of the strong interference signal and combined with sensitivity cancellation coefficients to overcome the sensitivity differences between actual array elements, the strong interference signal in the actual array is suppressed. Then, the virtual array elements are expanded a second time to estimate the DOA of the target signal, thereby obtaining the time-domain waveform signal of the target.

[0006] To achieve the above objectives, the method employed in this invention is: an interference suppression method for a small-scale underwater acoustic array under strong interference background, comprising the following steps:

[0007] (1) Initialize array parameters and data processing parameters (including the number of extended virtual array elements, scanning angle and scanning interval);

[0008] (2) Based on the received data of the actual array elements, perform the first virtual array element expansion and estimate the strong interference signal DOA;

[0009] (3) Based on the estimated value of the strong interference signal DOA, beamforming is used to suppress the strong interference signal;

[0010] (4) Based on the actual array data after strong interference suppression obtained in step 3, perform a second virtual array element expansion to estimate the target signal DOA;

[0011] (5) Based on the DOA estimate of the target signal, obtain the time-domain waveform signal of the target.

[0012] As a further improvement to the present invention, step (1) initializes the following parameters:

[0013] (1.1) Set the actual number of array elements M and the array element spacing d;

[0014] (1.2) Set the number of extended virtual array elements N (N≤M), and the scanning angle [θ] estimated by DOA. l ,θ h ], scan interval Δθ.

[0015] As a further improvement of the present invention, step (2) includes the following specific steps:

[0016] Based on actual array data, a linear prediction algorithm is used to expand N / 2 virtual array elements forward and backward at both ends of the actual array for the first time to estimate the strong interference signal DOA.

[0017] (2.1) For forward expansion, the result of the Kth snapshot of the m-th element is expressed in vector form, denoted as x. m =[x m (1),x m (2),...,x m (K)] T Then the sequence of M-1 array elements starting from array element m can be written as X m =[x m ,x m+1 ,...,x m+M-2 ], m=1,2,...,N / 2;

[0018] (2.2) Let c be the coefficient of the linear prediction algorithm. m =[cm ,c m+1 ,...,c m+M-2 ] T The coefficients c of the linear prediction algorithm are obtained by using the received data from the (m+M-1)th array element, where m = 1, 2, ..., N / 2. m To make an estimate, i.e., to satisfy the equation X m c m =x m+M-1 ;

[0019] (2.3) Solve the equation X using the least squares method. m c m =x m+M-1 The linear prediction coefficients are obtained as follows: in For X m The pseudo-inverse matrix;

[0020] (2.4) Based on the calculated linear prediction coefficient c m The estimated received data for the (m+M)th array element is...

[0021] (2.5) For backward expansion, taking the array element m and the M-1 array elements before it, we have X' m =[x m-M+2 ,x m-M+3 ,...,x m ] and c' m =[c m-M+2 ,c m-M+3 ,...,c m ] T The expansion method is the same as the forward expansion, estimating the received data of the m-th array element as follows:

[0022] (2.6) After the first virtual array element expansion, DOA estimation is performed on the received signal of the expanded array to obtain the energy of each beam. The angle corresponding to the maximum beam energy is taken as the estimated value of DOA of the strong interference signal.

[0023] As a further improvement of the present invention, step (3) includes the following specific steps:

[0024] (3.1) Based on the angle of the strong interference signal estimated in step (2) and the actual array element received data x m (t), m=1,2,...,M, the strong interference signal estimated by beamforming is:

[0025] (3.2) Considering the propagation delay, the estimate of the strong interference signal received by the m-th array element can be written as:

[0026] (3.3) Based on the strong interference signal estimated in (3.1) Let its discrete sampling result be right Perform autocorrelation to obtain Where E[·] represents the mathematical expectation;

[0027] (3.4) x, the result of the Kth quickshot of the m-th array element m Cross-correlation of (k), k = 1, 2, ..., K yields...

[0028] (3.5) Calculate the sensitivity cancellation coefficient of each actual array element

[0029] (3.6) Considering the sensitivity cancellation coefficient, the received signal x of each array element m Subtract strong interference from (t) The result after strong interference suppression is:

[0030] As a further improvement of the present invention, step (4) includes the following specific steps:

[0031] (4.1) Based on the actual array data after strong interference suppression obtained in step (3), the same as in step (2), the linear prediction algorithm is used to extend N / 2 virtual array elements in the forward and backward directions at both ends of the array for the second time.

[0032] (4.2) After the second virtual array element expansion, DOA estimation is performed on the received signal of the expanded array to obtain the energy of each beam. The angle corresponding to the maximum beam energy is taken as the estimated value of the target signal DOA.

[0033] As a further improvement of the present invention, step (5) includes the following specific steps:

[0034] Based on the target signal angle estimated in step (4) Obtain the time-domain waveform signal of the target.

[0035] The purpose of this invention is to address the interference suppression problem in underwater acoustic target identification by providing an interference suppression method for small-scale underwater acoustic arrays under strong interference backgrounds. This method first uses a linear prediction algorithm to expand virtual array elements at both ends of the actual array and performs DOA estimation on the received signal of the expanded array to obtain an estimated DOA value for the strong interference signal. Based on beamforming estimation of the strong interference signal and combined with sensitivity cancellation coefficients to overcome the sensitivity differences between actual array elements, the strong interference signal in the actual array is suppressed. Then, the virtual array elements are expanded a second time to estimate the DOA of the target signal, thereby obtaining the time-domain waveform signal of the target.

[0036] Beneficial effects:

[0037] Compared with conventional beamforming, the method disclosed in this invention has the following advantages: 1. When there is strong interference signal, beamforming based on virtual array element expansion can more accurately estimate the azimuth of the interference signal; 2. After expanding the virtual array element, the signal-to-noise ratio gain and resolution of the array are improved, and the main lobe width of the target azimuth is narrowed, which is beneficial for estimating the incident direction of the target signal. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of a virtual array element.

[0039] Figure 2 The results show the DOA estimation of the CBF after the first expansion and the beamforming based on virtual array elements.

[0040] Figure 3 This is the sensitivity coefficient.

[0041] Figure 4 The DOA estimation results are obtained after a second expansion to suppress strong interference.

[0042] Figure 5 The output power spectra before and after strong interference suppression are shown, where (a) is the CBF output power spectrum before interference suppression, (b) is the CBF output power spectrum after interference suppression, and (c) is the output power spectrum of the extended 20 virtual array elements after interference suppression. Detailed Implementation

[0043] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments:

[0044] This invention first expands virtual array elements at both ends of a small-scale underwater acoustic array based on actual array data, using a linear prediction algorithm. Then, it obtains the direction of strong interference signals through DOA estimation, estimates the strong interference signals through beamforming, and introduces sensitivity cancellation coefficients to each actual array element to suppress the strong interference signals. Finally, it expands the virtual array elements a second time to estimate the DOA of the target signal, and finally obtains the time-domain waveform signal of the target.

[0045] Example 1:

[0046] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments.

[0047] An interference suppression method for a small-scale underwater acoustic array under strong interference background includes the following steps:

[0048] Step 1: Initialize signal parameters, array parameters, and data processing parameters, specifically as follows:

[0049] (1.1) Given a target signal and a strong interference signal, the direction of the target signal is set to α1 = 15°, the direction of the interference signal is set to β1 = -45°, the signal-to-interference ratio (SIR) is -20dB, the signal-to-noise ratio (SNR) is -10dB, the frequency of the target signal is f1 = 980Hz, the frequency of the interference signal is f1' = 1000Hz, and the sampling rate is f s =8000Hz, speed of sound c=1500m / s;

[0050] (1.2) Set the actual number of array elements M = 20, and the array element spacing is 1 / 10 of the wavelength corresponding to 1000Hz, that is, d = λ2 / 10 = c / f2 / 10 = 0.15m;

[0051] (1.3) Set the number of extended virtual array elements N = 20, and the scanning angle [θ] estimated by DOA. l ,θ h ] = [-90°, 90°], scan interval Δθ = 0.5°, and a Gaussian random model with variance of 0.01 is used to simulate the sensitivity coefficient of each array element.

[0052] Step 2 is as follows:

[0053] Based on the actual array data, a linear prediction algorithm is used to first extend 10 virtual array elements forward and backward at both ends of the actual array to estimate the strong interference signal DOA. The extension diagram is shown below. Figure 1 As shown.

[0054] (2.1) For forward expansion, the result of K = 8000 snapshots of the m-th element is expressed in vector form, denoted as x. m =[x m (1),x m (2),...,x m (8000)] T Then the sequence of 19 array elements starting from array element m can be written as X m =[x m ,x m+1 ,...,x m+18 ], m=1,2,...,10;

[0055] (2.2) Let c be the coefficient of the linear prediction algorithm. m =[c m ,c m+1 ,...,c m+18 ] T The coefficients c of the linear prediction algorithm are obtained by using the received data from the (m+19th)th array element. m To make an estimate, i.e., to satisfy the equation X m c m =x m+19 ;

[0056] (2.3) Solve the equation X using the least squares method. m c m =x m+19 The linear prediction coefficients are obtained as follows: in For X m The pseudo-inverse matrix;

[0057] (2.4) Based on the calculated linear prediction coefficient c m The estimated received data for the (m+20th)th array element is...

[0058] (2.5) For backward expansion, taking the array element m and the 19 array elements preceding it, we have X' m =[x m-18 ,x m-17 ,...,x m ] and c' m =[c m-18 ,c m-17 ,...,c m ] T The expansion method is the same as the forward expansion, estimating the received data of the (m-20th)th array element as follows:

[0059] (2.6) After the first expansion with 20 virtual array elements, DOA estimation is performed on the received signal of the expanded array to obtain the energy of each beam. The angle corresponding to the maximum beam energy is taken as the estimated DOA of the strong interference signal. like Figure 2 As shown, compared with conventional beamforming, when there is strong interference signal, beamforming based on virtual array elements cannot accurately estimate the target direction, but the azimuth of the interference signal can be estimated more accurately by introducing virtual array elements.

[0060] Step 3 specifically involves:

[0061] (3.1) Based on the angle of the strong interference signal estimated in step 2 and the actual array element received data x m(t), m=1,2,...,20, the strong interference signal estimated by beamforming is:

[0062] (3.2) Considering the propagation delay, the estimate of the strong interference signal received by the m-th array element can be written as:

[0063] (3.3) Based on the strong interference signal estimated in (3.1) Let its discrete sampling result be right Perform autocorrelation to obtain Where E[·] represents the mathematical expectation;

[0064] (3.4) x, the result of 8000 snapshots of the m-th array element m Cross-correlation of (k), k=1,2,...,8000 yields...

[0065] (3.5) Calculate the sensitivity cancellation coefficient of each actual array element Figure 3 A comparison curve between the actual sensitivity coefficient and its estimated value in the experiment is given. It can be seen from the figure that it is feasible to use the formula to calculate the sensitivity cancellation coefficient.

[0066] (3.6) Considering the sensitivity cancellation coefficient, the received signal x of each array element m Subtract strong interference from (t) The result after strong interference suppression is:

[0067] Step 4 is as follows:

[0068] (4.1) Based on the actual array data after strong interference suppression obtained in step 3, the same as in step 2, the linear prediction algorithm is used to extend 10 virtual array elements forward and backward at both ends of the array for the second time.

[0069] (4.2) After the second virtual array element expansion, DOA estimation is performed on the received signal of the expanded array to obtain the energy of each beam. The angle corresponding to the maximum beam energy is taken as the estimated DOA of the target signal. Figure 4 The DOA estimation results are shown for conventional beamforming and virtual element-based beamforming. Both methods exhibit nulls in the -45° interference direction. Compared to CBF, virtual element-based beamforming has a narrower main lobe width in the target direction, and the array resolution and signal-to-noise ratio gain are also improved. Furthermore, due to the introduction of a sensitivity cancellation coefficient, interference signals can still be effectively suppressed even when there are sensitivity differences among the elements.

[0070] Step 5 specifically involves:

[0071] Based on the target signal angle estimated in step 4 Obtain the time-domain waveform signal of the target.

[0072] The power spectrum diagrams obtained by beamforming the target signal direction before and after interference suppression are shown below. Figure 5 The figure shows that before the interference signal was suppressed, the power spectrum of the array output still contained a 1000Hz interference component; however, after the interference was suppressed, the power spectra obtained by CBF and beamforming based on virtual array elements both contained only the 980Hz target signal, indicating that the interference signal had been suppressed at this time, further verifying the effectiveness of the method.

[0073] The above embodiments show that the signal-to-noise ratio gain and resolution of the array are improved after expanding the virtual array elements, and the incident direction of strong interference signals can be estimated more accurately. Furthermore, the influence of the sensitivity difference of each array element is eliminated by introducing a sensitivity cancellation coefficient, and the interference signal is effectively suppressed in the end.

[0074] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any other way. Any modifications or equivalent changes made based on the technical essence of the present invention shall still fall within the scope of protection claimed by the present invention.

Claims

1. A method for suppressing interference under strong interference background on a small-scale underwater acoustic array, characterized in that, Includes the following steps: (1) Initialize array parameters and data processing parameters, including the number of extended virtual array elements, DOA estimated scanning angle and scanning interval; (2) Based on the received data of the actual array elements, perform the first virtual array element expansion and estimate the strong interference signal DOA; (3) Based on the estimated value of the strong interference signal DOA, beamforming is used to suppress the strong interference signal; (4) Based on the actual array data after strong interference suppression obtained in step (3), perform a second virtual array element expansion to estimate the target signal DOA; (5) Based on the estimated DOA of the target signal, obtain the time-domain waveform signal of the target; Step (2) includes the following specific steps: Based on the actual array data, a linear prediction algorithm is used to expand N / 2 virtual array elements forward and backward at both ends of the actual array for the first time to estimate the strong interference signal DOA; (2.1) For forward expansion, the result of the Kth snapshot of the m-th element is expressed in vector form, denoted as Then, the sequence of M-1 array data elements starting from array element m is written as: ; (2.2) Let the coefficients of the linear prediction algorithm be . The coefficients of the linear prediction algorithm are obtained by using the received data from the (m+M-1)th array element, where m = 1, 2, ..., N / 2. To make an estimate, that is, to satisfy the equation ; (2.3) Solve the equation using the least squares method The linear prediction coefficients are obtained as follows: ,in for The pseudo-inverse matrix; (2.4) Based on the calculated linear prediction coefficients The estimated received data for the (m+M)th array element is... ; (2.5) For backward expansion, taking the array element m and the M-1 array elements preceding it, we have and The expansion method is the same as the forward expansion, estimating the received data of the m-th array element as follows: ; (2.6) After the first virtual array element expansion, DOA estimation is performed on the received signal of the expanded array to obtain the energy of each beam. The angle corresponding to the maximum beam energy is taken as the estimated value of DOA of the strong interference signal. ; Step (3) includes the following specific steps: (3.1) Based on the angle of the strong interference signal estimated in step (2) and the actual array element received data The strong interference signal is estimated by beamforming. ; (3.2) Considering the propagation delay, the estimate of the strong interference signal received by the m-th array element is written as: ; (3.3) Based on the strong interference signal estimated in (3.1) Let its discrete sampling result be ,right Perform autocorrelation to obtain ,in This indicates the calculation of the expected value. (3.4) The result of the Kth quickshot with the m-th array element Perform cross-correlation to obtain ; (3.5) Calculate the sensitivity cancellation coefficient of each actual array element ; (3.6) Considering the sensitivity cancellation coefficient, the received signal of each array element Subtract strong interference The result after strong interference suppression is: .

2. The interference suppression method for a small-scale underwater acoustic array under strong interference background according to claim 1, characterized in that, Step (1) Initialize the following parameters: (1.1) Set the actual number of array elements M and the array element spacing d; (1.2) Set the number of extended virtual array elements N, N≤M, and the scanning angle estimated by DOA. Scan interval .

3. The interference suppression method for a small-scale underwater acoustic array under strong interference background according to claim 1, characterized in that, Step (4) includes the following specific steps: (4.1) Based on the actual array data after strong interference suppression obtained in step (3), the same as in step (2), the linear prediction algorithm is used to extend N / 2 virtual array elements in the forward and backward directions at both ends of the array for the second time; (4.2) After the second virtual array element expansion, DOA estimation is performed on the received signal of the expanded array to obtain the energy of each beam. The angle corresponding to the maximum beam energy is taken as the estimated value of the DOA of the target signal. .

4. The interference suppression method for a small-scale underwater acoustic array under strong interference background according to claim 1, characterized in that, Step (5) includes the following specific steps: based on the target signal angle estimated in step (4) To obtain the time-domain waveform signal of the target.

Images