Method for extracting line spectrum of anti-platform jamming pulse signal of far-field target airspace characteristic

Abstract

The application discloses a kind of based on far-field target airspace characteristic anti-platform interference pulse signal detection line spectrum extraction method, it can be used to the non-cooperative underwater acoustic pulse signal interception detection of underwater small platform carried passive sonar.For the short-time power spectrum of each wave beam data obtained by far-field beam forming to the sonar array signal collected;Extract line spectrum;Find the output power of signal in each wave beam data for line spectrum frequency point, obtain signal wave direction θ0 by finding the peak point of the normalized directivity of sonar receiving signal;According to beam forming algorithm and signal wave direction θ0, the theoretical normalized directivity of far-field signal is obtained;Absolute difference value is obtained for the normalized directivity of receiving signal and theoretical directivity;Difference interval [0,1] is divided into n sections, and the distribution ratio p of difference value in each interval is counted i ;Difference entropy H is obtained;If H is greater than threshold, the extracted line spectrum is judged as platform interference and removed;If it is less than threshold, the extracted line spectrum is regarded as far-field signal and retained.

Description

Technical Field

[0001] This invention relates to a method for detecting and extracting the line spectrum of anti-platform interference pulse signals based on the spatial characteristics of far-field targets, belonging to the field of underwater acoustic signal processing technology. Background Technology

[0002] Non-cooperative underwater acoustic pulse signal interception and detection has important applications in underwater target detection and underwater biological signal capture. Traditional methods for non-cooperative underwater acoustic pulse signal interception and detection include energy detection, time-frequency analysis, and spectral entropy detection. Among these, time-frequency analysis analyzes the signal from the perspectives of time and frequency, facilitating real-time monitoring and parameter estimation, and is widely used in the detection and parameter estimation of CW and LFM pulse signals.

[0003] In recent years, with the development of unmanned and intelligent underwater equipment, the research and application of underwater unmanned platforms have increased significantly. When underwater unmanned platforms are equipped with passive sonar for underwater operations, electromagnetic interference from the platform's own motors, as well as flow noise, mechanical noise, and propeller noise during platform navigation, result in a large number of platform interference lines appearing in the passive sonar detection frequency band. Traditional time-frequency analysis methods cannot effectively identify platform interference and pulse signals. This leads to a high signal-to-noise ratio of platform interference, making it easy to misinterpret the platform interference lines as signals, increasing the false alarm rate and affecting the detection performance of pulse signals. Summary of the Invention

[0004] Objective: To address the issue of numerous platform interference lines appearing in the detection frequency band of passive sonar on small underwater platforms during non-cooperative underwater acoustic pulse signal detection, due to the platform's own influence. Traditional pulse signal detection methods have poor ability to identify platform interference, and when the signal-to-noise ratio of platform interference is high, it may be misjudged as a signal, leading to an increased false alarm rate. This invention provides a platform interference-resistant pulse signal detection line spectrum extraction method based on the spatial characteristics of far-field targets.

[0005] Since pulse signals are typically located in the far field of the array, they conform to far-field characteristics. Platform interference, such as mechanical noise and propagator noise, propagates through the hull and water, exhibiting randomness and uncertainty in its propagation patterns, thus exhibiting non-far-field propagation characteristics. The spatial characteristics of a signal can be reflected by its beam normalization directivity. Furthermore, since entropy is a measure of the uncertainty of a random variable, when the normalized directivity of the actual signal and the theoretical far-field signal show the same trend, their difference distribution is relatively concentrated, resulting in a small difference entropy. However, when the actual signal is platform interference, its normalized directivity exhibits a degree of randomness, and the difference distribution between it and the theoretical far-field normalized directivity is more random, resulting in a larger difference entropy. Using the magnitude of the difference entropy to perform secondary discrimination on the line spectrum signal extracted by the time-frequency analysis method can distinguish between platform interference and pulse signals, thereby eliminating platform interference line spectra and reducing the impact of platform interference on pulse signal detection.

[0006] Technical Solution: A method for extracting the line spectrum of a pulse signal detection signal based on the spatial characteristics of far-field targets and resisting platform interference. This method is based on the far-field characteristics of the pulse signal and the non-far-field characteristics of platform interference. It determines whether the signal conforms to the far-field characteristics by calculating the entropy difference between the normalized directivity of the actual signal and the normalized directivity of the theoretical far-field signal. This enables the identification of the pulse signal and platform interference, removes the interference line spectrum, and retains the pulse signal line spectrum, thereby achieving the purpose of resisting platform interference and improving the anti-platform interference capability of traditional pulse signal detection methods.

[0007] Includes the following steps:

[0008] Step S1: Perform far-field beamforming on the acquired sonar array signals to obtain multi-beam data;

[0009] Step S2: Calculate the short-time power spectrum of each beam data;

[0010] Step S3: Extract the line spectrum according to the traditional non-cooperative pulse signal detection method;

[0011] Step S4: Find the output power of the signal at the extracted line spectrum frequency point in each beam data, and calculate the normalized directivity of the sonar received signal at that frequency point.

[0012] Step S5: Find the peak point by normalized directivity to obtain the azimuth of the incoming signal θ0;

[0013] Step S6: Based on the beamforming algorithm used in step S1 and the signal azimuth θ0 in step S5, the theoretical normalized directivity of the far-field signal is obtained.

[0014] Step S7: Calculate the absolute difference between the normalized directivity and the theoretical directivity of the received signal;

[0015] Step S8: Divide the difference interval [0, 1] into n segments, and calculate the distribution ratio p of the difference in each interval. i (i = 1, 2, ..., n);

[0016] Step S9, calculate the difference entropy H;

[0017] Step S10: Set a decision threshold. If H is greater than the threshold, the extracted line spectrum is judged as platform interference and discarded; if it is less than the threshold, the extracted line spectrum is regarded as a far-field signal and retained.

[0018] In step S1, the far-field beamforming employs a delay-addition beamforming algorithm. By calculating and compensating for the propagation delay of each array element in the desired direction, the array output can be in phase superimposed on the incoming wave signal in the desired direction, thereby generating a main lobe beam in that direction and a smaller response in other directions. The specific process is as follows:

[0019] Assume there are M array elements, a reference point O, and a signal direction θ0. When the received signal at the reference point is s0(n), the received signal at the i-th array element is s. i (n)=s0(n+τ i (θ0)), where τ i (θ0) is the time difference between the signal received by the i-th element and the signal received by the reference point. If the beamforming azimuth is θ j Then each array element needs to be compensated for a delay τ. i (θ j The array output after delay-addition beamforming is:

[0020]

[0021] In step S2, the short-time power spectrum is estimated using the periodogram method based on the short-time Fourier transform of the beam data. The specific process is as follows:

[0022]

[0023]

[0024] Where n is a discrete time series, W N (n) is the sliding window function, N is the window length, l is the sliding window step, m represents the number of sliding window movements, k corresponds to the frequency point, and S N (θ j (k, m) represents the relationship between θ j Short-time spectrum obtained from directional beamforming data, P(θ) j , k, m) is the short-time power spectrum obtained from the corresponding data.

[0025] In step S4, the line spectrum frequency point is extracted, and the normalized directivity of the sonar received signal at that frequency point is calculated. This is obtained from the short-time power spectrum of the signal at that frequency point in each beam data. Assuming the line spectrum frequency point extracted by the traditional pulse signal detection method is k, its normalized directivity calculation formula is:

[0026]

[0027] Where θ is the beamforming azimuth set, p(θ) j (k, m) represents the relationship between θ j The short-time power spectrum of azimuth beamforming data at the k-th frequency point in the m-th window.

[0028] In step S6, the theoretical normalized directivity calculation formula for the M-element array is:

[0029]

[0030] Depending on the array configuration, the above equation can be simplified. For an M-element uniform linear array, the element spacing is d, and the sampling frequency is f. s The sliding window has a length of N, the signal frequency is k, and the corresponding signal wavelength is λ = cN / (kf). s ), where c is the speed of sound. When the incident direction of the signal is θ0, its theoretical directivity formula is:

[0031]

[0032] When d / λ << 1, the linear array can be considered as a continuous linear array, with L = Md representing the array length. Its theoretical directivity formula is:

[0033]

[0034] For a circular array of M uniformly distributed elements with radius r, the angle between two adjacent elements is α = 2π / M, the signal incident direction is θ0, and the beamforming direction is θ. j When the signal frequency is k and the sliding window length is N, the corresponding frequency is f = kf. s / N, its theoretical orientation formula is:

[0035]

[0036] Where, Δ i (θ j )=τ i (θ j )-τ i (θ0), τ i (θ j The formula for calculating ) is:

[0037]

[0038] The formula for calculating the absolute difference in step S7 is as follows:

[0039] in, For beamforming azimuth set.

[0040] In step S8, the distribution ratio of the absolute difference in each interval is calculated using a histogram statistical method. First, the absolute difference is statistically analyzed. The number of distributions x in each of the n intervals. i (i = 1, 2, ..., n), then divide by the total number of absolute differences, i.e., the number of azimuths formed by the beam, K, to obtain the distribution ratio of the absolute differences in each interval:

[0041] p i =x i / K(i=1,2,…,n)

[0042] in,

[0043] In step S9, the difference entropy calculation process is as follows:

[0044]

[0045] Where a is the base, when p i When p = 0, i log a p i =0.

[0046] In step S10, a decision threshold is selected from a value within the range of zero to the minimum value of the difference entropy when only interference exists. The median of the range can be selected, and the magnitude of the difference entropy H is compared with the decision threshold to identify the source of the line spectrum signal extracted by the traditional method. When the difference entropy is large, it indicates that the distribution range of the difference between the actual data directivity and the theoretical beam directivity is relatively dispersed, the actual data directivity deviates from the theoretical beam directivity, and the difference distribution is relatively random. At this time, the line spectrum signal does not conform to the far-field characteristics and is judged as platform interference, and the line spectrum is discarded. When the difference entropy is small, the distribution of the difference between the actual beam directivity and the theoretical beam directivity is relatively concentrated, and the overall distribution trend of the actual beam directivity is similar to that of the theoretical directivity. The extracted line spectrum can be regarded as a far-field pulse signal, and the line spectrum can be retained.

[0047] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the line spectrum extraction method for anti-platform interference pulse signal detection based on the spatial characteristics of far-field targets as described above.

[0048] A computer-readable storage medium storing a computer program that performs the line spectrum extraction method for anti-platform interference pulse signal detection based on far-field target spatial characteristics as described above.

[0049] Beneficial effects: Compared with the prior art, the anti-platform interference pulse signal detection line spectrum extraction method based on far-field target spatial characteristics provided by the present invention can improve the anti-platform interference capability of traditional pulse signal detection methods. When platform interference exists, it can effectively identify platform interference and pulse signals, thereby eliminating platform interference line spectra, reducing the impact of platform interference on pulse signal detection, and reducing the false alarm rate. Attached Figure Description

[0050] Figure 1 This is a schematic diagram of the overall process of the method according to an embodiment of the present invention;

[0051] Figure 2 This is a schematic diagram of the short-time power spectrum of beamforming data in the 0° direction when only platform interference exists in an embodiment of the present invention;

[0052] Figure 3 This is a comparison chart of the normalized directivity of the 40kHz interference line spectrum beam and the theoretical far-field normalized directivity when only platform interference exists in this embodiment of the invention.

[0053] Figure 4 This is a comparison diagram of the normalized directivity of the 50kHz pulse signal beam and the theoretical far-field normalized directivity in an embodiment of the present invention.

[0054] Figure 5 This is a graph showing the entropy difference of platform interference at 40kHz when only platform interference exists in an embodiment of the present invention.

[0055] Figure 6 This is a graph showing the entropy difference of a 50kHz pulse signal in an embodiment of the present invention.

[0056] Figure 7 This is a graph showing the entropy difference of the 40kHz line spectrum when a pulse signal exists in the first 100ms of an embodiment of the present invention. Detailed Implementation

[0057] The present invention will be further illustrated below with reference to specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. After reading the present invention, any modifications of the present invention in various equivalent forms by those skilled in the art will fall within the scope defined by the appended claims.

[0058] This invention discloses a method for detecting line spectrum of anti-platform interference pulse signals based on the spatial characteristics of far-field targets. The overall process of the method is as follows: Figure 1As shown, this method calculates the entropy difference between the actual normalized directivity and the theoretical far-field normalized directivity of the extracted line spectrum signal, and compares it with a set decision threshold to determine whether the line spectrum signal conforms to far-field characteristics. Based on this, it distinguishes between platform interference and pulse signals, eliminating platform interference line spectra above the threshold and retaining pulse signal line spectra below the threshold, thereby achieving anti-platform interference and improving the problem of high false alarm rates in traditional pulse signal detection methods caused by platform interference.

[0059] This example uses lake test data acquired by a 21-element uniform linear array mounted on an underwater unmanned platform for method verification. The element spacing d is 5.6 mm, the signal sampling frequency is 500 kHz, and the operating frequency band is 30 kHz-130 kHz. Pre-formed beamforming is performed in 25 azimuths within the range of [-90°, 90°] using isocosine selection. Short-time Fourier transform is applied to the 0° beam data... Figure 2 It can be seen that in the absence of a pulse signal, multiple plateau interference lines exist within its detection frequency band. Taking the 40kHz interference line spectrum as an example, its normalized directivity is compared with the theoretical far-field beam normalized directivity. Figure 3 It can be seen that there is a significant difference between the two, and the spatial characteristics of the interference line spectrum do not conform to the far-field propagation characteristics; a segment of data containing a 50kHz pulse signal is selected for beam normalization directivity comparison, by Figure 4 It can be seen that although the sidelobes of the actual data's directivity are affected by interference and noise, resulting in some discrepancies with the theoretical directivity, the overall trend conforms to the theoretical far-field normalized directivity. The 40kHz line spectrum with only plateau interference, the 50kHz line spectrum with a 50kHz pulse signal, and the first 100m of the spectrum were analyzed respectively. s The difference entropy of the 40kHz line spectrum in the presence of a 40kHz pulse signal is calculated, and the results are as follows: Figure 5 , Figure 6 and Figure 7 As shown, the difference entropy is larger when only platform interference exists, while it is smaller when pulse signals are present. For signals of the same frequency and interference, the difference entropy can be effectively distinguished by setting an appropriate threshold and utilizing the magnitude of the difference entropy.

[0060] The specific steps of this invention are as follows:

[0061] (1) Far-field beamforming is performed on the acquired sonar array signals to obtain multi-beam data. In this example, 25 beamforming azimuths are selected with equal cosine within the range of [-90°, 90°]. The time difference of the received signals of the array elements in each azimuth is calculated, and the multi-beam data is obtained by summing the data of each array element after delay compensation. For a uniform linear array, it is assumed that the angle between the beamforming azimuth and the normal direction of the array is θ. j The array elements are numbered 1, 2, ..., 21. Taking array element 1 as the reference element, the delay τ of each array element is...i The calculation formula is as follows:

[0062] τ i (θ j )=(i-1)d sinθ j / c

[0063] (2) Calculate the short-time power spectrum of each beam data. Apply a sliding window to each beam data and perform a short-time Fourier transform. Then, use the periodogram method to calculate the power spectrum. In this example, the sliding window length is 2048 points and the step size is 1024 points.

[0064] (3) Line spectrum is extracted based on the traditional non-cooperative pulse signal detection method. Since the signal-to-noise ratio of the platform interference line spectrum is high, platform interference line spectra such as 40kHz, 77.64kHz, and 117.4kHz will be extracted during the line spectrum extraction. In this example, the 40kHz platform interference line spectrum is used as an example for analysis and verification.

[0065] (4) Find the output power of the extracted line spectrum frequency point in each beam data, and calculate the normalized directivity of the sonar received signal at that frequency point. Assuming the line spectrum frequency point is k, its normalized directivity calculation formula is:

[0066]

[0067] in, This is the beamforming azimuth set; in this example, the set contains 25 azimuths, p(θ) j (k, m) represents the relationship between θ j The short-time power spectrum of azimuth beamforming data at the k-th frequency point in the m-th window.

[0068] (5) Find the peak point by normalized directivity and obtain the azimuth of the incoming signal θ0.

[0069] (6) Based on the beamforming algorithm used in step (1) and the signal azimuth θ0 from step (5), the theoretical normalized directivity of the far-field signal is obtained. In this example, the array is a uniform linear array, and the beamforming algorithm in step (1) is a delayed-addition conventional beamforming algorithm. Also, since d / λ << 1 in this example, the array can be regarded as a continuous linear array, and its directivity formula adopts the following form:

[0070]

[0071] Where L represents the array length, k is the line spectrum frequency, and f s Where is the sampling frequency, N is the sliding window length, and c is the speed of sound in water.

[0072] (7) The absolute difference between the normalized directivity and the theoretical directivity of the received signal is calculated using the following formula:

[0073] Dd (θ j ,k,m)=abs(D s (θ j ,k,m)-D t (θ j ,k)),j=1,2,…,25

[0074] (8) The average difference interval [0, 1] is n segments. Calculate the distribution ratio p of the differences in each interval. i (i = 1, 2, ..., n), in this example n is 5, and the number of wavenumbers is 25. First, calculate the difference. The number of distributions x in each interval i Then, dividing by the total number of differences, i.e., the number of beams K, yields the distribution ratio of the differences in each interval:

[0075] p i =x i / K(i=1,2,…,n)

[0076] in,

[0077] (9) Calculate the difference entropy H. The calculation process for the difference entropy is as follows:

[0078]

[0079] Where a is the base, in this example a is 10, and it is stipulated that when p i When p = 0, i log a p i =0.

[0080] (10) Set a decision threshold and compare the difference entropy with the decision threshold. If H is greater than the threshold, the extracted line spectrum is judged as platform interference and is removed. If it is less than the threshold, the extracted line spectrum is regarded as far-field signal and is retained. In this example, the decision threshold is set to 0.2.

[0081] A method for extracting the line spectrum of a pulse signal detection system based on the spatial characteristics of far-field targets, designed for use by passive sonar mounted on small underwater platforms to intercept and detect non-cooperative underwater acoustic pulse signals. This method identifies both the extracted line spectrum and platform interference by determining whether the line spectrum obtained by traditional pulse signal detection methods conforms to far-field characteristics in the spatial domain. Platform interference line spectra are then eliminated to reduce their impact on pulse signal detection, lower the false alarm rate, and improve the anti-platform interference capability of traditional pulse signal detection.

[0082] Obviously, those skilled in the art should understand that the steps of the anti-platform interference pulse signal detection line spectrum extraction method based on far-field target spatial characteristics in the above embodiments of the present invention can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by the computing device. Furthermore, in some cases, the steps shown or described can be performed in a different order than presented here, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the embodiments of the present invention are not limited to any particular hardware and software combination.

Claims

1. A method for line spectrum extraction of anti-platform interference pulse signals based on the spatial characteristics of far-field targets, characterized in that, Includes the following steps: Step S1: Perform far-field beamforming on the acquired sonar array signals to obtain multi-beam data; Step S2: Calculate the short-time power spectrum of each beam data; Step S3: Extract the line spectrum based on the non-cooperative pulse signal detection method; Step S4: Find the output power of the signal at the extracted line spectrum frequency point in each beam data, and calculate the normalized directivity of the sonar received signal at that frequency point. Step S5: Find the peak point using normalized directivity to obtain the signal arrival direction. ; Step S6: Determine the beam azimuth based on the beamforming algorithm used in step S1 and the signal from step S5. The theoretically normalized directivity of the far-field signal is obtained; Step S7: Calculate the absolute difference between the normalized directivity and the theoretical directivity of the received signal; Step S8: Divide the difference into n equal segments and calculate the distribution ratio of the difference in each interval. ; Step S9, calculate the difference entropy. ; Step S10: Set the decision threshold; if If the value is greater than the threshold, the extracted line spectrum is judged as platform interference and discarded; if the value is less than the threshold, the extracted line spectrum is regarded as a far-field signal and retained. In step S1, the far-field beamforming employs a delay-addition beamforming algorithm. By calculating and compensating for the propagation delay of each array element in the desired direction, the array output is in phase with the incoming wave signal in the desired direction, thereby generating a main lobe beam in that direction and a smaller response in other directions. The specific process is as follows: There are M array elements, and the reference point is... azimuth of incoming signal wave When the received signal at the reference point is At that time, the first Each element receives the signal as follows: ,in For the first The time difference between the signal received by each array element and the signal received by the reference point; if the beamforming azimuth is Then each array element needs to be compensated for the delay. The array output after delay-addition beamforming is: ； In step S6, the theoretical normalized directivity calculation formula for the M-element array is: Depending on the formation, the above formula can be simplified; for A uniform linear array of elements with an element spacing of 1. The sampling frequency is The sliding window has a length of N and a signal frequency of N. The corresponding signal wavelength is ,in The speed of sound; when the incident direction of the signal is Its theoretical orientation formula is: when At this time, the linear array can be regarded as a continuous linear array, using The theoretical directional formula for the matrix length is: right A circular array with uniformly distributed elements and a radius of . The included angle between two adjacent primitives The signal incident direction is The beamforming direction is The signal frequency is When the sliding window length is N, the corresponding frequency is Its theoretical orientation formula is: in, , The calculation formula is: 。 2. The method for detecting and extracting the line spectrum of a pulse signal with anti-platform interference based on the spatial characteristics of a far-field target according to claim 1, characterized in that, In step S2, the short-time power spectrum is estimated using the periodogram method based on the short-time Fourier transform of the beam data. The specific process is as follows: in, It is a discrete time series. For sliding window functions, For the length of the sliding window, For sliding window stepping, This represents the number of times the sliding window has been moved. Corresponding frequency points To Short-time spectrum obtained from directional beamforming data The short-time power spectrum is obtained from the corresponding data.

3. The method for line spectrum extraction of anti-platform interference pulse signal detection based on far-field target spatial characteristics according to claim 2, characterized in that, In step S4, the line spectrum frequency point is extracted, and the normalized directivity of the sonar received signal at that frequency point is calculated. This is obtained from the short-time power spectrum of the signal at that frequency point in each beam data. Let the line spectrum frequency point extracted by the pulse signal detection method be... Its normalized orientation calculation formula is: in, For beamforming azimuth set, To Azimuth beamforming data in the first The window The short-time power spectrum at each frequency point.

4. The method for detecting and extracting the line spectrum of anti-platform interference pulse signals based on the spatial characteristics of far-field targets according to claim 3, characterized in that, The formula for calculating the absolute difference in step S7 is as follows: 。 5. The method for detecting and extracting the line spectrum of anti-platform interference pulse signals based on the spatial characteristics of far-field targets according to claim 4, characterized in that, In step S8, the distribution ratio of the absolute difference in each interval is calculated using a histogram statistical method. First, the absolute difference is statistically analyzed. The number of distributions in the n intervals. Divide by the total number of absolute differences to get the number of azimuths formed by the beam. This allows us to obtain the distribution ratio of the absolute difference across each interval: in, .

6. The method for detecting and extracting the line spectrum of a pulse signal with anti-platform interference based on the spatial characteristics of a far-field target according to claim 5, characterized in that, In step S9, the difference entropy calculation process is as follows: in, With base, when hour, .

7. A computer device, characterized in that: The computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the anti-platform interference pulse signal detection line spectrum extraction method based on the spatial characteristics of far-field targets as described in any one of claims 1-6.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program that performs the line spectrum extraction method for anti-platform interference pulse signal detection based on the spatial characteristics of far-field targets as described in any one of claims 1-6.

Images