A method for echo detection of underwater targets in a strong reverberation background

Abstract

The present application relates to the technical field of underwater acoustic signal processing, and particularly relates to a method for detecting underwater target echo in a strong reverberation background, comprising the following steps: performing band-pass filtering on received sonar echo signals to obtain filtered signals; performing sliding window segmentation on the filtered signals to form a plurality of short-time signal segments; extracting the envelope of the short-time signal segments and performing normalization processing on the envelope; performing differential processing on the normalized envelope to obtain a differential envelope signal; performing frequency spectrum transformation and frequency weighting processing on the differential envelope signal to obtain a weighted envelope spectrum; performing feature extraction on the weighted envelope spectrum, calculating the spectral entropy and spectral variance, and constructing a spectral variance-entropy ratio; and determining whether a target echo exists. The method can reduce the false alarm rate while maintaining a high detection probability, and avoids the performance degradation of existing matched filtering methods in a strong reverberation condition.

Description

Technical Field

[0001] This invention relates to the field of underwater acoustic signal processing technology, specifically to a method for detecting underwater target echoes under strong reverberation backgrounds, which can be applied to target detection and identification in active sonar systems. Background Technology

[0002] In active sonar systems, target echo detection is often hampered by strong reverberation. Reverberation primarily originates from the seabed, air-water interfaces, water column inhomogeneities, and marine life, and its formation mechanism is similar to that of target echoes, exhibiting a high correlation in the time and frequency domains. This correlation leads to target echoes being easily submerged under low signal-to-reverberation ratio (SRR) conditions, significantly reducing detection performance.

[0003] In existing technologies, matched filters (MF) are a commonly used detection method. However, in low signal-to-mixing environments, the output of MF often contains clutter peaks with amplitudes close to or even exceeding those of the target echo, making it difficult for the amplitude threshold method to distinguish between the target and the reverberant signal.

[0004] To suppress reverberation interference, researchers both domestically and internationally have proposed various methods. For example, narrow-beam spatial filtering improves spatial resolution by increasing the transmit bandwidth and receiver array aperture, which can reduce the impact of reverberation to some extent. However, it also alters the statistical characteristics of the reverberation envelope, causing it to deviate from the Rayleigh distribution, leading to a decline in the performance of detectors based on the Gaussian assumption. Furthermore, the statistical whitening method uses a reverberation covariance model to discorrelate the interference, which is effective when the reverberation has a significant Doppler frequency shift. However, its performance deteriorates significantly when using FM signals or in the context of strong seabed reverberation. Other methods, such as time-domain reversal mirrors (TRM) and fractional Fourier transform (FRFT) based filtering techniques, have also been proposed. TRM relies on a vertical receiver array, making its engineering implementation complex; while the FRFT method can suppress reverberation to some extent while focusing energy, it is sensitive to signal conditions. Doppler-based detection uses the frequency shift generated by target motion to distinguish reverberation from echo, but its reliability decreases significantly when spatial resolution decreases.

[0005] Existing suppression methods based on spatial filtering, Doppler frequency shift, or covariance whitening have limitations, insufficient stability, and complex engineering implementation under low signal reverberation ratio conditions. Summary of the Invention

[0006] This invention addresses the difficulties in target echo detection under strong reverberation environments and the high false alarm rate and degraded detection performance of traditional matched filtering methods due to high-amplitude clutter interference. It proposes a method for underwater target echo detection in strong reverberation conditions. This method extracts the envelope of the received signal and constructs its envelope spectrum features, utilizing the difference in envelope fluctuation characteristics between the target echo and the reverberation to achieve robust detection under low signal-to-reverberation ratio conditions. The technical means employed in this invention are as follows:

[0007] A method for detecting underwater target echoes under strong reverberation backgrounds, specifically including the following steps:

[0008] The received sonar echo signal is bandpass filtered to obtain the filtered signal;

[0009] The filtered signal is segmented using a sliding window to form multiple short-time signal segments;

[0010] Extract the envelope of the short-time signal segment and normalize the envelope;

[0011] The differential envelope signal is obtained by performing differential processing on the normalized envelope;

[0012] The differential envelope signal is subjected to spectral transformation and frequency weighting to obtain the weighted envelope spectrum;

[0013] Feature extraction is performed on the weighted envelope spectrum, spectral entropy and spectral variance are calculated, and the spectral variance-entropy ratio is constructed.

[0014] An adaptive threshold is set based on local reverberation statistics. When the detected spectral variance-entropy ratio is lower than the adaptive threshold, it is determined that a target echo exists. When the detected spectral variance-entropy ratio is higher than the adaptive threshold, it is determined that no target echo exists.

[0015] Furthermore, the normalized envelope is differentially processed using a first-order difference method, specifically as follows:

[0016]

[0017] This represents the first-order difference result of the normalized envelope signal, used to reflect the rate of change of the envelope over time. Indicates the first The signal segment in the first Normalized envelope amplitude at each sampling point Indicates the first The signal segment in the first Normalized envelope amplitude at each sampling point. Represents the discrete sampling index of the envelope sequence. Indicates the first The numbering of the signal segment.

[0018] Furthermore, the spectral variance-entropy ratio is constructed based on the spectral entropy and spectral variance:

[0019]

[0020] in, Indicates the first The spectral variance-entropy ratio of a segment signal is used to represent the energy distribution characteristics of the envelope spectrum in the frequency domain. The spectral entropy of the envelope spectrum reflects the uniformity of the spectral energy distribution. is the spectral variance of the envelope spectrum, representing the intensity of fluctuations in the spectral amplitude.

[0021] Furthermore, the envelope is normalized using the root mean square amplitude method.

[0022] Furthermore, frequency indexing is used to perform frequency weighting on the differential envelope signal, increasing the role of high-frequency components in the weighted envelope spectrum.

[0023] Furthermore, an adaptive threshold is dynamically set based on the statistical characteristics of local reverberation samples to adapt to the detection requirements under different signal reverberation ratio conditions.

[0024] This invention discloses a method for detecting underwater target echoes under strong reverberation backgrounds. This method involves envelope extraction and differentiation of the received signal, combined with spectral transformation to construct the spectral variance-entropy ratio. This method employs an adaptive threshold for decision-making, achieving effective detection of target echoes under low signal-to-mixing ratio conditions. It maintains a high detection probability while reducing the false alarm rate, avoiding the significant performance degradation of existing matched filtering methods under strong reverberation conditions. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a comprehensive architecture diagram of the underwater target echo detection method under strong reverberation background of the present invention;

[0027] Figure 2 Simulation results for a signal reverberation ratio of 3 dB: (a) Results of the method of this invention; (b) Time-domain waveform of the received signal; (c) Matched filter output;

[0028] Figure 3Simulation results for a signal reverberation ratio of 0 dB: (a) Results of the method of this invention; (b) Time-domain waveform of the received signal; (c) Matched filter output;

[0029] Figure 4 Simulation results for a signal reverberation ratio of -3 dB: (a) Results of the method of this invention; (b) Time-domain waveform of the received signal; (c) Matched filter output;

[0030] Figure 5 The simulation results are as follows: (a) Results of the method of this invention; (b) Time-domain waveform of the received signal; (c) Matched filter output. Detailed Implementation

[0031] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0032] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0033] like Figure 1 The method for detecting underwater target echoes under strong reverberation backgrounds, as shown, specifically includes the following steps:

[0034] Step 1: Bandpass filtering.

[0035] The sonar echo signal received by the bandpass filter unit is filtered to remove out-of-band interference, retaining only the effective signal within the operating frequency band, thus obtaining the filtered signal. .

[0036] (1)

[0037] In the formula, For active sonar signal reception, For finite impulse response filters, This is the filtered signal.

[0038] Step 2: Slide the window into segments.

[0039] A segmented processing unit is used to divide the filtered signal into several frames to adapt to the non-stationary characteristics of reverberation.

[0040] (2)

[0041] In the formula, The first segment extracted from the filtered complete signal sequence Segment of signal, This is a segment index used to identify segment numbers. This represents the sampling point interval between adjacent segments. This is the index of the sampling points within the segment, used to represent the first... The location of the sampling point in the segment signal. The number of sampling points for each signal segment. This represents the total number of signal segments.

[0042] Step 3, envelope extraction.

[0043] An analytic signal is constructed in the envelope extraction unit using the Hilbert transform.

[0044] (3)

[0045] In the formula, H{ } is the Hilbert operator. From this, the instantaneous envelope amplitude is obtained:

[0046] (4)

[0047] In the formula, For the first The signal segment in the first The instantaneous envelope amplitude at each sampling point is used to characterize the change in the instantaneous intensity of the signal.

[0048] Step 4: Envelope normalization.

[0049] To eliminate amplitude dependence while preserving fluctuation characteristics, and to ensure the spectral variance quotient is... The features are unaffected by absolute power; only fluctuation characteristics are considered. The extracted envelope is normalized using root-mean-square normalization within the normalization unit.

[0050] (5)

[0051] In the formula, This is the envelope amplitude after root mean square normalization of the instantaneous envelope, used to eliminate amplitude dependence.

[0052] Step 5, envelope differentiation.

[0053] To enhance high-frequency fluctuations and amplify discriminative features, a first-order difference is performed on the normalized envelope within the envelope differential unit:

[0054] (6)

[0055] It represents the first-order difference result of the normalized envelope signal, which reflects the rate of change of the envelope over time. It characterizes the "slope" of the signal envelope fluctuations, and will show a large change at abrupt changes in the echo envelope, such as the target echo position. Indicates the first The signal segment in the first Normalized envelope amplitude at each sampling point Indicates the first The signal segment in the first Normalized envelope amplitude at each sampling point. The sampling frequency is expressed in Hz, representing the number of samples taken per second.

[0056] Step 6: Calculate the weighted envelope spectrum.

[0057] In the spectrum transformation unit, a Fourier transform is performed on the differential envelope signal, followed by high-frequency weighting to obtain the weighted envelope spectrum:

[0058] (7)

[0059] Step 7: Feature extraction.

[0060] The feature extraction process for the weighted envelope spectrum in the feature calculation unit includes the following four steps:

[0061] (a) Envelope spectrum normalization.

[0062] Normalizing the envelope spectral power yields the probability distribution:

[0063] (8)

[0064] In the formula, For the first The weighted envelope spectrum of the segment signal, The maximum frequency index of the weighted envelope spectrum corresponds to the total number of frequency points. .

[0065] (ii) Calculation of spectral entropy.

[0066] Calculate the spectral entropy based on the normalized distribution:

[0067] (9)

[0068] In the formula, To prevent abnormally small constants in logarithmic operations.

[0069] (III) Calculation of Spectral Variance.

[0070] Calculate the spectral variance using the mean of the normalized envelope spectrum as a benchmark:

[0071] (10)

[0072] In the formula, For the first The average value of the weighted envelope spectrum of the segment signal.

[0073] (iv) Constructing the spectral variance-entropy ratio index.

[0074] Based on spectral entropy and spectral variance, construct the spectral variance-entropy ratio:

[0075] (11)

[0076] Indicates the first The spectral variance-entropy ratio of a segment signal is used to represent the energy distribution characteristics of the envelope spectrum in the frequency domain. The spectral entropy of the envelope spectrum reflects the uniformity of the spectral energy distribution. The larger the entropy value, the flatter the spectrum and the more dispersed the energy. The spectral variance of the envelope spectrum represents the intensity of fluctuations in the spectral amplitude. A larger variance indicates more concentrated energy and more pronounced fluctuations.

[0077] It combines the randomness (entropy) of energy distribution with the discreteness (variance) of amplitude, making it more sensitive to subtle differences in spectral structure between transient echoes and sustained reverberation, and its high spectral variance quotient... A random, wide-bandwidth spectrum with irregular energy fluctuations is a typical characteristic of reverberation; a low spectral variance quotient... This indicates that the spectrum is concentrated, coherent, and stable, which is a significant characteristic of target echoes. Therefore, this index shows a low value at the target echo location and a high value at the reverberation location, effectively distinguishing the target from the reverberation.

[0078] Step 8: Adaptive decision.

[0079] In the decision unit, a dynamic threshold is set based on the local reverberation statistics. When detected Below the threshold When the signal segment is detected, a target echo is determined to be present. Robust target classification is achieved under different SRR conditions by dynamically adjusting the decision boundary.

[0080] (12)

[0081] like Figures 2-5 As shown, the methods were compared under different signal reverberation ratios. In the output of the matched filter, as the signal reverberation ratio decreases, the number of clutter peaks gradually increases. At 3 dB, more than 6 peak values ​​close to the target amplitude appear. At 6dB, with more than 10 peak values, the target signal is gradually masked, increasing the detection difficulty. However, using the method of this invention, a stable spectral valley characteristic is maintained at the target location. This valley contrasts with the reverberation background under all tested signal reverberation ratio conditions, thus enabling detection at low reverberation levels. Target identification and detection are still achieved even in a 6 dB reverberation environment.

[0082] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting underwater target echoes under strong reverberation background, characterized in that, include: The received sonar echo signal is bandpass filtered to obtain the filtered signal; The filtered signal is segmented using a sliding window to form multiple short-time signal segments; Extract the envelope of the short-time signal segment and normalize the envelope; The differential envelope signal is obtained by performing differential processing on the normalized envelope; The differential envelope signal is subjected to spectral transformation and frequency weighting to obtain the weighted envelope spectrum; Feature extraction is performed on the weighted envelope spectrum, the spectral entropy and spectral variance are calculated, and the spectral variance-entropy ratio is constructed. An adaptive threshold is set based on local reverberation statistics. When the detected spectral variance-entropy ratio is lower than the adaptive threshold, it is determined that a target echo exists. When the detected spectral variance-entropy ratio is higher than the adaptive threshold, it is determined that no target echo exists.

2. The method for detecting underwater target echoes under strong reverberation background according to claim 1, characterized in that, The normalized envelope is differentially processed using a first-order difference method, specifically as follows: This represents the first-order difference result of the normalized envelope signal, used to reflect the rate of change of the envelope over time. Indicates the first The signal segment in the first Normalized envelope amplitude at each sampling point Indicates the first The signal segment in the first Normalized envelope amplitude at each sampling point Represents the discrete sampling index of the envelope sequence. Indicates the first The numbering of signal segments, The sampling frequency.

3. The method for detecting underwater target echoes under strong reverberation background according to claim 1, characterized in that, Construct the spectral variance-entropy ratio based on spectral entropy and spectral variance: in, Indicates the first The spectral variance-entropy ratio of a segment signal is used to represent the energy distribution characteristics of the envelope spectrum in the frequency domain. The spectral entropy, representing the envelope spectrum, reflects the uniformity of the spectral energy distribution. This represents the spectral variance of the envelope spectrum, reflecting the intensity of fluctuations in the spectral amplitude.

4. The method for detecting underwater target echoes under strong reverberation background according to claim 1, characterized in that, The envelope is normalized using the root mean square amplitude method.

5. The method for detecting underwater target echoes under strong reverberation background according to claim 1, characterized in that, Frequency indexing is used to perform frequency weighting on the differential envelope signal, thereby increasing the role of high-frequency components in the weighted envelope spectrum.

6. The method for detecting underwater target echoes under strong reverberation background according to claim 1, characterized in that, An adaptive threshold is dynamically set based on the statistical characteristics of local reverberation samples to adapt to the detection requirements under different signal reverberation ratio conditions.

Images