Underwater acoustic communication leading signal detection method, device and system

Abstract

The invention discloses an underwater acoustic communication preamble signal detection method, device and system, and belongs to the field of underwater acoustic communication. The underwater acoustic communication preamble signal detection method comprises the following steps: after a signal is received, carrying out windowing processing on the received signal by adopting a sliding window; performing a hard iteration threshold algorithm on the data in each window to obtain a reconstruction coefficient of each window; performing signal reconstruction based on the dictionary matrix and the reconstruction coefficient, and calculating the energy concentration degree of each window after signal reconstruction; and determining whether a preamble signal exists in the received signal based on the energy concentration degree. According to the scheme, the hard iteration threshold algorithm is adopted to replace a matching tracking algorithm in the prior art, on the basis of ensuring that the detection performance is not changed, the calculation amount and the processing time consumption are greatly reduced, and the requirements for real-time performance and low power consumption of underwater acoustic communication are better met.

Description

Technical Field

[0001] This application relates to the field of underwater acoustic communication technology, and in particular, to a method, apparatus and system for detecting underwater acoustic communication preamble signals. Background Technology

[0002] Marine information technology has become one of the core driving forces leading the high-quality development of my country's marine economy. my country has achieved remarkable results in the construction of its marine information infrastructure. The efficient acquisition and reliable transmission of underwater data are of great significance for ensuring the monitoring of the marine ecological environment and safeguarding national maritime rights. However, compared with the diversified development of space-based and surface communication technologies, underwater acoustic communication remains the main means of achieving long-distance information exchange underwater.

[0003] Preamble detection is a crucial step in establishing an underwater acoustic communication link, and its detection performance directly determines the demodulation quality of subsequent communication data. This step faces multiple constraints related to accuracy, false alarm rate, and real-time performance. A low detection rate will directly lead to data frame loss, preventing the establishment of an effective link; a high false alarm rate will cause frequent false wake-ups, increasing the ineffective power consumption of the communication system; and if the detection algorithm takes too long to process, it will not only reduce the system's real-time performance but may even cause communication failure due to missing the optimal capture window. Therefore, improving the detection rate, reducing the false alarm rate, and controlling processing latency are critical issues that urgently need to be addressed in underwater acoustic communication. Traditional methods for preamble detection mainly rely on the physical or statistical characteristics of the signal, with energy detection and matched filtering being the most representative. Energy detection detects the preamble by accumulating the energy present during the signal period and the energy present only in the background noise, using the difference between the two. This method is simple in principle and implementation, does not require prior information about the signal, and is one of the earliest solutions applied to preamble detection in underwater acoustic communication. However, in situations with low signal-to-noise ratios or interference, this method struggles to distinguish whether a signal has arrived, resulting in severe performance degradation.

[0004] Matched filtering, with matched filters at its core, is widely used in radar and sonar communication systems. Matched filters are optimal detectors under Gaussian background noise. They determine whether a preamble signal has arrived by comparing the correlation peak obtained from the correlation operation between the received signal and a local template signal with a threshold value. Diamant et al. provided closed-form expressions for the detection rate and false alarm rate of normalized matched filters in underwater acoustic communication and verified their correctness using sea trial data. Normalized matched filters offer higher noise immunity. However, neither method addresses multipath propagation. With a high level of multipath propagation in the channel, signal energy becomes further dispersed, causing the matched filtering result to fail to meet the threshold, leading to a decrease in the detection rate.

[0005] With the development of technology, various modern detection methods have emerged in recent years, mainly including data-driven neural network methods and sparse detection methods based on channel characteristics. Neural network detectors utilize deep learning models to mine signal features. Although they can approximate the theoretical optimal solution under ideal conditions, their over-reliance on offline training distribution leads to significant performance degradation in real non-steady-state channels. Furthermore, once the network parameters are fixed, they lose their ability to be adjusted online, making it difficult to meet the stringent requirements of real-time robust detection in high-dynamic scenarios.

[0006] The sparse detector method leverages the sparse multipath characteristics of underwater acoustic channels to offer a novel solution for preamble detection in complex interference environments. Unlike traditional matched filtering, which relies solely on the strongest path for decision-making, this method employs overlapping sliding windows to divide the received signal into blocks, constructs an overcomplete dictionary using preamble templates, and applies a matched pursuit method for signal reconstruction in each block. Li et al. proposed two detection statistics—Energy Concentration (EC) and Accumulated Correlation Coefficient (ACC)—as decision criteria. Experiments demonstrate that its performance significantly outperforms traditional methods in complex interference scenarios.

[0007] However, the signal reconstruction methods used by sparse detectors involve matched pursuit and orthogonal matched pursuit, which require large-dimensional complex matrix multiplications, resulting in a high computational cost. Especially in complex multipath scenarios, the computational cost of sparse detector methods increases significantly, making it difficult to meet the real-time and low-power requirements of underwater acoustic communication. Summary of the Invention

[0008] To overcome the shortcomings of related technologies, this application provides a method, apparatus and system for detecting underwater acoustic communication preamble signals, in order to solve the problems that existing sparse detector methods have high computational load and are difficult to meet the real-time and low power consumption requirements of underwater acoustic communication.

[0009] The technical solution adopted by this application to solve its technical problem is: Firstly, a method for detecting a preamble signal in underwater acoustic communication is provided, comprising: After receiving the signal, a sliding window is used to process the received signal into windows, where the window length is greater than the preset template signal length. The reconstruction coefficients for each window are obtained by applying a hard iterative thresholding algorithm to the data within each window. The signal is reconstructed based on the dictionary matrix and the reconstruction coefficients, and the energy concentration of each window after signal reconstruction is calculated. The time delay of each atom in the dictionary matrix is ​​introduced with a preset sparse displacement step size. The presence of a preamble signal in the received signal is determined based on the energy concentration.

[0010] Furthermore, the step of obtaining the reconstruction coefficients for each window by applying a hard iterative thresholding algorithm to the data within each window includes: Initialize coefficients, estimate coefficients, and set sparsity. And step size, maximum number of iterations; Gradient descent is performed based on the step size, and the estimated value is updated along the negative gradient direction of the objective function; A hard thresholding operator is applied to the updated estimate, which preserves the largest amplitude. One component is set to zero and the other components are set to zero; When the maximum number of iterations is reached or the residual energy is less than the preset residual energy, the iteration stops and the reconstruction coefficients are output.

[0011] Further, the signal reconstruction based on the dictionary matrix and the reconstruction coefficients includes: Reconstructed signal = dictionary matrix * reconstruction coefficients.

[0012] Furthermore, the dictionary matrix construction step includes: Define the template signal's first atoms, where the time delay = *Preset sparse displacement step size; The set of vectors composed of all atoms is used as a dictionary matrix.

[0013] Furthermore, the calculation of the energy concentration of each window after signal reconstruction includes: Calculate the ratio of the reconstructed signal to the original signal for each window; The ratio is used as the energy concentration.

[0014] Further, determining whether a preamble signal exists in the received signal based on the energy concentration includes: A vector is constructed based on the energy concentration of each serial port, which serves as the energy value vector of the received signal. The presence of a preamble signal in the received signal is determined based on the energy value vector.

[0015] Further, determining whether a preamble signal exists in the received signal based on the energy value vector includes: Obtain the maximum energy concentration in the energy value vector; When the maximum energy concentration is greater than a first preset energy threshold, it is determined that a preamble signal exists in the received signal; when the maximum energy concentration is less than a second preset energy threshold, it is determined that no preamble signal exists in the received signal, wherein the first preset energy threshold is greater than the second preset energy threshold.

[0016] Furthermore, it also includes: Calculate the variance of the energy value vector; When the maximum energy concentration is less than or equal to a first preset energy threshold and greater than or equal to a second preset energy threshold, and when the variance is greater than a preset variance, it is determined that a preamble signal exists in the received signal; when the variance is less than a preset variance, it is determined that no preamble signal exists in the received signal.

[0017] Secondly, an underwater acoustic communication preamble signal detection device is provided, comprising: The receiving signal processing module is used to perform windowing processing on the received signal using a sliding window after receiving the signal, wherein the window length is greater than the preset template signal length; The reconstruction coefficient acquisition module is used to obtain the reconstruction coefficient of each window by applying a hard iterative thresholding algorithm to the data within each window; The energy concentration acquisition module is used to reconstruct the signal based on the dictionary matrix and the reconstruction coefficients, and calculate the energy concentration of each window after signal reconstruction. The time delay of each atom in the dictionary matrix is ​​introduced with a preset sparse displacement step size. A preamble determination module is used to determine whether a preamble signal exists in the received signal based on the energy concentration.

[0018] Thirdly, a computer-readable storage medium is provided, on which a computer program or instructions are stored, wherein when the computer program or instructions are executed by a processor, the steps of the underwater acoustic communication preamble signal detection method provided in the first aspect of the technical solution are implemented.

[0019] Fourthly, a preamble signal detection system for underwater acoustic communication is provided, comprising: At least one processor and at least one memory; The memory stores the executable instructions of the processor; The processor is configured to execute the underwater acoustic communication preamble signal detection method provided by the first aspect of the technical solution.

[0020] Beneficial effects: This application provides a method, apparatus, and system for detecting preamble signals in underwater acoustic communication. The method includes: after receiving a signal, performing windowing processing on the received signal using a sliding window; obtaining reconstruction coefficients for each window using a hard iterative thresholding algorithm; reconstructing the signal based on a dictionary matrix and the reconstruction coefficients, and calculating the energy concentration of each window after signal reconstruction; and determining whether a preamble signal exists in the received signal based on the energy concentration. This application uses a hard iterative thresholding algorithm instead of the matching tracking algorithm in the prior art, significantly reducing computational load and processing time while maintaining detection performance, thus better meeting the real-time and low-power requirements of underwater acoustic communication. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a flowchart of a method for detecting a preamble signal in underwater acoustic communication provided in an embodiment of this application; Figure 2 This is a diagram of the core algorithm structure of a sparse reconstruction detection method for underwater acoustic preamble signals based on hard iterative cascaded thresholds provided in this application embodiment; Figure 3 This is a schematic diagram of the structure of an underwater acoustic communication preamble signal detection device provided in an embodiment of this application. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be described in detail below with reference to the accompanying drawings and embodiments. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments of this application, all other implementation methods obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] First, it's necessary to introduce existing underwater acoustic communication systems. In such systems, a preamble signal arrives at the receiving device before the communication data packet, helping the receiver identify its arrival and begin demodulation. In sonar and radar systems, the preamble signal is typically a Doppler-insensitive signal sequence, such as linear frequency modulated (LFM) and hyperbolic frequency modulated (HFM) signals. Taking hyperbolic frequency modulated (HFM) signals as an example: Existing techniques model underwater acoustic channels as a multipath impulse response in the passband. Its mathematical expression is: (1) Therefore, the channel response model using hyperbolic frequency modulated signals as preambles can be simplified to equation (2): (2) Generated passband preamble template After propagation through the channel, the signal at the receiving end It can be expressed as equation (3): (3) in Represents convolution. Indicates noise. Based on received signal samples. and known preamble template The preamble detection task can be modeled as a binary hypothesis testing problem. At each detection time, the receiver needs to make a decision between the two hypotheses in equation (4): (4) When the receiver makes a decision When the receiver switches to demodulation / decoding mode for communication data packets, it will operate in a mode that allows for the determination of communication data packets. At this time, the receiver continues to operate in preamble detection mode.

[0025] At the receiving end of the communication system, the received bandpass signal undergoes a Hilbert transform to obtain a complex signal from the real signal, and then passes through a bandpass filter to remove out-of-band noise, thereby improving the input signal-to-noise ratio. (Filtered signal sample) As input to matched filtering methods. Based on local template signal samples. Based on the established preamble detection system model, the following sections will present representative methods currently used in preamble detection tasks: 1. Normalized matched filtering, widely used in practical communication devices. 2. A sparse detector method based on the sparsity of underwater acoustic channels, combined with signal reconstruction algorithms.

[0026] 1. Normalized matched filtering method The matched filter measures the similarity between the received signal and the local template signal by calculating the cross-correlation function between the two signals, and its output is shown in Equation (5).

[0027] (5) Here, "y[n]" represents the output of the matched filter. m represents the iterator in the calculation process; in the calculation on the right side of the equals sign, m represents the range from 0 to N. T-1 is calculated each time. n represents each point of the output of the received signal x and the matched filter y. x and y are of equal length and are both n points. This means that if you want to obtain the first point value of the matched filter output, i.e. y[1], you can get it by taking n as 1 in the formula. The asterisk on s, i.e. *, represents taking the conjugate. Here s is the "local template signal".

[0028] However, the peak value of the output of a traditional matched filter directly depends on the amplitude of the signal. When the signal amplitude is unknown or fluctuates, the decision performance of a fixed threshold will decrease, making it difficult to adapt to dynamic environments. To solve this problem, the normalized matched filter introduces an energy normalization mechanism, which eliminates the dependence of the output result on the energy signal and enhances the robustness of detection. Its output is shown in Equation (6).

[0029] (6) A normalized matched filter uses the energy product of the template signal and the input signal as a normalization factor, outputting the correlation coefficient between the two. Theoretically, the output value is located at... Within the interval. In practical engineering implementation, to avoid square root operations and reduce computational complexity, underwater acoustic communication systems typically employ... As a detection statistic, the detection problem of the preamble signal can then be transformed into a binary hypothesis test as shown in equation (7).

[0030] (7) in, This represents the decision threshold of the normalized matched filter.

[0031] In underwater acoustic channels, multipath propagation causes energy dispersion in the template signal, resulting in a decrease in the signal's correlation peak value, which approaches the correlation peak value of the background noise. This makes it difficult to select the threshold of the normalized matched filter, limiting its robustness in complex multipath environments.

[0032] 2. Sparse Detector Method The sparse detector method proposed by Li et al. leverages the inherent sparsity of underwater acoustic channels, abandoning the traditional approach of matched filters that rely solely on the single strongest path. Instead, it enhances detection robustness through sparse signal reconstruction. By aggregating the energy of multiple paths, the algorithm improves its detection robustness in channel scenarios with numerous multipaths and interference.

[0033] This method first divides the received signal into overlapping data blocks and then reconstructs the signal on each data block. Its core lies in constructing a signal based on a known hyperbolic frequency-modulated preamble signal. The overcomplete dictionary is composed of integer sample delay versions. Let the preamble length be... The data block length is Then the first in the dictionary Atoms The definition is given by equation (8): (8) That is, by filling in the blanks before and after the normalized leader template respectively. Individual and A zero element, simulating the preamble signal in time delay. The propagation response at the point. The overcomplete dictionary consisting of all possible time-delay atoms is shown in equation (9).

[0034] (9) Based on this dictionary, the Matching Pursuit (MP) algorithm is used to iteratively reconstruct the observed signals within the window. In this process, Li et al. proposed two detection statistics: energy concentration and cumulative correlation coefficient. Experiments show that this method exhibits significantly better detection performance than traditional matched filters in channel environments with multiple paths and interference.

[0035] However, the dictionary dimension constructed by this method is as high as [missing information]. Here, C represents the complex field. In each iteration of the matching pursuit algorithm, the inner product of the residual vector and the dictionary matrix needs to be calculated, a process involving a large number of complex multiplication operations. Furthermore, the more preset channel multipaths there are, the more iterations are required, further exacerbating the computational burden. This surge in computational complexity makes the algorithm unsuitable for the real-time and low-power processing requirements of underwater acoustic communication systems.

[0036] To solve the above problems, refer to Figure 1 This application provides a method for detecting underwater acoustic communication preamble signals. Addressing the issue of drastically increased computational load for sparse detectors in complex multipath scenarios, this application proposes a sparse reconstruction detection method for underwater acoustic preamble signals based on hard iterative cascaded thresholds. The core algorithm structure is shown in the diagram below. Figure 2 As shown.

[0037] The method includes: S11: After receiving the signal, a sliding window is used to perform windowing processing on the received signal, wherein the window length is greater than the preset template signal length; Compared to existing normalized matched filtering methods that make a decision based on the matched filter output of the received sample signal, this application's embodiments, referencing sparse detector methods, use a sliding window to perform windowing processing on the received signal. The sliding window length is set to... The sliding step size is 1 sampling point. For a length of 1 input signal sequence , No. individual windows It can be expressed as equation (10).

[0038] (10) in, This step divides a segment of the received signal into... Multiple overlapping windows, window length Greater than the template signal length .exist Under the assumption that each overlapping window of the received signal should be independent of the transmitted signal; Under the assumption that some sliding windows of the received signal should contain the components of the transmitted template signal after multipath effect, as shown in Equation (11).

[0039] (11) Referring to the sparse detector method, set the window length. Unlike the step size set by sparse detectors This method is set (Understandably, it can also be divided by 8 or 16, etc.) In the process of processing a signal sample of the same length, the sliding window number and the number of signal feature indices obtained by this method are relatively more than those of the sparse detector method, which is beneficial for constructing test statistics.

[0040] Within the binary hypothesis testing framework of the leader detection task, the first... Signal fragment within a window It can be expressed as equation (12): (12) in These are the observation vectors actually used for reconstruction. It is a dictionary matrix. It is noise.

[0041] S12: Use a hard iterative thresholding algorithm to obtain the reconstruction coefficients for each window; As a preferred implementation of this application, the step of obtaining the reconstruction coefficients of each window by applying a hard iterative thresholding algorithm to the data within each window includes: S121: Initialize coefficients, estimate coefficients, and set sparsity. And step size, maximum number of iterations; S122: Perform gradient descent based on the step size, and update the estimated value along the negative gradient direction of the objective function; S122: Apply a hard thresholding operator to the updated estimate, the hard thresholding operator retaining the largest amplitude. One component is set to zero and the other components are set to zero; S123: Stop iteration and output reconstruction coefficients when the maximum number of iterations is reached or the residual energy is less than the preset residual energy.

[0042] For example, for each window signal In this embodiment, the IHT algorithm is used for signal reconstruction, as shown in equation (15): (15) In order to obtain from the observed values Accurate recovery of sparse observation vectors The goal is to minimize measurement errors under sparse constraints. The core logic of IHT combines gradient descent with a nonlinear projection operator. In each iteration, we first update the estimate along the negative gradient direction of the objective function. Then, a hard thresholding operator is applied to the updated estimate, which preserves the largest amplitude. One component is set to zero and the other components are set to zero. This iterative process ensures that the expected sparse structure is maintained while converging to a local minimum.

[0043] The core iterative formula of the IHT algorithm is shown in equation (16): (16) in It is a nonlinear hard thresholding operator that preserves the largest amplitude in the input vector. Set one element to zero and set the other elements to zero. The specific steps are as follows: Step 1: Initialization. Initialize the sparse coefficient estimation. Set sparsity and step length Maximum number of iterations ; Step 2: Gradient Descent. Calculate the gradient update caused by the residuals and apply the hard threshold operator: (17) (18) By descending in the gradient direction, and by... The operator forcibly projects the solution back. - Sparse space.

[0044] Step 3: Termination check. If the maximum number of iterations has been reached... If the residual energy falls below a set threshold, the iteration stops and the reconstruction coefficients are output. Finally, the reconstructed signal is obtained. .

[0045] sparsity This corresponds to the assumption in this method regarding the maximum number of multipath paths. It can be seen that the computational cost remains constant when using the same dictionary with different sparsity settings. Consequently, the computational cost does not surge when dealing with complex multipath scenarios.

[0046] S13: Reconstruct the signal based on the dictionary matrix and the reconstruction coefficients, and calculate the energy concentration of each window after signal reconstruction. The time delay of each atom in the dictionary matrix is ​​introduced with a preset sparse displacement step size. The signal reconstruction based on the dictionary matrix and the reconstruction coefficients includes: Reconstructed signal = dictionary matrix * reconstruction coefficients.

[0047] It should be noted that the dictionary matrix can be constructed using existing conventional techniques. Alternatively, it can be constructed in the following ways: exist Assuming that, given the presence of a signal, the received signal should contain the transmitted template signal. The multipath components. In order to apply the sparse reconstruction method to obtain the sparse reconstruction results within this window, it is necessary to construct a dictionary with a certain restricted isometry property (RIP).

[0048] This represents the maximum search latency. Considering that in sparse detector methods, closely adjacent latency replicas lead to high correlation between dictionary columns, thus violating the RIP condition and preventing the effective application of the hard iterative thresholding method, this application introduces a sparse shift step. Equation (13) defines The A sparse, time-delayed copy.

[0049] (13) Among them, delay These vector sets generated by the large-step time shift are combined to form a dictionary matrix suitable for this method, as shown in equation (14): (14) This construction method reduces the number of columns in the dictionary matrix from... Become This achieved the following effect: The resolution of the reconstructed signal is reduced, but acceptable. The timing resolution of the reconstructed signal relative to the preamble signal is reduced to [value missing]. It has little impact on binary hypothesis testing problems; The computational cost of matrix multiplication is reduced with each sliding window, significantly improving computation speed; This reduces the correlation coefficient between columns in the dictionary matrix, improving the effectiveness of the hard iterative thresholding algorithm.

[0050] As a preferred implementation of this application, the calculation of the energy concentration of each window after signal reconstruction includes: Calculate the ratio of the reconstructed signal to the original signal for each window; The ratio is used as the energy concentration.

[0051] For example, for each reconstruction result Using energy concentration as an energy characteristic index, it is defined by equation (19): (19) or, .

[0052] S14: Determine whether a preamble signal exists in the received signal based on the energy concentration.

[0053] The step of determining whether a preamble signal exists in the received signal based on the energy concentration includes: A vector is constructed based on the energy concentration of each serial port, which serves as the energy value vector of the received signal; the presence of a preamble signal in the received signal is determined based on the energy value vector.

[0054] Further, determining whether a preamble signal exists in the received signal based on the energy value vector includes: Obtain the maximum energy concentration in the energy value vector; When the maximum energy concentration is greater than a first preset energy threshold, it is determined that a preamble signal exists in the received signal; when the maximum energy concentration is less than a second preset energy threshold, it is determined that no preamble signal exists in the received signal, wherein the first preset energy threshold is greater than the second preset energy threshold.

[0055] Also includes: Calculate the variance of the energy value vector; when the maximum energy concentration is less than or equal to a first preset energy threshold and greater than or equal to a second preset energy threshold, and when the variance is greater than a preset variance, determine that a preamble signal exists in the received signal; when the variance is less than a preset variance, determine that no preamble signal exists in the received signal.

[0056] For each window, the corresponding... A value can be obtained for each signal sample. The value vector is shown in equation (20).

[0057] (20) based on We construct first-order and second-order statistics to determine the presence or absence of a leading sequence: Maximum reconstruction energy: This reflects the relative strength of the reconstruction signal of the strongest window; Energy variance between windows: This reflects the salience of the signal; the preamble signal exists only within a few windows, leading to... The signal reconstruction energy is relatively large; while the signal reconstruction energy within the noise window is close. Smaller.

[0058] Based on the two statistics constructed above, three thresholds are set. , , The specific judgment steps are as follows: Step 1: Compare the maximum reconstruction energy with the energy threshold. If the maximum reconstruction energy falls within the positive or negative region, immediately output the decision result. : (twenty one) Step 2: Compare the energy variance between windows with the variance threshold. If the maximum reconstruction energy falls into the fuzzy region, activate the secondary decision logic and use the energy variance between windows to give the decision result. : (twenty two) The reason for introducing variance statistics in this method is: The IHT reconstruction method fits the noise segment better than the matching pursuit method, and the single threshold judgment is still insufficient to distinguish whether the preamble signal exists. Using second-order moments can better distinguish between windows with preamble signals and windows with only noise, which is equivalent to automatically calibrating the energy of the noise segment, thereby improving the detection rate and reducing the false alarm rate.

[0059] This application uses the IHT algorithm instead of the matching pursuit algorithm, and with the designed dictionary matrix, significantly reduces the computational cost of matrix multiplication. The reduction in computational cost is mainly reflected in the number of complex multiplications. Next, we will calculate the number of complex multiplications when the two algorithms process an 8192-point sample.

[0060] In existing matching pursuit methods, complex multiplication occurs in the calculation of the residual-to-lexicographic inner product, the calculation of the residual, and the signal reconstruction. For sparse detectors, complex multiplication is concentrated in the matching pursuit algorithm in the calculation of the residual-to-lexicographic inner product, the calculation of the residual, and the signal reconstruction. Based on the parameter settings of the sparse detector, with a sliding window number of 5, a leader sequence length of 2048 points, a window length of 4096 points, and a maximum of 15 multipaths, the number of complex multiplications is shown in Equation (23).

[0061] (twenty three) Keeping the number of sliding windows, the length of the leading sequence, and the window length constant, when the number of multipaths increases to at most 20, the number of complex multiplications is as shown in equation (24).

[0062] (twenty four) It can be seen that the total number of complex multiplications increases as the number of multipaths increases in sparse detectors.

[0063] Complex multiplication in the proposed method appears in the calculation of support vectors, iteration step size, residuals, and signal reconstruction. The sparsity is set to 15, the number of sliding windows is 9, the maximum number of IHT iterations is 5, the length of the leader sequence is 2048 points, the window length is 4096 points, and the number of complex multiplications is shown in Equation (25).

[0064] (25) As can be seen, when the number of at most multipaths increases, i.e. when sparsity needs to be improved, the complex multiplier does not increase accordingly. Increasing the number of at most multipaths only changes the number of supporting elements retained in each iteration.

[0065] Combining equations (23), (24), and (25), it can be seen that, under the same assumption of at most 15 multipaths, the number of complex multiplications required by the sparse detector is 24.75 times that of the proposed method. The reduction in the number of complex multiplications mainly comes from the reduction in the dictionary matrix. The same parameter configuration will be used for validation on experimental data.

[0066] In summary, the method provided in this application consists of four stages: window partitioning, data preparation, signal reconstruction, and cascaded threshold decision. To more intuitively illustrate the data interaction logic between each step, a complete description of the processing flow from the original input signal to the final detection output is provided below: Input: Received signal sequence; sliding window parameters; signal reconstruction parameters; threshold decision parameters; Output: Signal detector decision result; 1: Execute formula (11) to construct the sliding window signal from the received signal sequence and sliding window parameters; 2: Execute formulas (12) and (13) to construct a dictionary matrix using the signal reconstruction parameters; 3: For each signal within a window, execute formula (16) to obtain the reconstructed signal; 4: Obtain the energy concentration of each window. ; 5: According to Construction Statistics: Maximum Reconstruction Energy Energy variance between windows ; 6: Based on the threshold decision parameters, execute (21) and (22) to obtain the decision result.

[0067] The method provided in this application addresses the problem of severe multipath effects in underwater acoustic communication, which makes preamble signal detection difficult. An improved algorithm is proposed, which effectively enhances the performance and real-time performance of preamble signal detection algorithms in complex underwater acoustic channels by introducing a novel sparse reconstruction algorithm and designing a cascaded threshold discrimination mechanism. Test results on actual data show that, compared with existing sparse detector methods, the improved algorithm proposed in this application significantly reduces the average processing time per sample while maintaining comparable detection performance, thus well meeting the real-time requirements and low-power consumption requirements of underwater acoustic communication devices.

[0068] Based on the same inventive concept, this application provides a computer-readable storage medium storing a computer program or instructions thereon, wherein when the computer program or instructions are executed by a processor, the steps of the underwater acoustic communication preamble signal detection method provided in any of the above embodiments are implemented.

[0069] Based on the same inventive concept, this application provides an underwater acoustic communication preamble signal detection device 30, comprising: The receiving signal processing module 31 is used to perform window processing on the received signal using a sliding window after receiving the signal, wherein the window length is greater than the preset template signal length.

[0070] The reconstruction coefficient acquisition module 32 is used to obtain the reconstruction coefficient of each window by applying a hard iterative threshold algorithm to the data in each window; The process of obtaining the reconstruction coefficients for each window using a hard iterative thresholding algorithm includes: Initialize coefficients, estimate coefficients, and set sparsity. And step size, maximum number of iterations; Gradient descent is performed based on the step size, and the estimated value is updated along the negative gradient direction of the objective function; A hard thresholding operator is applied to the updated estimate, which preserves the largest amplitude. One component is set to zero and the other components are set to zero; When the maximum number of iterations is reached or the residual energy is less than the preset residual energy, the iteration stops and the reconstruction coefficients are output.

[0071] The energy concentration acquisition module 33 is used to reconstruct the signal based on the dictionary matrix and the reconstruction coefficients, and to calculate the energy concentration of each window after signal reconstruction. The time delay of each atom in the dictionary matrix is ​​introduced with a preset sparse displacement step size. As a preferred implementation of this application, the signal reconstruction based on the dictionary matrix and the reconstruction coefficients includes: Reconstructed signal = dictionary matrix * reconstruction coefficients.

[0072] The dictionary matrix construction step includes: Define the template signal's first atoms, where the time delay = *Preset sparse displacement step size; The set of vectors composed of all atoms is used as a dictionary matrix.

[0073] As a preferred implementation of this application, the calculation of the energy concentration of each window after signal reconstruction includes: Calculate the ratio of the reconstructed signal to the original signal for each window; The ratio is used as the energy concentration.

[0074] The preamble determination module 34 is used to determine whether a preamble signal exists in the received signal based on the energy concentration.

[0075] The step of determining whether a preamble signal exists in the received signal based on the energy concentration includes: A vector is constructed based on the energy concentration of each serial port, which serves as the energy value vector of the received signal. The presence of a preamble signal in the received signal is determined based on the energy value vector.

[0076] The step of determining whether a preamble signal exists in the received signal based on the energy value vector includes: Obtain the maximum energy concentration in the energy value vector; When the maximum energy concentration is greater than a first preset energy threshold, it is determined that a preamble signal exists in the received signal; when the maximum energy concentration is less than a second preset energy threshold, it is determined that no preamble signal exists in the received signal, wherein the first preset energy threshold is greater than the second preset energy threshold.

[0077] Furthermore, it also includes: Calculate the variance of the energy value vector; When the maximum energy concentration is less than or equal to a first preset energy threshold and greater than or equal to a second preset energy threshold, and when the variance is greater than a preset variance, it is determined that a preamble signal exists in the received signal; when the variance is less than a preset variance, it is determined that no preamble signal exists in the received signal.

[0078] Based on the same inventive concept, this application provides an underwater acoustic communication preamble signal detection system, comprising: At least one processor and at least one memory; The memory stores the executable instructions of the processor; The processor is configured to perform the underwater acoustic communication preamble signal detection method provided in any of the above embodiments.

[0079] It is understood that the same or similar parts in the above embodiments can be referred to each other, and the contents not described in detail in some embodiments can be referred to the same or similar contents in other embodiments.

[0080] It should be noted that in the description of this application, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, in the description of this application, unless otherwise stated, "a plurality of" means at least two.

[0081] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the function involved, as will be understood by those skilled in the art to which embodiments of this application pertain.

[0082] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0083] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0084] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0085] The storage media mentioned above can be read-only memory, disk, or optical disk, etc.

[0086] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0087] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for detecting a preamble signal in underwater acoustic communication, characterized in that, include: After receiving the signal, a sliding window is used to perform windowing processing on the received signal, wherein the window length is greater than the preset template signal length; The reconstruction coefficients for each window are obtained by applying a hard iterative thresholding algorithm to the data within each window. The signal is reconstructed based on the dictionary matrix and the reconstruction coefficients, and the energy concentration of each window after signal reconstruction is calculated. The time delay of each atom in the dictionary matrix is ​​introduced with a preset sparse displacement step size. The presence of a preamble signal in the received signal is determined based on the energy concentration.

2. The method according to claim 1, characterized in that, The process of obtaining the reconstruction coefficients for each window using a hard iterative thresholding algorithm includes: Initialize coefficients, estimate coefficients, and set sparsity. And step size, maximum number of iterations; Gradient descent is performed based on the step size, and the estimated value is updated along the negative gradient direction of the objective function; A hard thresholding operator is applied to the updated estimate, which preserves the largest amplitude. One component is set to zero and the other components are set to zero; When the maximum number of iterations is reached or the residual energy is less than the preset residual energy, the iteration stops and the reconstruction coefficients are output.

3. The method according to claim 1, characterized in that, The signal reconstruction based on the dictionary matrix and the reconstruction coefficients includes: Reconstructed signal = dictionary matrix * reconstruction coefficients.

4. The method according to claim 3, characterized in that, The dictionary matrix construction steps include: Define the template signal's first atoms, where the time delay = *Preset sparse displacement step size; The set of vectors composed of all atoms is used as a dictionary matrix.

5. The method according to claim 1, characterized in that, The calculation yields the energy concentration of each window after signal reconstruction, including: Calculate the ratio of the reconstructed signal to the original signal for each window; The ratio is used as the energy concentration.

6. The method according to claim 1, characterized in that, The step of determining whether a preamble signal exists in the received signal based on the energy concentration includes: A vector is constructed based on the energy concentration of each serial port, which serves as the energy value vector of the received signal. The presence of a preamble signal in the received signal is determined based on the energy value vector.

7. The method according to claim 6, characterized in that, Determining whether a preamble signal exists in the received signal based on the energy value vector includes: Obtain the maximum energy concentration in the energy value vector; When the maximum energy concentration is greater than a first preset energy threshold, it is determined that a preamble signal exists in the received signal; when the maximum energy concentration is less than a second preset energy threshold, it is determined that no preamble signal exists in the received signal, wherein the first preset energy threshold is greater than the second preset energy threshold.

8. The method according to claim 7, characterized in that, Also includes: Calculate the variance of the energy value vector; When the maximum energy concentration is less than or equal to a first preset energy threshold and greater than or equal to a second preset energy threshold, and when the variance is greater than a preset variance, it is determined that a preamble signal exists in the received signal. When the variance is less than a preset variance, it is determined that there is no preamble signal in the received signal.

9. A preamble signal detection device for underwater acoustic communication, characterized in that, include: The receiving signal processing module is used to perform windowing processing on the received signal using a sliding window after receiving the signal, wherein the window length is greater than the preset template signal length; The reconstruction coefficient acquisition module is used to obtain the reconstruction coefficient of each window by applying a hard iterative thresholding algorithm to the data within each window; The energy concentration acquisition module is used to reconstruct the signal based on the dictionary matrix and the reconstruction coefficients, and calculate the energy concentration of each window after signal reconstruction. The time delay of each atom in the dictionary matrix is ​​introduced with a preset sparse displacement step size. A preamble determination module is used to determine whether a preamble signal exists in the received signal based on the energy concentration.

10. A preamble signal detection system for underwater acoustic communication, characterized in that, include: At least one processor and at least one memory; The memory stores the executable instructions of the processor; The processor is configured to perform the method according to any one of claims 1-8.

Images