Underwater acoustic communication codebook generation method, device and equipment and readable storage medium

Abstract

The invention discloses an underwater acoustic communication codebook generation method, device and equipment and a readable storage medium, and the method comprises the steps: normalizing the characteristics of candidate signal segments through preprocessing, and weakening the inherent correlation influence caused by a stable line spectrum and a similar envelope in real ship noise; a similarity evaluation index between candidate signal segments is defined, and similarity misjudgment caused by similar segment characteristics is avoided; based on the similarity, a simulated annealing optimization algorithm is adopted to screen a target signal segment from a candidate signal set, the algorithm does not adopt a simple random or greedy selection mode, and low-cross-correlation segments are screened in a targeted mode through heuristic search logic, so that the defect that random or greedy selection is easy to select high-cross-correlation segments is effectively avoided, and the selection accuracy of the target signal segment is improved. Further, the codebook separability is improved, and the problem that the codebook separability is poor due to the fact that real ship noise contains stable line spectrums and similar envelopes is solved; meanwhile, in the screening process of the method, a low-correlation starting point set is obtained through farthest point sampling, and a traditional exhaustion or global optimization mode does not need to be adopted.

Description

Technical Field

[0001] This application relates to the field of underwater acoustic communication, specifically to an underwater acoustic communication codebook generation method, apparatus, device, and readable storage medium. Background Technology

[0002] Currently, concepts such as "digital ocean" and "underwater Internet of Things" have become the research focus in the field of underwater acoustic communication networks. The security and information protection requirements of underwater acoustic communication continue to increase, and there is an urgent need to expand the existing covert underwater acoustic communication codebook based on camouflage in order to enhance the concealment of communication behavior, prevent communication signals from being detected or intercepted by non-target receivers, and ensure the information security of underwater networks.

[0003] In related technologies, a method for generating underwater acoustic covert communication codebooks based on ship radiated noise signals uses a simple random or greedy selection strategy to extract signal segments from measured ship radiated noise data as candidate codebook sequences.

[0004] However, for attempts to generate codebooks based on ship radiated noise, on the one hand, real ship noise contains stable line spectra and similar envelopes, and simple random or greedy selection can easily filter out segments with high cross-correlation, resulting in poor codebook separability and affecting communication reliability; on the other hand, selecting M subsets with the worst cross-correlation from the candidate set is an NP-hard problem, and traditional exhaustive or global optimization methods are not feasible. There is a lack of practical approximate optimal search strategies, making it difficult to meet the codebook performance requirements of underwater acoustic covert communication. Summary of the Invention

[0005] This application provides a method, apparatus, device, and readable storage medium for generating underwater acoustic communication codebooks, which can solve the technical problems of poor separability and difficulty in searching traditional communication codebooks.

[0006] In a first aspect, embodiments of this application provide a method for generating an underwater acoustic communication codebook, the method comprising: The original signal of ship radiated noise is acquired, the original signal is discretized, candidate signal segments are extracted and preprocessed, and a candidate signal set is constructed. Define a similarity evaluation index between candidate signal segments, and calculate the similarity between each candidate signal segment in the candidate signal set; Based on the similarity, a simulated annealing optimization algorithm is used to filter target signal segments from the candidate signal set and construct an underwater acoustic communication codebook with low cross-correlation characteristics.

[0007] In conjunction with the first aspect, in one implementation, the discretization of the original signal, extraction of candidate signal segments, and construction of a candidate signal set includes: Set a fixed length for the candidate signal segment and an extraction step size; According to the fixed length and extraction step size, multiple signal segments are extracted from the discretized original signal as candidate signal segments. The candidate signal segments constitute the candidate signal set, and the number of candidate signal segments in the candidate signal set is greater than the number of signal segments required by the preset underwater acoustic communication codebook. In conjunction with the first aspect, in one implementation, defining a similarity evaluation index between candidate signal segments and calculating the similarity between each candidate signal segment in the candidate signal set includes: Define a finite lag range to clarify the time lag boundary used to determine the similarity of candidate signal segments; Normalized cross-correlation is defined as a similarity evaluation index between candidate signal segments. This index is used to quantify the waveform similarity between any two candidate signal segments within the finite hysteresis range. Based on the calculation results of the normalized cross-correlation, a distance index is derived. The distance index is used to characterize the degree of difference between two candidate signal segments and is negatively correlated with the similarity between the candidate signal segments.

[0008] In conjunction with the first aspect, in one implementation, the step of using a simulated annealing optimization algorithm to filter target signal segments from the candidate signal set includes: The farthest point sampling method is used to select an initial combination of signal segments from the candidate signal set as the initial solution for simulated annealing optimization; The number of signal segments in the initial signal segment combination is consistent with the number of signal segments required by the preset underwater acoustic communication codebook.

[0009] In conjunction with the first aspect, in one implementation, the step of using a simulated annealing optimization algorithm to filter target signal segments from the candidate signal set includes: Set the initial temperature, temperature decay coefficient, and iteration limit for simulated annealing; Based on the initial solution, a new solution is generated through a neighborhood exchange strategy, wherein any signal segment in the current solution is replaced with a signal segment from the candidate signal set that was not selected for the current solution. The new solution is accepted based on a preset acceptance probability criterion, which is related to the current temperature and the difference in the objective function between the new solution and the current solution. The temperature is gradually reduced according to the temperature decay coefficient, and the solution is iteratively updated until the upper limit of the iteration is reached. The signal segments corresponding to the optimal solution of the objective function during the iteration process are combined as the target signal segments.

[0010] In conjunction with the first aspect, in one implementation, the objective function of the simulated annealing optimization algorithm includes a cross-correlation minimization function, which weighs the maximum cross-correlation and average cross-correlation between candidate signal segments through weight coefficients, and introduces a penalty term to constrain candidate signal segments with too close starting point distances.

[0011] In conjunction with the first aspect, in one embodiment, the preprocessing of the candidate signal segment includes at least one of mean removal, mild pre-whitening, and unit norm processing.

[0012] Secondly, embodiments of this application provide an underwater acoustic communication codebook generation device, the underwater acoustic communication codebook generation device comprising: The data acquisition and candidate set construction module is used to acquire the original signal of ship radiated noise, discretize the original signal, extract candidate signal segments and preprocess the candidate signal segments to construct a candidate signal set. The similarity calculation module is used to define similarity evaluation indicators between candidate signal segments and to calculate the similarity between each candidate signal segment in the candidate signal set. The codebook optimization and construction module is used to select target signal segments from the candidate signal set based on the similarity and employ a simulated annealing optimization algorithm to construct an underwater acoustic communication codebook with low cross-correlation characteristics.

[0013] Thirdly, embodiments of this application provide an underwater acoustic communication codebook generation device, which includes a processor, a memory, and an underwater acoustic communication codebook generation program stored in the memory and executable by the processor. When the underwater acoustic communication codebook generation program is executed by the processor, it implements the steps of the underwater acoustic communication codebook generation method as described in some of the above embodiments.

[0014] Fourthly, embodiments of this application provide a computer-readable storage medium storing an underwater acoustic communication codebook generation program, wherein when the underwater acoustic communication codebook generation program is executed by a processor, it implements the steps of the underwater acoustic communication codebook generation method as described in some of the above embodiments.

[0015] The beneficial effects of the technical solutions provided in this application include: Preprocessing is used to regularize the characteristics of candidate signal segments, reducing the inherent correlation effects of stable line spectra and similar envelopes in real ship noise. A similarity evaluation index is defined among candidate signal segments, and this index is used to accurately quantify the correlation between each candidate signal segment in the candidate signal set, avoiding misjudgment of similarity due to similar segment characteristics. Based on the similarity, a simulated annealing optimization algorithm is used to select target signal segments from the candidate signal set. This algorithm does not use a simple random or greedy selection method, but uses heuristic search logic to selectively filter low cross-correlation segments. This method effectively avoids the drawbacks of random or greedy selection, which easily leads to the selection of highly cross-correlated segments, thereby improving codebook separability and overcoming the problem of poor codebook separability caused by the presence of stable line spectra and similar envelopes in real ship noise. At the same time, in the screening process, this method first obtains a set of low-correlation starting points by sampling the farthest point, without the need for traditional exhaustive search or global optimization methods. It achieves the screening of target signal segments with limited computing resources using an approximately optimal search strategy, successfully solving the NP-hard problem of selecting M subsets with the worst cross-correlation from the candidate set, and finally constructing a codebook for underwater acoustic communication with low cross-correlation characteristics. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating an embodiment of the underwater acoustic communication codebook generation method of this application; Figure 2 This is a flowchart illustrating another embodiment of the underwater acoustic communication codebook generation method of this application; Figure 3 This is a thermal comparison diagram of codebook segment cross-correlation in one embodiment of the underwater acoustic communication codebook generation method of this application; Figure 4 This is a schematic diagram of the hardware structure of the underwater acoustic communication codebook generation device involved in the embodiments of this application. Detailed Implementation

[0017] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0018] Concepts such as "digital ocean" and "underwater Internet of Things" are gradually becoming one of the research focuses in the field of underwater acoustic communication networks, which places higher demands on the security and information protection of underwater acoustic communication. Hiding the information transmission process through secure underwater acoustic communication technology ensures that the communication activity cannot be detected by non-target receivers and that communication information cannot be intercepted, greatly improving the information security of underwater networks. This has significant practical implications and research value for the security of underwater communication networks.

[0019] Covert communication in terrestrial radio frequency and radar applications has long been extensively studied. However, underwater communication faces unique challenges. Underwater acoustic communication links are affected by various factors in the underwater environment (water depth, temperature, etc.), resulting in high uncertainty in the communication channel. Signal attenuation, delay, noise interference, and multipath propagation effects manifest complexly under different environmental conditions. This makes it impossible to directly apply airborne covert wireless communication technologies to underwater applications.

[0020] Existing physical layer security methods, both domestically and internationally, can achieve secure communication using signal modulation, waveform design, frequency modulation time hopping, camouflage, and spatial focusing. Extensive research has been conducted in low signal-to-noise ratio (SNR) communication and biomimetic underwater acoustic communication. Low SNR communication reduces power spectral density by widening the bandwidth of the transmitted signal using spreading codes. Existing methods for generating spreading sequences in underwater acoustic spread spectrum communication are typically based on classical chaotic mappings, utilizing one-dimensional discrete mappings or two-dimensional continuous systems, such as Logistic mappings, Tent mappings, Lorenz systems, and Chen systems. These methods binarize the continuous chaotic sequence to generate binary spreading codes, and finally improve correlation through parameter fine-tuning or post-processing. However, the Lyapunov exponent distribution and recurrence graph of traditional chaotic sequences exhibit a regular structure. When a third party uses a multi-element receiver system to monitor for an extended period, the communication signal can still be detected, and this phenomenon becomes increasingly apparent as the distance between the third party and the transmitter increases. Therefore, the security of this method still faces some challenges. In response to this phenomenon, biomimetic underwater acoustic communication technology has emerged and gradually gained attention. By studying the characteristics of marine organism vocalizations, with dolphins as the primary research subject, underwater communication is achieved using real or simulated biological signals. However, biofriendliness and regional requirements impose certain limitations on this method. Existing methods also utilize steganography and decoys to increase security. Steganography techniques are used to embed communication signals within camouflage signals, and related research can support the development of covert underwater communication, but the problem of low communication speeds still exists.

[0021] Currently, there is limited research on underwater acoustic covert communication codebook generation methods based on ship radiated noise signals, and the main problems include: (1) Real ship noise often contains stable line spectra and similar envelopes. Simple random or greedy selection methods can easily select a set of segments with high cross-correlation, resulting in poor codebook separability.

[0022] (2) Selecting M subsets with the worst cross-correlation in the candidate set is an NP (Non-deterministic Polynomial time) problem. Traditional exhaustive search or global optimization is not feasible, and an achievable approximate optimal search strategy is needed.

[0023] Therefore, given these shortcomings of the underwater acoustic safety communication codebook generation method based on ship radiated noise, it is urgent to address them through new design methods and technical means to meet the needs of underwater acoustic safety communication.

[0024] In a first aspect, embodiments of this application provide a method for generating underwater acoustic communication codebooks.

[0025] In one embodiment, reference is made to Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the underwater acoustic communication codebook generation method of this application. Figure 1 As shown, the underwater acoustic communication codebook generation method includes: S100: Acquire the original signal of ship radiated noise, discretize the original signal, extract candidate signal segments and preprocess the candidate signal segments to construct a candidate signal set; S200: Define a similarity evaluation index between candidate signal segments, and calculate the similarity between each candidate signal segment in the candidate signal set; S300: Based on the similarity, a simulated annealing optimization algorithm is used to select target signal segments from the candidate signal set to construct an underwater acoustic communication codebook with low cross-correlation characteristics.

[0026] In this embodiment, based on measured ship radiated noise signals, the original ship radiated noise signal is acquired and discretized. Candidate signal segments are extracted and preprocessed to construct a candidate signal set. Preprocessing regularizes the characteristics of the candidate signal segments, reducing the inherent correlation effects of stable line spectra and similar envelopes in real ship noise. A similarity evaluation index is defined between candidate signal segments. Based on this index, the correlation between each candidate signal segment in the candidate signal set is accurately quantified, avoiding misjudgments of similarity due to similar segment characteristics. Based on the similarity, a simulated annealing optimization algorithm is used to select target signal segments from the candidate signal set. This algorithm does not use simple random or greedy selection, but rather uses heuristic search logic to specifically select low cross-correlation segments, effectively avoiding the shortcomings of random or greedy selection that easily selects segments with high cross-correlation. This method improves codebook separability, overcoming the poor separability of real ship noise due to its stable line spectrum and similar envelope. Furthermore, during the selection process, it first obtains a set of low-correlation starting points through sampling from the farthest point, eliminating the need for traditional exhaustive search or global optimization. It employs an approximately optimal search strategy to select target signal segments with limited computational resources, successfully solving the NP-hard problem of selecting the M worst cross-correlation subsets from the candidate set. Ultimately, it constructs a codebook for underwater acoustic communication with low cross-correlation characteristics, ensuring high similarity between the communication signal and real ship radiated noise. This enhances the physical layer's anti-interception capability of underwater acoustic communication, while maintaining controllable computational complexity. It is feasible and efficient under typical parameter scenarios, solving the difficulties of direct modulation of ship radiated noise signals and optimizing the selection of relevant characteristic noise sequences. This reduces the difficulty of generating communication codebooks and meets the information protection requirements for safe underwater acoustic communication.

[0027] Furthermore, in one embodiment, S100 includes the following steps: S101: Set the fixed length and extraction step size of the candidate signal segment; S102: According to the fixed length and extraction step size, extract multiple signal segments from the original signal after discretization as candidate signal segments; The candidate signal segments constitute the candidate signal set, and the number of candidate signal segments in the candidate signal set is greater than the number of signal segments required by the preset underwater acoustic communication codebook.

[0028] In this embodiment, after discretizing the original ship radiated noise signal, a fixed length and extraction step size are set for the candidate signal segments. Then, multiple signal segments are extracted from the discretized original signal according to the fixed length and extraction step size as candidate signal segments. All the candidate signal segments form a candidate signal set, and the number of candidate signal segments in the candidate signal set is greater than the number of signal segments required for the preset underwater acoustic communication codebook. This provides a sufficient number of candidate samples with consistent length, standardized extraction logic, and sufficient quantity for subsequent similarity evaluation between candidate signal segments and selection of target signal segments. This ensures that there is sufficient selection space when using the simulated annealing optimization algorithm based on similarity to select low cross-correlation target signal segments. It helps to avoid the selection of high cross-correlation segments that are prone to occur with simple random or greedy selection. At the same time, it lays the foundation for the implementation of the near-optimal search strategy, thereby assisting in the construction of an underwater acoustic communication codebook with low cross-correlation characteristics. This more effectively addresses the problem of poor codebook separability caused by stable line spectra and similar envelopes in real ship noise, as well as the nondeterministic polynomial time difficulty in selecting the preset number of worst cross-correlation subsets from the candidate set.

[0029] Furthermore, in one embodiment, step S200 includes the following steps: S201: Set a finite lag range to define the time lag boundary used to determine the similarity of candidate signal segments; S202: Define normalized cross-correlation as a similarity evaluation index between candidate signal segments, and use this index to quantify the waveform similarity between any two candidate signal segments within the finite hysteresis range; S203: Based on the calculation results of the normalized cross-correlation, a distance index is derived. The distance index is used to characterize the degree of difference between two candidate signal segments and is negatively correlated with the similarity between the candidate signal segments.

[0030] In this embodiment, after constructing the candidate signal set, a finite lag range is set to clarify the time lag boundary used to judge the similarity of candidate signal segments. Normalized cross-correlation is defined as the similarity evaluation index between candidate signal segments. This index quantifies the waveform similarity between any two candidate signal segments within the finite lag range. Then, a distance index is derived based on the calculation result of the normalized cross-correlation. The distance index is used to characterize the difference between two candidate signal segments and is negatively correlated with the similarity between candidate signal segments. This provides an accurate and reliable quantitative basis for subsequent screening of target signal segments based on similarity using simulated annealing optimization algorithm. It helps to accurately identify candidate signal segments with low cross-correlation and avoids misjudgment of similarity caused by the presence of stable line spectra and similar envelopes in real ship noise. This provides support for improving codebook separability and solving the nondeterministic polynomial time difficulty of selecting a preset number of worst cross-correlation subsets from the candidate set.

[0031] Furthermore, in one embodiment, step S300 includes the following steps: S301: The farthest point sampling method is used to select an initial combination of signal segments from the candidate signal set as the initial solution for simulated annealing optimization; The number of signal segments in the initial signal segment combination is consistent with the number of signal segments required by the preset underwater acoustic communication codebook.

[0032] In this embodiment, after calculating the similarity between each candidate signal segment in the candidate signal set, the farthest point sampling method is used to select an initial signal segment combination from the candidate signal set. The number of signal segments in the initial signal segment combination is consistent with the number of signal segments required for the preset underwater acoustic communication codebook. This initial signal segment combination is used as the initial solution for simulated annealing optimization, providing a high-quality starting foundation for the iterative search of the subsequent simulated annealing optimization algorithm. This helps reduce the probability of getting trapped in local optima during the search process, improves the execution efficiency of the near-optimal search strategy, and provides preliminary support for accurately screening target signal segments with low cross-correlation. This further helps to improve the problem of poor codebook separability caused by stable line spectra and similar envelopes in real ship noise. It also lays the initial conditions for solving the nondeterministic polynomial time difficulty problem of selecting a preset number of worst cross-correlation subsets from the candidate set.

[0033] Furthermore, in one embodiment, step S300 includes the following steps: S302: Set the initial temperature, temperature decay coefficient, and iteration limit for simulated annealing; S303: Based on the initial solution, a new solution is generated through a neighborhood exchange strategy, wherein the neighborhood exchange strategy is to replace any signal segment in the current solution with a signal segment in the candidate signal set that was not selected for the current solution; S304: Determine whether to accept the new solution based on a preset acceptance probability criterion, wherein the acceptance probability criterion is related to the current temperature and the difference in the objective function between the new solution and the current solution; S305: Gradually reduce the temperature according to the temperature decay coefficient, iteratively update the solution until the upper limit of the iteration is reached, and combine the signal segments corresponding to the optimal solution of the objective function during the iteration process as the target signal segment.

[0034] In this embodiment, after obtaining the initial solution for simulated annealing optimization, the initial temperature, temperature decay coefficient, and iteration upper limit of simulated annealing are set. Based on the initial solution, a new solution is generated through a neighborhood exchange strategy. The neighborhood exchange strategy involves replacing any signal segment in the current solution with a signal segment from the candidate signal set that was not selected for the current solution. Then, a preset acceptance probability criterion is used to determine whether to accept the new solution. The acceptance probability criterion is related to the current temperature and the difference between the objective function of the new solution and the current solution. Subsequently, the temperature is gradually reduced according to the temperature decay coefficient, and the solution is iteratively updated until the iteration upper limit is reached. The combination of signal segments corresponding to the solution with the optimal objective function during the iteration process is taken as the target signal segment. This provides clear execution parameters and standardized search logic for the simulated annealing optimization algorithm, ensuring that the near-optimal search process proceeds in an orderly and efficient manner, accurately selecting combinations of low cross-correlation signal segments, and further supporting the improvement of the poor codebook separability problem caused by stable line spectra and similar envelopes in real ship noise. It provides a complete search execution scheme for effectively solving the nondeterministic polynomial time difficulty of selecting a preset number of worst cross-correlation subsets from the candidate set, and helps to construct underwater acoustic communication codebooks with low cross-correlation characteristics.

[0035] Furthermore, in one embodiment, the objective function of the simulated annealing optimization algorithm includes a cross-correlation minimization function, which weighs the maximum cross-correlation and average cross-correlation between candidate signal segments through weight coefficients, and introduces a penalty term to constrain candidate signal segments with too close starting point distances.

[0036] In this embodiment, the objective function of the simulated annealing optimization algorithm includes a cross-correlation minimization function. This cross-correlation minimization function weighs the maximum cross-correlation and average cross-correlation between candidate signal segments through weighting coefficients. At the same time, a penalty term is introduced to constrain candidate signal segments with too close starting point distances. This provides a precise guiding basis for the evaluation and selection of solutions during the simulated annealing optimization process, improves the pertinence and effectiveness of the near-optimal search strategy, and more accurately avoids the problem of selecting high cross-correlation segments caused by stable line spectra and similar envelopes in real ship noise. It provides a more targeted optimization direction for solving the nondeterministic polynomial time difficulty problem of selecting a preset number of worst cross-correlation subsets from the candidate set, and further supports the construction of underwater acoustic communication codebooks with low cross-correlation characteristics.

[0037] Furthermore, in one embodiment, the preprocessing of the candidate signal segment includes at least one of mean removal, mild pre-whitening, and unit norm processing.

[0038] In this embodiment, after extracting candidate signal segments from the discretized original ship radiated noise signal, at least one preprocessing operation is performed on the candidate signal segments, including mean removal, mild pre-whitening, and unit norm processing. This preprocessing regularizes the signal characteristics of the candidate signal segments, weakens the inherent correlation interference caused by stable line spectra and similar envelopes in real ship noise, and provides a more reliable candidate sample basis for the accurate calculation of similarity evaluation index between candidate signal segments and the selection of target signal segments based on simulated annealing optimization algorithm. This helps to improve the accuracy of low cross-correlation signal segment selection, and provides a prerequisite guarantee for improving codebook separability and solving the nondeterministic polynomial time difficulty problem of selecting a preset number of worst cross-correlation subsets from the candidate set. In this way, it supports the construction of underwater acoustic communication codebook with low cross-correlation characteristics.

[0039] In summary, as Figure 2 As shown, the underwater acoustic communication codebook generation method provided in this application embodiment is described in its entirety as follows: Step 1: Data Acquisition and Discretization (converting raw data into a processable signal) Purpose of the operation: The continuous signal of the measured ship radiated noise is converted into a discrete digital signal that can be processed by a computer, in preparation for subsequent segment extraction.

[0040] Specific operations 1. Collect a continuous signal of the measured ship radiated noise (referred to as the original continuous signal). ); 2. According to the preset sampling frequency (Unit: Hz, i.e., number of sampling points per second) Sampling a continuous signal yields a discrete-time signal x[n], mathematically expressed as:

[0041] Symbol Explanation x[n]: Discrete-time signal (digital signal), each x[n] corresponds to the value of a sampling point; n: Sampling point index (counted starting from 0); Sampling frequency, core parameters (e.g.) =1000Hz means 1000 points are sampled per second). N a The total number of sampling points, i.e., the length of the discrete signal (e.g., sampling for 10 seconds). =1000Hz, then N a =10×1000=10000).

[0042] Step 2: Constructing the candidate signal set (extracting candidate segments from discrete data) Operational Purpose From the discrete data before preprocessing, a sufficient number of candidate segments are extracted according to rules to form a candidate pool (the number of candidates > the final codebook number, to ensure the selection space).

[0043] Specific operations 1. Set 3 key parameters: M: The number of signal segments required for the final communication codebook (e.g., if 10 communication instructions are needed, then M=10). L: The fixed length of each candidate signal segment (number of sampling points, for example, L=50, that is, each segment contains 50 sampling points); h: Extraction step size of candidate segments (the number of sampling points between the starting points of two adjacent candidate segments, for example, h=20, that is, the starting point of the next segment is 20 sampling points later than the previous one); 2. Extract candidate fragments according to parameters and construct a candidate set. The mathematical expression is:

[0044] 3. Starting point rule for candidate segments: The starting point of the i-th candidate segment And must satisfy constraints (Ensure that the fragment does not exceed the range of the original data); 4. Specific representation of the i-th candidate segment (original unprocessed segment): , ,Right now ; Symbol Explanation K: The total number of candidate signal sets (e.g., K=100, meaning 100 candidate segments are extracted). : The starting index of the i-th candidate segment in the original discrete data x[n]; : The nth sampling point of the i-th candidate segment (e.g. When n=10 and n=0, [0]=x

[10] ).

[0045] Step 3: Candidate fragment preprocessing (standardization, interference elimination) Operational Purpose By removing the mean and unit norm, the baseline shift and amplitude difference of the segments are eliminated, so that subsequent similarity comparisons focus only on the "waveform shape" rather than the "amplitude" or "baseline position".

[0046] Specific operations For each candidate fragment Perform the "mean removal + mild pre-whitening + unit norm" processing to obtain the preprocessed fragment. The mathematical expression is:

[0047] Detailed calculations and explanations 1. Remove the mean: Calculate

[0048] : The i-th segment The mean (e.g.) =[2,4,6], then =(2+4+6) / 3=4, after removing the mean, it becomes [-2,0,2]); Objective: To eliminate the DC component (baseline offset) of the segments and make all segments fluctuate around 0.

[0049] 2. Calculate the L2 norm:

[0050] Calculation method: The square root of the sum of squares of vector elements (e.g., the L2 norm of [-2, 0, 2] is...). ; Meaning: Measures the magnitude of a vector after removing the mean (similar to "vector length").

[0051] 3. Unity norm: Divided by L2 norm Objective: To improve the preprocessed fragments The "length" is fixed at 1 ( ), unify the amplitude scale of all segments.

[0052] Note: The following text is omitted. The wavy line is the default. After normalization, the index is sequence.

[0053] Step 4: Definition of Similarity and Distance Metrics (Quantifying Segment Differences) Operational Purpose Define "similarity index" (to measure how similar segments are) and "distance index" (to measure how dissimilar segments are) to provide a quantitative basis for subsequent screening of "low cross-correlation segments".

[0054] Step 1: Define normalized cross-correlation (similarity index) Mathematical expression (Formula 4):

[0055] Detailed explanation 1. Key parameters: K: Lag upper limit (allowed time offset range, for example, K=20, then the offset q can take -20, -19,...,19,20); q: Time offset (let) relatively Shift left and right by q sampling points to simulate signal delay in actual communication.

[0056] 2. Calculation logic: Numeric: The sum of the products of the corresponding sampling points of the two segments after the offset (the larger the sum of the products, the more overlapping the waveforms are after the offset); Denominator: Due to preprocessing = =1, the denominator is actually 1, and the result range is fixed in [-1,1].

[0057] 3. Meaning: The closer to 1: the more similar the two segments are at offset q; The closer to 0, the less similar they are.

[0058] Step 2: Define the finite lag maximum correlation and distance Mathematical expression (Formula 5):

[0059] Detailed explanation 1. Finite lag maximum correlation : Calculation: Among all allowed offsets q, take The maximum value (i.e., the similarity between two segments at the "most matching offset"). Meaning: The "maximum similarity" (the similarity in the worst case) between two segments.

[0060] 2. Distance : Calculation: Subtract the maximum similarity from 1; Meaning: To convert "similarity" into "difference". The larger the size, the less similar the fragments (e.g.) When =0.3, =0.7, indicating a high degree of difference.

[0061] Step 5: Optimize target definition (clarify screening criteria) Operational Purpose The core requirements for "selecting M segments from K candidates" are clearly defined: the differences between each pair must be large enough and the starting interval must be compliant, which is transformed into a mathematical optimization objective.

[0062] Core optimization objective (Formula 6)

[0063] Detailed explanation 1. Optimization objective: "Maximize the minimum distance" Inner layer : Among the selected M segments, find the minimum distance between all pairs of segments (i.e., the difference between the most similar pair of segments); Outer max: Make this "minimum value" as large as possible (for example, if you select 3 segments with pairwise distances of 0.8, 0.7, and 0.6, the minimum value is 0.6; after optimization, the minimum value is increased to 0.7 to ensure the distinguishability of all segments).

[0064] 2. Constraints:

[0065] The minimum allowed interval for the start of a segment (e.g.) =50 sampling points); Objective: To avoid two segments starting too close together (or even overlapping), because the original noise is continuous, segments with close starting points are prone to repetition, resulting in high similarity.

[0066] Step 6: Initial solution selection (farthest point sampling, paving the way for optimization) Operational Purpose This provides a "high-quality starting point" (initial solution) for subsequent simulated annealing optimization, avoiding the optimization from getting stuck in "local optimum" (i.e. the selected fragment seems good, but it is not the best globally).

[0067] Specific operations Using farthest point sampling, from the candidate set Select M segments to form the initial index set Mathematical expression (Formula 7):

[0068] Detailed explanation arg max: Find the set that maximizes the subsequent "minimum value". ; Logic: Consistent with the optimization objective of formula (6), a set of segments with "the largest possible differences between each pair" is selected as the initial solution, and subsequent optimization is iterated on this basis, which is more efficient.

[0069] Step 7: Objective function construction (quantifying codebook quality to provide a basis for optimization) Operational Purpose By integrating "fragment similarity" and "starting point compliance" into a computable objective function, the simulated annealing algorithm can find the optimal codebook by "minimizing the objective function".

[0070] Step 1: Define 3 core sub-indicators (Formula 8-10) 1. Maximum cross - correlation (worst - case similarity, Equation 8)

[0071] Meaning: Among the selected M segments, the similarity of "the most similar pair" ( The smaller it is, the better the worst - case situation); : The maximum normalized cross - correlation of the pre - processed segments (i.e., ) in Equation 5.

[0072] 2. Average cross - correlation (overall situation similarity, Equation 9)

[0073] Calculation logic: Denominator M(M - 1): The total number of pairwise combinations of M segments (for example, when M = 3, the number of combinations is 3×2 = 6); Numerator : The sum of pairwise similarities without repetition (i < j to avoid double - counting, multiplied by 2 and then canceled with the denominator to get the average value); Meaning: The average similarity of all segments ( The smaller it is, the better the overall discrimination).

[0074] 3. Starting - point violation count (Equation 10)

[0075] Symbol explanation: 1{condition} is an indicator function (takes 1 when the condition is satisfied, otherwise takes 0); Meaning: Counts the number of pairs of violations where the "starting - point interval is less than " among the selected segments ( The smaller it is, the better the compliance).

[0076] Step 2: Combined objective function (Equation 11)

[0077] Parameters and meanings : Weight coefficient (for example = 0.6, indicating more emphasis on controlling the worst - case situation ); λ≥0: Penalty coefficient (for example, λ = 10, and when there is 1 pair of violations, the objective function will increase by 10, which forces the starting - point compliance); Objective: Minimize , that is, simultaneously achieve "low worst - case similarity, low average similarity, and few violations".

[0078] Step 8: Simulated annealing optimization (iteratively find the optimal codebook) Operational Purpose The objective function is searched in the candidate set using the simulated annealing algorithm. The smallest M segments are the optimal codebook (because this problem is NP hard, it cannot be exhaustively enumerated, and an approximate optimal algorithm is required).

[0079] 1. Define the state: any subset of size M (That is, the set of M currently selected segments, each state corresponds to a candidate codebook).

[0080] 2. Neighborhood swapping (rules for generating new solutions)

[0081] Operation: From the current state Remove a fragment with index i from the pool and add it to the unselected fragment with index j in the candidate pool to obtain the new state. (Minor adjustments to ensure search continuity).

[0082] 3. Acceptance probability (Formula 12)

[0083] Parameter: T>0 represents the current temperature (controls search flexibility); Acceptance logic: If the new solution is better ( < : - <0, then >1, probability of acceptance = 1 (must be accepted); If the new solution is worse ( > : - >0, then We accept the difference in solution with that probability (to avoid getting trapped in local optima).

[0084] 4. Temperature update rules (geometric cooling) ,0< <1 Cooling coefficient (usually taken as 0.8~0.95, for example) =0.9, and the temperature drops to 90% of the original value with each iteration). Objective: To explore more at the initial high temperature and focus on convergence at the later low temperature.

[0085] Complete iterative steps (closed-loop execution) 1. Initialization: Input initial solution (obtained in step 6), initial temperature T0 (e.g., T0 = 100), cooling coefficient , iteration upper limit Tmax (e.g., 1000 times); Record the current optimal solution = , the optimal objective function value J* = J( ).

[0086] 2. Iterative loop (t = 1, 2,..., T max ): a. Generate a new solution: According to the swap neighborhood rule, generate a new solution from the current solution ; b. Calculate the cost difference: ; c. Decide to accept / reject: If ΔJ < 0 (the new solution is better): Accept , let = ; If ΔJ ≥ 0 (the new solution is worse): Generate a random number r between 0 and 1. If r < , then accept , otherwise retain ; d. Update the optimal solution: If J( ) < J*, then let = , J* = J( ); e. Cool down: .

[0087] 3. Terminate the iteration: Reach the iteration upper limit T max , output the optimal solution .

[0088] Step 9: Final codebook output The optimal solution corresponding preprocessed segment , which is the final "low cross-correlation characteristic underwater acoustic communication codebook" and can be directly used for underwater acoustic covert communication (because the segment is derived from ship radiated noise, with strong concealment; and the differences between pairs are large, so commands will not be confused during communication).

[0089] Based on this, taking Figure 3 as an example, this is a heat map comparison of the cross-correlation of codebook segments, used to visually show the "similarity change between pairs of segments in the codebook before and after optimization". The core information is analyzed as follows: I. Explanation of the basic elements of the figure Graph type: Both graphs are heatmaps, with color intensity representing the degree of cross-correlation (similarity) between any two segments in the codebook.

[0090] Coordinate axes: The horizontal and vertical axes (labeled "m") represent the indices of segments in the codebook (e.g., "m=5" corresponds to the 5th segment in the codebook), covering all pairwise combinations of segments in the codebook.

[0091] Color and numerical correspondence: The color bar on the right represents the "cross-correlation value ρ" (range 0~1): the more yellow / light green the color, the closer ρ is to 1, the more similar the two segments are; the more dark blue the color, the closer ρ is to 0, the less similar the two segments are.

[0092] Yellow dashed line: diagonal line, representing "the cross-correlation between the same segment and itself", at which point ρ=1 (complete similarity), so the diagonal area is yellow, which is a reasonable baseline value.

[0093] II. Left figure (before optimization): Max ρ within ±20 ms Title meaning: It calculates "the maximum cross-correlation value ρ between any two segments within a time lag range of ±20ms".

[0094] The characteristics of the graph are: except for the diagonal, most areas are light colors such as cyan and light yellow, and the corresponding ρ values ​​are generally above 0.5. This indicates that before optimization, the similarity between different segments in the selected codebook is high, and the waveform differences of many segments are small, which makes them easy to confuse in communication.

[0095] III. Right image (optimized) The graph's characteristics are as follows: except for the diagonal lines, most areas are dark blue, and the corresponding ρ values ​​are generally below 0.5, indicating that after optimization, the similarity between different segments in the selected codebook is greatly reduced, and the waveform differences between segments are significantly improved.

[0096] Secondly, embodiments of this application also provide an underwater acoustic communication codebook generation device, comprising: a data acquisition and candidate set construction module, which acquires the original signal of ship radiated noise, discretizes the original signal, extracts candidate signal segments and preprocesses the candidate signal segments to construct a candidate signal set; a similarity calculation module, which defines a similarity evaluation index between candidate signal segments and calculates the similarity between each candidate signal segment in the candidate signal set; and a codebook optimization and construction module, which, based on the similarity, uses a simulated annealing optimization algorithm to select target signal segments from the candidate signal set to construct an underwater acoustic communication codebook with low cross-correlation characteristics.

[0097] The functions of each module in the above-mentioned underwater acoustic communication codebook generation device correspond to the steps in the above-mentioned underwater acoustic communication codebook generation method embodiment, and their functions and implementation processes will not be described in detail here.

[0098] Thirdly, embodiments of this application provide an underwater acoustic communication codebook generation device, which can be a personal computer (PC), laptop computer, server, or other device with data processing capabilities.

[0099] Reference Figure 4 , Figure 4 This is a schematic diagram of the hardware structure of the underwater acoustic communication codebook generation device involved in the embodiments of this application. In the embodiments of this application, the underwater acoustic communication codebook generation device may include a processor, a memory, a communication interface, and a communication bus.

[0100] The communication bus can be of any type and is used to interconnect the processor, memory, and communication interface.

[0101] The communication interface includes input / output (I / O) interfaces, physical interfaces, and logical interfaces used for interconnecting internal components of the underwater acoustic communication codebook generation device, as well as interfaces used for interconnecting the underwater acoustic communication codebook generation device with other devices (such as other computing devices or user equipment). Physical interfaces can be Ethernet interfaces, fiber optic interfaces, ATM interfaces, etc.; user equipment can be displays, keyboards, etc.

[0102] Memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.

[0103] The processor can be a general-purpose processor, which can call the underwater acoustic communication codebook generation program stored in the memory and execute the underwater acoustic communication codebook generation method provided in the embodiments of this application. For example, the general-purpose processor can be a central processing unit (CPU). The method executed when the underwater acoustic communication codebook generation program is called can be referred to the various embodiments of the underwater acoustic communication codebook generation method of this application, and will not be repeated here.

[0104] Those skilled in the art will understand that Figure 4 The hardware structure shown does not constitute a limitation of this application and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0105] Fourthly, embodiments of this application also provide a readable storage medium.

[0106] The present application has a readable storage medium storing an underwater acoustic communication codebook generation program, wherein when the underwater acoustic communication codebook generation program is executed by a processor, it implements the steps of the underwater acoustic communication codebook generation method described above.

[0107] The method implemented when the underwater acoustic communication codebook generation program is executed can be referred to in the various embodiments of the underwater acoustic communication codebook generation method of this application, and will not be repeated here.

[0108] It should be noted that the sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0109] The terms "comprising" and "having," and any variations thereof, in the specification, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. The terms "first," "second," and "third," etc., are used to distinguish different objects, etc., and do not indicate a sequence, nor do they limit "first," "second," and "third" to different types.

[0110] In the description of the embodiments of this application, terms such as "exemplary," "for example," or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplary," "for example," or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary," "for example," or "for instance" is intended to present the relevant concepts in a concrete manner.

[0111] In the description of the embodiments of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of this application, "multiple" means two or more.

[0112] In some processes described in the embodiments of this application, multiple operations or steps are included in a specific order. However, it should be understood that these operations or steps may not be executed in the order they appear in the embodiments of this application, or they may be executed in parallel. The sequence number of the operation is only used to distinguish different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed sequentially or in parallel, and these operations or steps may be combined.

[0113] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to execute the methods described in the various embodiments of this application.

[0114] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for generating an acoustic communication codebook, the method comprising: The method for generating the underwater acoustic communication codebook comprises the following steps: An original signal of ship radiated noise is acquired, and the original signal is discretized, candidate signal segments are extracted and preprocessed, and a candidate signal set is constructed; An evaluation index of similarity between the candidate signal segments is defined, and the similarity between each candidate signal segment in the candidate signal set is calculated; Based on the similarity, a simulated annealing optimization algorithm is used to screen target signal segments from the candidate signal set, and an underwater acoustic communication codebook with low cross-correlation characteristics is constructed.

2. The method for generating the underwater acoustic communication codebook according to claim 1, wherein the discretization of the original signal, the extraction of the candidate signal segments and the construction of the candidate signal set comprise the following steps: The fixed length and the extraction step length of the candidate signal segments are set; According to the fixed length and the extraction step length, a plurality of signal segments are extracted from the discretized original signal as the candidate signal segments; The candidate signal set is composed of all the candidate signal segments, and the number of the candidate signal segments in the candidate signal set is greater than the number of the required signal segments of the preset underwater acoustic communication codebook.

3. The method for generating the underwater acoustic communication codebook according to claim 1, wherein the definition of the evaluation index of similarity between the candidate signal segments and the calculation of the similarity between each candidate signal segment in the candidate signal set comprise the following steps: A limited lag range is set to clearly define the time lag boundary for judging the similarity of the candidate signal segments; The normalized cross-correlation is defined as the evaluation index of similarity between the candidate signal segments, and the waveform similarity degree of any two candidate signal segments within the limited lag range is quantified by the index; Based on the calculation result of the normalized cross-correlation, a distance index is derived, which is used to represent the difference degree of the two candidate signal segments and is negatively correlated with the similarity between the candidate signal segments.

4. The method for generating the underwater acoustic communication codebook according to claim 1, wherein the screening of the target signal segments from the candidate signal set by using the simulated annealing optimization algorithm comprises the following steps: The farthest point sampling method is used to select an initial signal segment combination from the candidate signal set as the initial solution of the simulated annealing optimization; The number of the signal segments in the initial signal segment combination is consistent with the number of the required signal segments of the preset underwater acoustic communication codebook.

5. The method for generating the underwater acoustic communication codebook according to claim 1, wherein the screening of the target signal segments from the candidate signal set by using the simulated annealing optimization algorithm comprises the following steps: The initial temperature, the temperature attenuation coefficient and the iteration upper limit of the simulated annealing are set; Based on the initial solution, a new solution is generated by using the exchange neighborhood strategy, which is to replace any signal segment in the current solution with a signal segment in the candidate signal set that is not selected in the current solution; According to the preset acceptance probability criterion, it is judged whether to accept the new solution, which is related to the current temperature and the difference of the objective function between the new solution and the current solution. ​ ​ ​ ​ The temperature is gradually reduced according to the temperature attenuation coefficient, and the solution is iteratively updated until the upper limit of iteration is reached. The signal segment corresponding to the optimal solution of the objective function in the iteration process is combined as the target signal segment.

6. The method of claim 1, wherein the similarity between the candidate signal segments is calculated by using a cross-correlation function.

7. The method of claim 1, wherein the pre-processing of the candidate signal segments comprises at least one of mean removal, mild pre-whitening, and unit norm processing. The water acoustic communication codebook generation device comprises: a data acquisition and candidate set construction module configured to acquire a ship radiated noise original signal, perform discrete processing on the original signal, extract candidate signal segments, pre-process the candidate signal segments, and construct a candidate signal set; 8. An underwater acoustic communication codebook generation device, characterized in that, a similarity calculation module configured to define a similarity evaluation index between the candidate signal segments, and calculate the similarity between each candidate signal segment in the candidate signal set; a codebook optimization construction module configured to select target signal segments from the candidate signal set based on the similarity and using a simulated annealing optimization algorithm, and construct a water acoustic communication codebook with low cross-correlation characteristics. The water acoustic communication codebook generation device comprises a processor, a memory, and a water acoustic communication codebook generation program stored in the memory and executable by the processor, wherein the water acoustic communication codebook generation program, when executed by the processor, implements the steps of the water acoustic communication codebook generation method according to any one of claims 1 to 7. The computer readable storage medium stores a water acoustic communication codebook generation program, wherein the water acoustic communication codebook generation program, when executed by a processor, implements the steps of the water acoustic communication codebook generation method according to any one of claims 1 to 7.

9. An underwater acoustic communication codebook generation device, characterized by, ​ 10. A computer-readable storage medium, characterized in that, ​

Images