Underwater acoustic communication Doppler factor accurate estimation method based on Gaussian Markov class weighted least squares

Abstract

The invention discloses an underwater acoustic communication Doppler factor accurate estimation method based on Gaussian Markov class weighted least squares, and the method comprises the following steps: firstly, carrying out the synchronous head detection of a communication signal received by a hydrophone, and determining the position index of each synchronous head of a direct wave; secondly, pre-estimating the length compression / expansion amount of the synchronous head based on the synchronous head position index sequence by adopting a maximum common factor method; judging and eliminating false synchronous head detection results based on the pre-estimated compression / expansion amount; and finally, accurately estimating Doppler factors based on the updated synchronous head position index sequence by adopting a Markov weighted least square method. According to the method, an analytic solution of Doppler factor estimation is obtained based on received communication signal synchronization head length compression / expansion measurement, and an inherent integer multiple relationship between multiple synchronization head compression / expansion and single synchronization head compression / expansion is utilized; and high-precision estimation of the Doppler factor can be obtained by using a Markov weighted least square method with a small calculation amount.

Description

Technical Field

[0001] The present invention relates to the field of underwater acoustic communication technology, and particularly relates to an accurate estimation method for the Doppler factor of underwater acoustic communication based on Gaussian Markov class weighted least squares, which is applicable to a robust communication system under complex underwater acoustic channel conditions. Background Art

[0002] In recent years, the country has vigorously developed the marine cause. As the only means of underwater long-distance wireless communication, underwater acoustic communication technology has attracted wide attention. Underwater acoustic communication generally refers to the communication between a moving object and a fixed object underwater, or between moving objects through an acoustic wave channel. When there is relative movement between the transmitting end and the receiving end, the Doppler effect will occur, resulting in the compression or broadening of the time-domain waveform of the received signal. Since the propagation speed of underwater acoustic waves is much lower than that of electromagnetic waves in the air, even a very small relative speed will cause serious Doppler frequency offset. To reduce the adverse effects of the Doppler effect on the reception and processing of underwater acoustic communication, it is necessary to accurately estimate and compensate the Doppler factor.

[0003] At present, many scholars at home and abroad have proposed many accurate estimation methods for the Doppler factor. The more typical ones are: (1) Dichotomy scanning: By gradually narrowing the search range, the Doppler factor closest to the true value can be quickly found. This type of method first finds a method (such as the autocorrelation method) to obtain a rough estimate value of the Doppler factor, and then sets the search range and search step size with the initial value as the center. Within this range, the search range is gradually narrowed by the dichotomy method, and the correlation between the received signal compensated by the Doppler factor at the midpoint of the range and the transmitted signal is calculated. According to the level of the correlation, decide to search left or right, and repeat the above steps until the predetermined accuracy or the number of iterations is reached. Although this method improves the estimation accuracy of the Doppler factor to a certain extent, firstly, the computational complexity is high, and multiple iterations and signal processing are required; secondly, when the initial estimate is inaccurate, the search range will become larger; and under low signal-to-noise ratio conditions, the signal matching degree is significantly affected, and the estimation accuracy is not high; finally, this method has a long calculation time and is not suitable for underwater communication scenarios with high real-time requirements. (2) Single-frequency pulse signal frequency estimation method: This method uses a single-frequency pulse signal as the synchronization header signal, and calculates the Doppler frequency offset of underwater acoustic communication by estimating the frequency of the received synchronization header signal. It is simple and intuitive, and the computational complexity is low. Usually, spectral analysis based on FFT (Fast Fourier Transformation) is used to estimate the frequency of the single-frequency pulse signal. The frequency estimation accuracy is closely related to the length of the single-frequency pulse signal. Under the condition of the same noise level, the longer the length of the single-frequency pulse signal, the higher the frequency estimation accuracy. However, to improve the communication efficiency, the length of the synchronization header signal is usually limited, resulting in low accuracy of the single-frequency pulse signal frequency estimation under low signal-to-noise ratio conditions, which in turn affects the accuracy of the Doppler factor estimation of underwater acoustic communication. Summary of the Invention

[0004] The object of the present invention is to provide an accurate estimation method for the Doppler factor in underwater acoustic communication based on Gaussian Markov - type weighted least squares.

[0005] The technical solution for achieving the object of the present invention is: an accurate estimation method for the Doppler factor in underwater acoustic communication based on Gaussian Markov - type weighted least squares, including the following steps:

[0006] Step S1: Detect the synchronization header of the communication signal received by the hydrophone to determine the position index of each synchronization header of the direct wave;

[0007] Step S2: According to the position index of the synchronization header determined in Step S1, use the greatest common divisor algorithm to determine the greatest common divisor of the difference sequence of the position indexes of the synchronization headers, and further obtain the pre - estimated compression / expansion amount of the synchronization header signal;

[0008] Step S3: According to the pre - estimated value of the compression / expansion amount of the synchronization header signal obtained in Step S2, discriminate and eliminate the false position indexes of the synchronization headers;

[0009] Step S4: Use the updated position indexes of the synchronization headers and use the Markov - type weighted least - squares estimator to accurately estimate the Doppler factor.

[0010] Further, the specific content of Step S1 includes:

[0011] Step S101: The transmitted communication signal contains (N sync + 1) synchronization header signals, the length of a single synchronization header signal is L sum sampling points, where the first L HFM sampling points are double - chirp signals, and the rest are 0. The interval between the first N sync synchronization header signals and the last synchronization header signal is N gap × L sum . A single synchronization header signal is expressed as follows:

[0012]

[0013] where f0 is the starting frequency of the double - chirp signal, k s is the modulation slope, A is the amplitude of a single synchronization header signal, cos is the cosine function, and ln is the natural logarithm function;

[0014] Step S102: Use the cross - correlation between a single synchronization header signal s sync [n] and the received communication signal to determine the position of the correlation peak of the received communication signal, and further obtain the position index L i of the correlation peak of the received communication signal, i = 1, 2... K, where K is the number of detected correlation peaks.

[0015] Further, the specific steps of step S2 include:

[0016] Step S201: Obtain the difference ΔL between the relevant peak position index sequences i,j :

[0017] ΔL i,j = L i - L j , i, j = 1, 2,... K, i > j;

[0018] Set the value range of ΔL i,j ≥ 0.95L sum , discard the differences that fall outside the set range, and record the difference array as ΔL i,j The number of elements is K';

[0019] Step S202: Arrange the obtained differences ΔL i,j in ascending order of the array elements. There may be equal or approximately equal elements. Count the number of differences in the difference array that are similar to the difference ΔL w , which is defined as the merging quality factor s n of this difference. The initial value of s n is set to 0. If

[0020]

[0021] then the merging quality factor s w corresponding to ΔL n is incremented by 1, and only keep this difference once in the new difference array, denoted as ΔL n . Denote the newly obtained difference array as {ΔL n}, n = 1, 2,... N, where N is the number of updated differences;

[0022] Step S203: For the difference ΔL n , divide each difference ΔL n in the difference array {ΔL k} by the difference ΔL n and define the multiple quality factor q n . The initial value of the multiple quality factor q n is set to 0. If

[0023]

[0024] then the multiple quality factor q n corresponding to ΔL n is incremented by 1, where round represents rounding the value to the nearest integer;

[0025] Define the comprehensive quality factor m n

[0026] m n = s n × q n

[0027] m n The ΔL corresponding to the maximum value n is what we want, denoted as ΔL g , if ΔL g satisfies:

[0028]

[0029] Then the finally obtained position index difference ΔL based on the greatest common divisor algorithm G = ΔL g , otherwise ΔL G = L sum .

[0030] Furthermore, the specific steps of step S3 include:

[0031] Step S301, the difference obtained by the greatest common divisor algorithm in step S2 is ΔL G , use the difference ΔL G to identify and eliminate the false position indexes obtained due to noise interference;

[0032] For each element in the received communication signal correlation peak array, set the initial loop index i = 1, and construct a discrimination array {T i,k}, the size of the array is 1×(K - 1), and the elements in the array satisfy

[0033]

[0034] Count the number of elements in the array {T i,k} whose values are less than 0.05. If the number is less than N sync / 2, then it is determined as a false interference related peak, and the i-th element is removed from the corresponding received communication signal correlation peak array {L k}; if the number is greater than or equal to N sync / 2, then it is considered as a true correlation peak of the synchronization header, and the i-th element in the corresponding correlation peak array {L k} is retained;

[0035] Step S302, increment the value of the loop index i by 1, and judge the size of i. If i > K, the loop stops and step S303 is executed; otherwise, step S301 is executed;

[0036] Step S303, denote the array after removing the false correlation peak indexes as {L m}, m = 1, 2,... M, M ≤ K, where M is the remaining number of correlation peaks after removal.

[0037] Further, step S4 specifically includes:

[0038] Step S401: Use the updated relevant peak array {L m} obtained in step S3 to subtract pairwise to construct a relevant peak difference observation vector n, where the m-th element n m satisfies:

[0039] n m = L i - L j , i, j = 1, 2,..., M, i > j

[0040] The size of vector n is [M × (M - 1)] / 2;

[0041] Step S402: Use the relevant peak difference vector n obtained in step S401 to construct a weighted vector h, where the m-th element h m satisfies:

[0042]

[0043] The size of vector h is [M × (M - 1)] / 2;

[0044] Step S403: After obtaining the relevant peak difference vector n and the weighted vector h, the Gaussian Markov - type weighted least - squares estimate of the length after compression / expansion of a single sync - head is given by the following formula:

[0045]

[0046] Step S404: Use the estimate of the length after compression / expansion of a single sync - head obtained in step S403 to solve for the accurate estimate of the Doppler factor

[0047]

[0048] An accurate estimation system for the Doppler factor in underwater acoustic communication based on Gaussian Markov - type weighted least - squares implements the above - mentioned method for accurately estimating the Doppler factor in underwater acoustic communication based on Gaussian Markov - type weighted least - squares, and realizes the accurate estimation of the Doppler factor in underwater acoustic communication based on Gaussian Markov - type weighted least - squares. It is divided into four modules, which respectively execute steps S1 - S4.

[0049] A computer device includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the method for accurately estimating the Doppler factor in underwater acoustic communication based on Gaussian Markov class weighted least squares, and realizes the accurate estimation of the Doppler factor in underwater acoustic communication based on Gaussian Markov class weighted least squares.

[0050] A computer-readable storage medium stores a computer program. When the computer program is executed by a processor, it implements the method for accurately estimating the Doppler factor in underwater acoustic communication based on Gaussian Markov class weighted least squares, and realizes the accurate estimation of the Doppler factor in underwater acoustic communication based on Gaussian Markov class weighted least squares.

[0051] Compared with the prior art, the remarkable advantages of the present invention are as follows: 1) The greatest common divisor algorithm is used to discriminate the correlation peaks detected by the cross-correlation algorithm, find the greatest common divisor of the peak positions to discriminate and eliminate false correlation peaks, obtain a pre-estimated compression / expansion amount of a single synchronization header signal, and improve the purity of the correlation peaks. 2) First, the greatest common divisor algorithm is used to discriminate and eliminate false correlation peaks, and a rough estimate value of the compression / expansion amount of a single synchronization header signal is obtained. Then, the Gaussian Markov class weighted least squares estimator is used to obtain an accurate estimate of the Doppler factor, which saves the repeated iteration process compared with the dichotomy scan estimation, and takes into account both the calculation accuracy and the calculation efficiency. Brief Description of the Drawings

[0052] Figure 1 It is a schematic flowchart of the method for accurately estimating the Doppler factor in underwater acoustic communication based on Gaussian Markov class weighted least squares.

[0053] Figure 2 It is a schematic diagram of the structure of the transmitted communication synchronization header signal.

[0054] Figure 3 It is a time-domain diagram of the transmitted communication synchronization header signal and the received synchronization header signal.

[0055] Figure 4 It is a comparison diagram of the compression of the received signal length caused by the Doppler effect for a single transmitted communication synchronization header signal.

[0056] Figure 5 It is a result diagram after cross-correlation of the transmitted communication synchronization header signal and the received signal. Detailed Embodiment

[0057] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0058] A method for accurately estimating the Doppler factor in underwater acoustic communication based on Gaussian Markov class weighted least squares uses the greatest common divisor algorithm to obtain a pre-estimated Doppler factor value, uses the pre-estimated Doppler factor value to discriminate and eliminate false autocorrelation peaks, and finally uses the Markov class weighted least squares method to accurately estimate the Doppler factor based on the updated synchronization header position index sequence. The specific steps are as follows:

[0059] Step 1: Detect the synchronization header of the communication signal received by the hydrophone to determine the position index of each direct wave synchronization header;

[0060] Step 1-1: The transmitted communication signal contains (N sync + 1) synchronization header signals, and the length of a single synchronization header signal is L sum sampling points, where the first L HFM sampling points are double-tone frequency modulation signals, and the rest are 0. The interval between the first N sync synchronization header signals and the last synchronization header signal is N gap × L sum . A single synchronization header signal can be expressed as follows:

[0061]

[0062] where f0 is the starting frequency of the double-tone frequency modulation signal, k s is the modulation slope, A is the amplitude of a single synchronization header signal, cos is the cosine function, and ln is the natural logarithm function;

[0063] Step 1-2: Use the cross-correlation between the single synchronization header signal s sync [n] and the received communication signal to determine the position of the correlation peak of the received communication signal, and then obtain the position index L i of the correlation peak of the received communication signal, i = 1, 2... K, and K is the number of detected correlation peaks.

[0064] Step 2: Use the greatest common divisor algorithm to determine the greatest common divisor of the synchronization header position index difference sequence, and then obtain the pre-estimated compression / expansion amount of the synchronization header signal;

[0065] Step 2-1: The position index L of the correlation peak of the received communication signal obtained in Step 1i , calculate the difference ΔL between each other i,j :

[0066] ΔL i,j = L i - L j , i, j = 1, 2,...K; i > j

[0067] Set the value range of ΔL i,j ≥0.95L sum , discard the differences that fall outside the set range, and record the difference array ΔL i,j The number of elements is K'.

[0068] Step 2-2: Arrange the elements of the obtained difference ΔL i,j array in ascending order. There may be equal or approximately equal elements. Count the number of elements in the difference array that are similar to the difference ΔL w , which is defined as the merging quality factor s n of this difference, and the initial value of s n is set to 0. If

[0069]

[0070] then the merging quality factor s w corresponding to ΔL n is incremented by 1. Only keep this difference once in the new difference array and record it as ΔL n , and record the newly obtained difference array as {ΔL n}, n = 1, 2,...N, where N is the number of updated differences;

[0071] Step 2-3: Divide each difference ΔL k by the difference ΔL n . Define the multiple quality factor q n , and each difference ΔL n corresponds to a multiple quality factor q n . The initial value of the multiple quality factor q n is set to 0. If

[0072]

[0073] holds, then the multiple quality factor q n corresponding to ΔL n is incremented by 1. Here, round represents rounding the value to the nearest integer. Define the comprehensive quality factor m n

[0074] m n = s n × q n

[0075] m n ΔL corresponding to the maximum value n is what we want, denoted as ΔL g , if ΔL g satisfies:

[0076]

[0077] Then the finally obtained position index difference ΔL based on the greatest common divisor algorithm G = ΔL g , otherwise ΔL G = L sum .

[0078] Step 3, discriminate and eliminate false position indexes according to the pre-estimated single sync header compression / dilation amount;

[0079] Step 3-1. The difference obtained by the greatest common factor algorithm in step S2 is ΔL G , use the difference ΔL G to discriminate and eliminate the false position indexes obtained due to noise interference. For each element in the received communication signal correlation peak array, set the initial loop index i = 1, and construct a discrimination array {T i,k}, the array size is 1×(K - 1), and the elements in the array satisfy

[0080]

[0081] Count the number of elements in the array {T i,k} whose values are less than 0.05. If the number is less than N sync / 2, then it is determined as a false interference related peak, and the i-th element is removed from the corresponding received communication signal correlation peak array {L k}; if the number is greater than or equal to N sync / 2, then it is considered as a true correlation peak of the sync header, and the i-th element in the corresponding correlation peak array {L k} is retained.

[0082] Step 3-2. Increase the value of the loop index i by 1, judge the size of i. If i > K, the loop stops and step S3-3 is executed; otherwise, step S3-1 is executed.

[0083] Step 3-3. Denote the array after removing the false correlation peak indexes as {L m}, m = 1, 2,...M, M ≤ K. Where M is the remaining number of correlation peaks after removal.

[0084] Step 4, estimate the compression / dilation length and accurately estimate the Doppler factor by using the Gaussian-Markov weighted least squares method according to the updated position index;

[0085] Step S4-1: Subtract the updated correlation peak arrays {L m} pairwise to construct the correlation peak difference observation vector n, where the m-th element n m satisfies:

[0086] n m = L i - L j , i, j = 1, 2,..., M, i > j

[0087] The size of the vector n is [M × (M - 1)] / 2.

[0088] Step S4-2: Use the correlation peak difference vector n obtained in Step S401 to construct the weighted vector h, where the m-th element h m satisfies:

[0089]

[0090] The size of the vector h is [M × (M - 1)] / 2.

[0091] Step S4-3: After obtaining the correlation peak difference vector n and the weighted vector h, the Gaussian Markov class weighted least squares estimate of the length after compression / dilation of a single sync header can be given by the following formula:

[0092]

[0093] Step S4-4: Solve the exact estimate of the Doppler factor for the estimate of the length after compression / dilation of a single sync header obtained in Step 4-3 Solve the exact estimate of the Doppler factor

[0094]

[0095] The present invention also proposes an underwater acoustic communication Doppler factor exact estimation system based on Gaussian Markov class weighted least squares, which implements the above-mentioned underwater acoustic communication Doppler factor exact estimation method based on Gaussian Markov class weighted least squares, realizes the underwater acoustic communication Doppler factor exact estimation based on Gaussian Markov class weighted least squares, and is divided into four modules, which respectively execute Steps S1 to S4.

[0096] A computer device includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the above-mentioned underwater acoustic communication Doppler factor exact estimation method based on Gaussian Markov class weighted least squares, and realizes the underwater acoustic communication Doppler factor exact estimation based on Gaussian Markov class weighted least squares.

[0097] A computer-readable storage medium stores a computer program thereon. When the computer program is executed by a processor, the method for accurately estimating the Doppler factor in underwater acoustic communication based on Gaussian Markov class weighted least squares is implemented, and the accurate estimation of the Doppler factor in underwater acoustic communication based on Gaussian Markov class weighted least squares is realized.

[0098] Embodiment

[0099] In order to verify the effectiveness of the proposed solution of the present invention, the following experimental design is carried out.

[0100] The simulation signal parameters are respectively set as follows: the sampling frequency is f s = 50 kHz, and the propagation speed of sound waves in water is c = 1500 m / s. The center frequency f c of the chirp signal in the transmitted synchronization header signal is 8 kHz, the bandwidth B = 4 kHz, and the length of a single synchronization header signal is L sum = 4096, where the length of the chirp signal is L HFM = 2500. N sync is set to 5, the total number of synchronization header signals is 6, N gap is set to 3, and the schematic diagram of the synchronization header signal structure is as shown in Figure 2 shown. The broadband signal-to-noise ratio of the received synchronization header signal is set to -20 dB, and the time-domain diagrams of the transmitted and received synchronization header signals are as shown in Figure 3 shown. The radial velocity of the transmitting end is set to v t = 3 m / s, and the radial velocity of the receiving end is set to v r = 3 m / s, and the true value of the Doppler factor is α = 0.996008. As shown in Figure 4 shown, due to the opposite movement of the transmitting end and the receiving end, there is length compression in the received synchronization header signal.

[0101] According to Step 1, after performing cross-correlation on multiple transmitted synchronization header signals and the received signal, and performing correlation peak detection, after setting the detection threshold, a total of 7 correlation peak position coordinate indices are detected, and they are 29526, 48934, 53014, 57093, 61173, 65252, 81571, as shown in Figure 5 shown.

[0102] According to Step 2, based on the correlation peak position coordinate indices, the difference array {ΔL n} of the correlation peak position indices is obtained = [4079, 8159, 12239, 16319, 19408, 20398, 23488, 24478, 27567, 28557, 31647, 32637, 35726, 52045], and after calculation, the merged quality factor s n= [4, 3, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], multiple quality factor q n = [8, 4, 2, 2, 1, 2, 2, 2, 2, 2, 2, 2, 1, 1], comprehensive quality factor m n = [32, 12, 4, 4, 1, 2, 2, 2, 2, 2, 2, 2, 1, 1]. The difference ΔL obtained by the greatest common divisor algorithm G = 4079, pre - estimated Doppler factor value

[0103] According to step 3, based on the pre - estimated Doppler factor and the difference ΔL obtained by the greatest common divisor algorithm G , one - by - one discrimination and elimination are performed on the seven sync - head position indexes. Finally, one false autocorrelation peak index is eliminated, and the six pure sync - head position indexes are: 48934, 53014, 57093, 61173, 65252, 81571.

[0104] According to step 4, based on the updated position indexes, a Gaussian - Markov - type weighted least - squares estimator is used to accurately estimate the length of a single sync - head signal, and the result obtained is The accurate estimated value of the Doppler factor is The relative error is 0.011‰.

[0105] Where the present invention is not described in detail are all well - known techniques to those skilled in the art.

[0106] The preferred specific embodiments of the present invention have been described in detail above. It should be understood that those of ordinary skill in the art can make many modifications and variations based on the concept of the present invention without creative labor. Therefore, all technical solutions that can be obtained by those skilled in the art in the technical field of the present invention based on the concept of the present invention through logical analysis, reasoning, or limited experiments on the basis of the prior art should be within the protection scope determined by the claims.

Claims

1. An accurate estimation method for the Doppler factor in underwater acoustic communication based on Gaussian Markov - like weighted least squares, characterized in that, It includes the following steps: Step S1: Detect the synchronization header of the communication signal received by the hydrophone to determine the position index of each synchronization header of the direct wave; Step S2: According to the position index of the synchronization header determined in Step S1, use the greatest common divisor algorithm to determine the greatest common divisor of the difference sequence of the position indexes of the synchronization headers, and then obtain the pre-estimated compression / expansion amount of the synchronization header signal; Step S3: According to the pre-estimated value of the compression / expansion amount of the synchronization header signal obtained in Step S2, discriminate and eliminate the false position indexes of the synchronization headers; Step S4: Use the updated position index of the synchronization header and use the Markov-class weighted least squares estimator to accurately estimate the Doppler factor.

2. The accurate estimation method for the Doppler factor of underwater acoustic communication based on Gaussian Markov class weighted least squares according to claim 1, characterized in that The specific content of Step S1 includes: Step S101, transmitting a communication signal includes (N sync +1) synchronization header signals, and the length of a single synchronization header signal is L sum sampling points, where the first L HFM sampling points are dual-tone frequency modulation signals, and the rest are 0. The interval between the first N sync synchronization header signals and the last synchronization header signal is N gap ×L sum , and a single synchronization header signal is represented as follows: where f0 is the starting frequency of the double-tone frequency modulation signal, k s is the modulation slope, A is the amplitude of a single synchronization header signal, cos is the cosine function, and ln is the natural logarithm function; Step S102: Cross-correlate the single synchronization header signal s sync [n] with the received communication signal to determine the position of the correlation peak of the received communication signal, and further obtain the correlation peak position index L i , where i = 1, 2... K, and K is the number of detected correlation peaks.

3. The accurate estimation method of the Doppler factor in underwater acoustic communication based on Gaussian Markov class weighted least squares according to claim 2, characterized in that The specific content of Step S2 includes: Step S201: Obtain the difference ΔL between the relevant peak position index sequences i,j : ΔL i,j = L i - L j , where i, j = 1, 2,... K and i > j; Set the value range to ΔL i,j ≥0.95L sum , discard the differences that fall outside the set range, and denote the difference array as ΔL i,j The number of elements is K′; Step S202: Take the obtained difference ΔL i,j Arrange the array elements from smallest to largest. There may be equal or approximately equal elements. Count the number of elements in the difference array that are similar to the difference ΔL w and define it as the merging quality factor s of this difference n , and set the initial value of s n to 0. If Then ΔL w The corresponding merging quality factor s n Increment by 1, and only keep this difference once in the new difference array, denoted as ΔL n , and denote the newly obtained difference array as {ΔL n}, where n = 1, 2,... N, and N is the number of differences after update; Step S203: For the difference ΔL n , divide each difference ΔL n in the difference array {ΔL k} by the difference ΔL n to define the multiple quality factor q n . Set the initial value of the multiple quality factor q n to 0. If Then ΔL n The corresponding multiple quality factor q n plus 1, where round represents rounding the logarithmic value to an integer; Define the comprehensive quality factor m n m n = s n × q n m n ΔL corresponding to the maximum value n is what we want, denoted as ΔL g , if ΔL g satisfies: Then the finally obtained position index difference ΔL based on the greatest common divisor algorithm G = ΔL g , otherwise ΔL G = L sum .

4. The method for accurately estimating the Doppler factor of underwater acoustic communication based on Gaussian Markov - like weighted least squares according to claim 3, wherein, The specific content of Step S3 includes: Step S301. The difference obtained by the greatest common divisor algorithm in Step S2 is ΔL G , and the difference ΔL G is used to identify and eliminate the false position indexes obtained due to noise interference; For each element in the peak array related to receiving communication signals, set the initial loop index i = 1, and construct a discrimination array {T i,k}, the size of the array is 1×(K - 1), and the elements in the array satisfy Count the number of elements in the statistical array {T i,k} whose values are less than 0.

05. If the number is less than N sync / 2, it is determined as a false interference related peak, and the i-th element is removed from the corresponding received communication signal related peak array {L k}; if the number is greater than or equal to N sync / 2, it is considered as a true correlation peak of the synchronization header, and the i-th element in the corresponding correlation peak array {L k} is retained; Step S302: Increment the value of the loop index i by 1, and determine the size of i. If i > K, the loop stops and Step S303 is executed; otherwise, Step S301 is executed; Step S303. Denote the array after removing the false correlation peak indices as {L m}, where m = 1, 2,... M, M ≤ K, and M is the remaining number of correlation peaks after removal.

5. A method for accurately estimating the Doppler factor of underwater acoustic communication based on Gaussian Markov - like weighted least squares according to claim 4, characterized in that, The specific content of Step S4 includes: Step S401. Use the updated correlation peak array {L m} obtained in step S3 to subtract pairwise to construct a correlation peak difference observation vector n, where the m-th element n m satisfies: n m = L i - L j , i, j = 1, 2, ..., M, i > j The size of vector n is [M×(M - 1)] / 2; Step S402. Construct a weighted vector h using the relevant peak difference vector n obtained in Step S401, where the m-th element h m satisfies: The size of vector h is [M×(M - 1)] / 2; Step S403, after obtaining the relevant peak difference vector n and the weighted vector h, the Gaussian Markov class weighted least squares estimation of the length after compression / dilation of a single synchronization header is given by the following formula: Step S404: Use the estimated value of the single synchronization header obtained in Step S403 after compression / dilation Solve for the accurate estimation of the Doppler factor 6. An underwater acoustic communication Doppler factor accurate estimation system based on Gaussian Markov class weighted least squares, characterized in that, Implement the method for accurately estimating the Doppler factor in underwater acoustic communication based on Gaussian Markov-class weighted least squares according to any one of claims 1-5, and realize the accurate estimation of the Doppler factor in underwater acoustic communication based on Gaussian Markov-class weighted least squares. It is divided into four modules, which respectively execute Steps S1 to S4.

7. A computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the method for accurately estimating the Doppler factor in underwater acoustic communication based on Gaussian Markov-class weighted least squares according to any one of claims 1-5, and realizes the accurate estimation of the Doppler factor in underwater acoustic communication based on Gaussian Markov-class weighted least squares.

8. A computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the method for accurately estimating the Doppler factor in underwater acoustic communication based on Gaussian Markov-class weighted least squares according to any one of claims 1-5, and realizes the accurate estimation of the Doppler factor in underwater acoustic communication based on Gaussian Markov-class weighted least squares.

Images