Dynamic leakage factor-based adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method

Abstract

The invention discloses an adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on a dynamic leakage factor, and belongs to the technical field of underwater acoustic communication. The method comprises the following steps: acquiring a baseband sampling signal x; a weight vector w0 is initialized to be equal to 0, and an input signal vector u0 is a zero vector with the length L; a baseband sampling signal x is used as an input signal, a training sequence is used as an expected signal d, and self-adaptive dynamic updating of a filter weight is realized by adjusting a leakage factor in real time, so that self-adaptive adjustment in a signal equalization process is completed. According to the method, a dynamic adjustment strategy of a leakage factor is introduced, and combined optimization of weight attenuation and power normalization is adopted, so that the equalization precision and the algorithm convergence speed are effectively balanced, finally, channel distortion is accurately compensated and equalized, and the communication quality and stability of an underwater acoustic communication system in a complex marine environment are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of underwater acoustic communication, and in particular relates to an adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on a dynamic leakage factor. Background Art

[0002] Problems such as equalization and adaptive beamforming of underwater acoustic communication channels can be attributed to the optimization problem of adaptive filtering. Real-time correction and adjustment of signals through adaptive algorithms have become an effective method to solve complex channel characteristics and noise suppression. Existing methods include the normalized least mean square error (NLMS) algorithm and its improved versions, such as the proportional normalized least mean square error (PNLMS) algorithm. For details of the algorithm based on proportional normalization, please refer to the article "Improved Proportionate Constrained Normalized Least Mean Square for Adaptive Beamforming", published in "Circuits, Systems, and Signal Processing" Volume 42, Issue 12 in 2023, with the article starting and ending pages 7651-7665. For details of the normalized minimum mean square error algorithm based on optimization design, please refer to the article "Optimal Design of NLMS Algorithm with a Variable Scaler Against Impulsive Interference", published in "Signal, Image and Video Processing" Volume 17, Issue 6, 2023, with the article starting and ending pages 2705-2712. Although these documents have proposed several optimization methods, they still face the problem of performance degradation in dealing with non-stationary channels and strong interference environments.

[0003] Due to the high time-varying and multipath effects of the underwater acoustic channel, equalization and interference suppression become extremely difficult. In addition, the transmission characteristics of the underwater acoustic channel may be affected by many factors, such as turbulence, temperature gradient, pressure change, etc. The traditional adaptive equalization method often loses the ability to effectively estimate the channel in a strong interference, nonlinear and rapidly changing underwater environment. Therefore, considering the unique characteristics and challenges of the underwater acoustic communication channel, it is necessary to finely adjust the weights and adaptively optimize the parameters of the equalization algorithm to improve the robustness of the signal and the overall stability of the system. The literature [Wu FY, Song YC, 2023 (Wu FY, Song Y C. Optimal design of NLMS algorithm with a variable scaleragainst impulsive interference [J]. Signal, Image and Video Processing, 2023, 17 (6): 2705-2712.)] proposes an improved NLMS algorithm based on a variable scaling factor. The algorithm aims to improve the system's ability to cope with pulse interference, but its convergence speed and stability are still lacking in the face of highly complex multipath environments.

[0004] However, the weight update strategy in the traditional NLMS and PNLMS algorithms cannot simultaneously meet the requirements of channel equalization accuracy and adaptive convergence speed, which makes it difficult to balance the equalization effect and algorithm convergence speed. Especially in underwater acoustic communications, the channel characteristics change very rapidly, and the traditional algorithms are not able to adapt to the environment. To solve this problem, the present invention proposes an adaptive non-uniform power normalized minimum mean square error algorithm based on a dynamic leakage factor, which uses the dynamic adjustment of the leakage factor to continuously adapt to channel changes, thereby improving the equalization performance in an environment with severe underwater multipath effects and noise interference. In addition, by introducing a weight attenuation strategy, a dynamic leakage factor is used to enhance the flexibility of weight updates, in order to achieve better equalization effects and faster convergence speeds, thereby significantly improving the robustness and signal processing performance of the underwater acoustic communication system. Summary of the invention

[0005] Technical issues to be solved:

[0006] In order to avoid the shortcomings of the prior art, the present invention provides an adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on dynamic leakage factor. The invention uses a dynamic weight adjustment framework, combined with an adaptive leakage factor adjustment mechanism, to continuously and finely update the filter weights, thereby achieving a robust equalization effect in a complex underwater environment. By introducing a dynamic adjustment strategy for the leakage factor and adopting joint optimization of weight attenuation and power normalization, the equalization accuracy and algorithm convergence speed are effectively balanced, and finally the channel distortion is accurately compensated and equalized, thereby improving the communication quality and stability of the underwater acoustic communication system in a complex marine environment.

[0007] The technical solution of the present invention is: an adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on dynamic leakage factor, and the specific steps are as follows:

[0008] Get the baseband sampling signal x;

[0009] Initialize the weight vector w 0 =0, input signal vector u 0 is a zero vector of length L; and sets the leakage factor α;

[0010] The baseband sampling signal x is used as the input signal and the training sequence is used as the expected signal d. The adaptive dynamic update of the filter weight is achieved by adjusting the leakage factor in real time to complete the adaptive adjustment in the signal equalization process.

[0011] A further technical solution of the present invention is: the method for obtaining the baseband sampling signal x is to obtain the signal through the underwater acoustic communication receiving end, multiply the signal with the carrier after signal synchronization and resampling, and then process it through a low-pass filter to obtain the baseband sampling signal x.

[0012] A further technical solution of the present invention is: the process of adaptive dynamic updating of the filter weights is as follows:

[0013] Given an input signal x[n], a desired signal d[n], and a learning rate μ, repeat the following for n iterations:

[0014] Calculate the input vector u n =[x[n],x[n-1],…,x[n-L+1]] T , where L is the filter length;

[0015] Calculate the estimated error Where d[n] is the expected signal, w n is the current weight vector;

[0016] Calculate the scale factor Among them, q 0is the initial proportional constant, δ is the regularization factor used to prevent the denominator from being zero; a is the algorithm constraint control factor;

[0017] Calculate weight update And calculate the weight w n+1 =(1-α)w n +f, where α is the current leakage factor.

[0018] A further technical solution of the present invention is: the number of iterative calculations is equal to the length N of the input signal.

[0019] A further technical solution of the present invention is: the leakage factor α is set to 0.01.

[0020] A further technical solution of the present invention is: the adaptive adjustment mechanism of the leakage factor balances the forgetting rate in the weight update process by introducing a dynamic leakage factor and an algorithm constraint control factor to adapt to the rapid changes of the channel and effectively suppress the impact of the multipath effect.

[0021] An adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization system based on dynamic leakage factor, comprising an underwater acoustic communication receiving module, a signal processing module, and an adaptive filtering module;

[0022] Acquiring an underwater acoustic signal through the underwater acoustic communication receiving module;

[0023] The signal processing module is used to expand the signal frequency into the range of the carrier frequency, and then the signal is processed by a low-pass filter to obtain a baseband sampling signal;

[0024] The filter weights are adaptively and dynamically updated through the adaptive filtering module.

[0025] An electronic device comprises at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on dynamic leakage factor.

[0026] A computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on dynamic leakage factor when executed.

[0027] Beneficial Effects

[0028] The beneficial effect of the present invention is that the present invention proposes an adaptive equalization method based on a dynamic leakage factor to optimize the channel equalization performance of an underwater acoustic communication system. The present invention utilizes a dynamic leakage factor adjustment mechanism combined with a non-uniform power normalized minimum mean square error (Leaky IPNLMS) method to achieve adaptive dynamic update of filter weights by adjusting the leakage factor in real time, thereby effectively improving the equalization accuracy of the system in complex underwater acoustic channels, and can still maintain stable convergence characteristics under complex conditions such as channel multipath effects and high noise interference, thereby improving the stability and robustness of signal transmission. This method not only improves the anti-interference capability of the communication system, but also can adapt to the dynamic changes of the channel more quickly and accurately, so that the underwater acoustic communication system has better transmission performance and higher communication quality in different underwater environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 Comparison of the computational time of four algorithms: LMS, VSS-NLMS, NLMS, and Leaky IPNLMS.

[0030] Figure 2 This is the constellation diagram after equalization of the four algorithms: LMS, VSS-NLMS, NLMS and Leaky IPNLMS.

[0031] Figure 3 The bit error rate performance of four algorithms, LMS, VSS-NLMS, NLMS and Leaky IPNLMS, in different signal-to-noise ratio ranges. DETAILED DESCRIPTION

[0032] The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, but should not be construed as limiting the present invention.

[0033] Based on the problems that the channel characteristics of traditional algorithms change very quickly in underwater acoustic communication and the traditional algorithms are not able to adapt to the environment, the present invention provides an adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on dynamic leakage factor, and the specific steps are as follows:

[0034] Step 1: At the receiving end of underwater acoustic communication, after signal synchronization and resampling, the processed received signal is multiplied by the carrier and then processed by a low-pass filter to obtain a baseband sampling signal x, which is used as the input signal of the equalizer;

[0035] Step 2: Set the leakage factor α to 0.01 and initialize the weight vector w 0 =0, input signal vector u 0 is a zero vector of length L;

[0036] Step 3: Given the input signal x[n], the desired output d[n] and the learning rate μ, repeat the following for the nth iteration:

[0037] Calculate the input vector u n =[x[n],x[n-1],…,x[n-L+1]] T , where x[n] is the current input signal and L is the filter length;

[0038] Calculate the estimated error Where d[n] is the expected signal, w n is the current weight vector;

[0039] Calculate the scale factor Among them, q 0 is the initial proportional constant, δ is the regularization factor used to prevent the denominator from being zero; a is the algorithm constraint control factor;

[0040] Calculate weight update And calculate the weight w n+1 =(1-α)w n +f, where α is the current leakage factor;

[0041] The number of iterations above is equal to the total length N of the input signal.

[0042] The adaptive leakage factor adjustment mechanism is to balance the forgetting rate in the weight update process by introducing a dynamic leakage factor and an attenuation coefficient to adapt to the rapid changes of the channel and effectively suppress the impact of the multipath effect.

[0043] The present invention discloses an adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization system based on dynamic leakage factor, comprising an underwater acoustic communication receiving module, a signal processing module and an adaptive filtering module; an underwater acoustic signal is acquired through the underwater acoustic communication receiving module; the signal frequency is expanded to the range of the carrier frequency through the signal processing module, and then processed by a low-pass filter to obtain a baseband sampling signal; and the filter weight is adaptively and dynamically updated through the adaptive filtering module.

[0044] The present invention provides an electronic device, comprising at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on dynamic leakage factor.

[0045] The present invention provides a computer-readable data storage medium, wherein the computer-readable data storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on dynamic leakage factor when executed.

[0046] The present invention discloses an adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on dynamic leakage factor, which is used for signal equalization processing in underwater acoustic communication systems. The method introduces a dynamically adjustable leakage factor into an adaptive filtering algorithm, gradually attenuates the leakage factor according to changes in the signal environment, realizes stable update of weights and optimization of equalization effects, thereby effectively suppressing noise interference in complex underwater acoustic environments and improving signal transmission quality. The present invention is particularly suitable for underwater acoustic communication environments under multipath interference and high noise backgrounds, and provides stronger anti-interference capabilities and reliability guarantees for underwater acoustic communication systems.

[0047] The above technical solution is further described below in conjunction with the accompanying drawings:

[0048] Figure 1 The comparison of the operation time of four algorithms, LMS, VSS-NLMS, NLMS, and Leaky IPNLMS, is shown. As can be seen from the box plot, the operation time of the Leaky IPNLMS algorithm is significantly longer than that of the other three algorithms. The LMS algorithm has the smallest computation time, followed by VSS-NLMS and NLMS. The distribution of the operation time of Leaky IPNLMS is wider, and the variation from the lowest to the highest value is also larger. This may be due to the introduction of more complex dynamic leakage factors and weight update mechanisms in each iteration, which increases the computational burden. Although Leaky IPNLMS has the longest operation time, this cost may be to achieve higher accuracy and robustness in balancing performance. For applications that require real-time processing and are very sensitive to latency, the long computation time may be a disadvantage, but for scenarios that require higher balancing effects, this computational cost may be worthwhile.

[0049] Figure 2The constellation diagrams after equalization of the four algorithms, LMS, VSS-NLMS, NLMS and Leaky IPNLMS, are shown. It can be seen that the constellation diagram of the LMS algorithm has a loose distribution of scattered points and a large error, indicating that the algorithm is obviously insufficient in the process of equalizing the channel. The performance of the VSS-NLMS algorithm after equalization is better than that of the LMS, but the points are still not concentrated enough, indicating that the equalization effect of the signal is not good under high noise conditions. The constellation diagram of the NLMS algorithm performs slightly better, with more concentrated scattered points, but there is still a certain offset. The constellation diagram of the Leaky IPNLMS algorithm shows a very concentrated distribution, and the scattered points are almost in the ideal position, indicating that the algorithm can effectively reduce channel distortion and improve the equalization accuracy of the signal. The excellent performance of Leaky IPNLMS is attributed to its use of a dynamic leakage factor update mechanism, which can better adapt to channel characteristics and improve the robustness of the system.

[0050] Figure 3 The bit error rate performance of different algorithms in different signal-to-noise ratio intervals is shown. As can be seen from the figure, the bit error rate of the LMS algorithm almost completely falls within the highest bit error rate interval, which shows that it shows a poor equalization effect under different signal-to-noise ratio conditions. The VSS-NLMS algorithm has some improvements, but the proportion in the higher bit error rate interval is still large. The bit error rate performance of the NLMS algorithm is further optimized, and the coverage area of ​​the low bit error rate interval is larger, showing an improved resistance to noise. However, the Leaky IPNLMS algorithm occupies the highest proportion in the low bit error rate interval, showing a significant performance advantage. This means that Leaky IPNLMS is not only stable under high noise conditions, but also can maintain a low bit error rate in various signal-to-noise ratio environments, greatly improving the reliability of communication.

[0051] Based on the analysis of the above three charts, it can be concluded that the LMS algorithm has a simple structure and the shortest calculation time, but it performs poorly in terms of equalization performance and bit error rate, especially under conditions of large signal interference or complex channel characteristics. VSS-NLMS and NLMS have improved in terms of equalization effect and bit error rate, but they are still insufficient when facing more complex channels. Although the Leaky IPNLMS algorithm has a relatively long calculation time, it has obvious advantages in the constellation diagram and bit error rate performance after signal equalization. Its dynamic leakage factor adjustment mechanism can adaptively adjust the weights according to channel changes, significantly improving the equalization accuracy and robustness of the system. For underwater acoustic communication scenarios with high requirements for communication quality, the Leaky IPNLMS algorithm obviously has greater application potential and practical value. By using this method in complex underwater environments, the system's anti-noise ability and signal equalization accuracy can be significantly improved, thereby ensuring the stability and efficiency of communication.

[0052] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and modify the above embodiments within the scope of the present invention without departing from the principles and intent of the present invention.

Claims

1. An adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on dynamic leakage factor, characterized in that The specific steps are as follows: Get the baseband sampling signal x; Initialize the weight vector w0=0, the input signal vector u0 is a zero vector of length L; and set the leakage factor α; The baseband sampling signal x is used as the input signal and the training sequence is used as the expected signal d. The adaptive dynamic update of the filter weight is achieved by adjusting the leakage factor in real time to complete the adaptive adjustment in the signal equalization process.

2. According to claim 1, an adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on dynamic leakage factor is characterized by: The method for acquiring the baseband sampling signal x is to acquire the signal through the underwater acoustic communication receiving end, multiply the signal with the carrier after signal synchronization and resampling, and then process it through a low-pass filter to obtain the baseband sampling signal x.

3. According to claim 1, an adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on dynamic leakage factor, characterized in that: The process of adaptive dynamic updating of the filter weights is as follows: Given an input signal x[n], a desired signal d[n], and a learning rate μ, repeat the following for n iterations: Calculate the input vector u n =[x[n],x[n-1],…,x[n-L+1]] T , where L is the filter length; Calculate the estimated error Where d[n] is the expected signal, w n is the current weight vector; Calculate the scale factor Among them, q0 is the initial proportional constant, δ is the regularization factor used to prevent the denominator from being zero; a is the algorithm constraint control factor; Calculate weight update And calculate the weight w n+1 =(1-α)w n +f, where α is the current leakage factor.

4. According to claim 3, an adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on dynamic leakage factor is characterized by: The number of iterative calculations is equal to the length N of the input signal.

5. According to claim 4, an adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on dynamic leakage factor is characterized by: The leakage factor α is set to 0.

01.

6. According to claim 5, an adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on dynamic leakage factor is characterized by: The adaptive adjustment mechanism of the leakage factor balances the forgetting rate in the weight update process by introducing a dynamic leakage factor and an algorithm constraint control factor to adapt to the rapid change of the channel and effectively suppress the influence of the multipath effect.

7. An adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization system based on dynamic leakage factor, used to implement the adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on dynamic leakage factor according to any one of claims 1 to 6; characterized in that: It includes an underwater acoustic communication receiving module, a signal processing module, and an adaptive filtering module; Acquiring an underwater acoustic signal through the underwater acoustic communication receiving module; The signal processing module is used to expand the signal frequency into the range of the carrier frequency, and then the signal is processed by a low-pass filter to obtain a baseband sampling signal; The filter weights are adaptively and dynamically updated through the adaptive filtering module.

8. An electronic device, characterized in that: It includes at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on dynamic leakage factor as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the adaptive non-uniform power normalized minimum mean square error underwater acoustic communication equalization method based on dynamic leakage factor as described in any one of claims 1-6 when executed.

Images