Underwater self-interference suppression method

Abstract

The invention discloses an underwater self-interference suppression method, which is based on a fast dichotomy coordinate descent recursive least square algorithm FDCDRLS of a variable forgetting factor, and comprises the following steps: initializing algorithm parameters including the forgetting factor, the step length, the number of iterations, an initial error and an initial value of an autocorrelation matrix; calculating an instantaneous error value at the current moment by updating the input signal vector; calculating a forgetting factor at the current moment according to the instantaneous error value, wherein an arc tangent function relationship exists between the value of the forgetting factor and the square of the instantaneous error; updating the circular index sequence, and updating the autocorrelation matrix and the pseudo-error vector through the circular index sequence; updating a weight increment and a pseudo-error vector through a binary coordinate descent method; updating an estimated self-interference channel parameter at the current moment according to the weight increment; and repeating the steps until convergence, and finally obtaining an estimated value of the underwater self-interference channel.

Description

Technical Field

[0001] The invention belongs to the field of underwater acoustic communication detection, in particular to a digital domain self-interference suppression method in self-interference suppression of an underwater transceiver. Background Art

[0002] In recent years, with the continuous development of the field of hydroacoustics, the application of hydroacoustic technology in underwater communication, underwater detection, environmental monitoring and other fields has become increasingly widespread. Whether it is underwater full-duplex communication technology [1-3] that aims to save spectrum resources and improve communication efficiency, or underwater communication and detection integrated technology [4,5] that aims to reduce equipment size, reduce power consumption, and enhance concealment and safety, they all need to rely on the transceiver system. However, during the operation of the transceiver system, the strong power signal at the transmitter will directly interfere with the receiver, causing the desired signal at the far end to be submerged by the strong self-interference signal, thereby hindering the reception of the desired signal at the far end. Therefore, self-interference suppression has become a key issue that needs to be urgently solved in the transceiver system and has been widely studied and developed.

[0003] The underwater environment is complex and changeable. When the underwater transceiver is working, it is affected by the movement of the position and the change of the water environment, and the self-interference underwater acoustic multipath channel will change significantly. Therefore, it is necessary to estimate the self-interference channel parameters in real time to accurately reconstruct and eliminate the self-interference signal from the transmitter to the receiver, so as to extract the desired signal.

[0004] In the digital domain self-interference suppression method, an adaptive filtering algorithm is often used to realize the real-time estimation of the self-interference channel. The multipath delay channel model is highly compatible with the FIR lateral filter model, so the actual underwater self-interference channel model can be simulated with the help of an adaptive lateral filter. The transmitting end signal and the receiving end signal of the transceiver system are used as the reference signal and the expected signal of the adaptive filter respectively, and the self-interference system parameter estimation is performed through iterative calculation. According to the real-time feedback of the error after cancellation, the filter coefficient can be quickly corrected, and then the estimated filter coefficient can be made closer to the actual real-time changing self-interference channel, so as to realize the real-time reconstruction and cancellation of the self-interference signal, and help the signal at the receiving end to eliminate the interference signal from its own transmitting end, and then realize the extraction of the desired signal at the far end.

[0005] The adaptive channel estimation algorithms used in underwater transceivers mainly include the LMS algorithm and its improved algorithm, the RLS algorithm and its improved algorithm, the DCD_RLS algorithm and its improved algorithm [6]. Among them, the LMS algorithm and its improved algorithm perform iterative calculations by gradient descent. The algorithm operation complexity is low, but the channel estimation accuracy is low, the convergence speed is slow, and the tracking performance is poor. The RLS algorithm introduces a forgetting factor and obtains the least squares solution by iteratively updating the inverse matrix of the covariance matrix. The channel estimation accuracy is high and the convergence speed is fast, but the amount of calculation is large, and the operation complexity is O(N). 2 ), especially when the order N is high, it consumes a lot of computing resources and computing time. The DCD_RLS algorithm no longer uses matrix operations, but instead obtains the least squares solution through the binary coordinate descent method, which greatly reduces the amount of calculation, and the estimation accuracy is still high, but during the algorithm iteration process, each iteration still needs to update the autocorrelation matrix in its entirety, which consumes a lot of computing resources and still takes a long time. The FDCD_RLS algorithm is an improved algorithm based on the DCD_RLS algorithm. In each iteration, it does not update the entire autocorrelation matrix in its entirety, but only updates one element in the matrix through a ring index, thereby further reducing the computational complexity and improving the computing speed.

[0006] At the same time, in the RLS, DCD_RLS and FDCD_RLS algorithms, the forgetting factor λ has a huge impact on the estimation accuracy and convergence speed of the algorithm. The value of λ is usually between 0 and 1, which represents the influence weight of the past data at the input end on the channel parameter estimation. The closer λ is to 1, the higher the estimation accuracy, but the lower the convergence speed. The smaller λ is, the lower the algorithm estimation accuracy, but the convergence speed is faster, and it can better track the changing channel. In the case of a fixed forgetting factor, the estimation accuracy and convergence speed are often not achieved at the same time. Therefore, in the application scenario of eliminating self-interference signals of underwater transceivers, it is urgent to propose an algorithm with high calculation accuracy, fast convergence speed and low computational complexity to realize real-time estimation of complex and changeable underwater self-interference channels and realize real-time cancellation of self-interference signals.

[0007] [1]Qu F, Hong Y, Yu G, et al. ln-Band Full-Duplex Communications for Underwater Acoustic Networks[J]. IEEE Network, 2017; 31(5):59-65.

[0008] [2]KOLODZIEJ KE,PERRY BT,HERD J S.In-band full-duplex technology:techniques and systems survey[J].IEEE Transactions on Microwave Theory andTechniques,2019;67(7):3025-3041.

[0009] [3] Zhao Yunjiang, Qiao Gang, Liu Songzuo, et al. Research status and prospects of in-band full-duplex underwater acoustic communication technology. Digital Ocean and Underwater Attack and Defense, 2021; 4(3): 195-205.

[0010] [4] Lu Jun, Zhang Qunfei, Shi Wentao, et al. Research status and development trend of detection and communication integration. Signal Processing, 2019, 35(09): 1484-1495.

[0011] [5]Bao Z, Zhang Y, Tai Y, et al. Integration of detection and communication system designed for autonomous underwater vehicles in deepwater. Ocean Engineering, 2024; 310(1):118607

[0012] [6]Lu S, Henson B, Zakharov Y et al.Digital Self-InterferenceCancellation for Full-Duplex Underwater Acoustic Systems.IEEE Transactions onCircuits and Systems II:Express Briefs.2020;67(1):192-196. Summary of the invention

[0013] The purpose of the present invention is to overcome the deficiencies in the prior art and to provide an underwater self-interference suppression method based on a fast DCD_RLS algorithm with a variable forgetting factor (VFF_FDCD_RLS), so that the transceiver has both high steady-state accuracy and fast convergence tracking speed when performing self-interference signal suppression.

[0014] The objective of the present invention is achieved through the following technical solutions:

[0015] An underwater self-interference suppression method based on a binary coordinate descent recursive least squares algorithm FDCD_RLS with a variable forgetting factor includes:

[0016] Initialize algorithm parameters, including forgetting factor, step size, number of iterations, initial error, and initial value of autocorrelation matrix;

[0017] By updating the input signal vector, the instantaneous error value at the current moment is calculated;

[0018] The forgetting factor at the current moment is calculated according to the instantaneous error value. There is an inverse tangent function relationship between the value of the forgetting factor and the square of the instantaneous error.

[0019] Performing a circular index sequence update, and updating an autocorrelation matrix and a pseudo error vector through the circular index sequence;

[0020] Update the weight increment and pseudo error vector by bisection coordinate descent method;

[0021] Update the estimated self-interference channel parameters at the current moment according to the weight increment;

[0022] Repeat the above steps until convergence, and finally obtain the estimated value of the underwater self-interference channel.

[0023] Furthermore, the initial value of the autocorrelation matrix is:

[0024]

[0025] δ is the preset initial value;

[0026] The input signal vector x(n) is a one-dimensional vector of order N

[0027] x(n)=[u(n),u(n-1),…,u(n-N+1)] T

[0028] Calculate the instantaneous error value e(n) at the current moment:

[0029] e(n)=d(n)-w T (n-1)x(n)

[0030] Where d(n) is the receiving end signal. In the scenario of a transceiver-in-one system, d(n) includes the self-interference signal from the transmitter, the far-end desired signal, and the environmental noise.

[0031] Furthermore, the calculation formula of the forgetting factor λ is as follows:

[0032] λ=λ mid -Aacrtan(B(e 2 -E))

[0033] The value of the forgetting factor λ is dynamically adjusted according to the square of the instantaneous error e at the previous moment, where λ mid is the middle value of the forgetting factor, A is the variable amplitude of the forgetting factor, parameter B controls the slope of the relationship curve between the forgetting factor and the instantaneous error e, that is, the rapidity of the change of the forgetting factor, and parameter E is the average value of the square of the instantaneous error; when e 2 When E, the forgetting factor λ=λ mid .

[0034] Furthermore, the circular index sequence is set to I(N), and its update method is:

[0035] I(i)=mod(I(i-1)-1,Ni)

[0036] For each row i∈{1,2,…,N}, mod is the remainder function, and the value of mod(a,b) is the remainder after a is divided by b. When a=0 or a=b, its mod value is b;

[0037] Then the autocorrelation matrix R is updated according to the ring index sequence. In each row i, only the elements of the corresponding column of the ring index sequence are updated. The updating method is:

[0038] R(i,I(i))=λR(i,I(i-1))+x(1)·x(i)

[0039] Then the pseudo error vector β(n) is updated:

[0040] β(n)=λβ(n-1)+e(n)x(n)

[0041] λ is the forgetting factor, e(n) is the instantaneous error value, and x(n) is the input signal vector.

[0042] Further, updating the weight increment and the pseudo error vector by the bisection coordinate descent method includes: adjusting the estimated self-interference channel parameters at each moment by iteratively calculating the step size h and the weight increment Δw to gradually reduce the error of the pseudo error vector.

[0043] Furthermore, according to the dynamic changes of the underwater self-interference channel, the forgetting factor can be reduced to accelerate convergence when the channel changes rapidly, and the forgetting factor can be increased to improve the estimation accuracy when the channel is stable, thereby ensuring efficient self-interference signal suppression capability.

[0044] The present invention also provides an application of an underwater self-interference suppression method, which is applied to an underwater transceiver integrated system. During underwater communication or detection, real-time self-interference signal suppression is performed through digital domain adaptive filtering to improve the reception quality of the remote desired signal.

[0045] The present invention also provides a self-interference suppression system for an underwater transceiver, comprising:

[0046] The signal acquisition module is used to simultaneously acquire the transmitting end signal and the receiving end signal of the underwater transceiver; the signal acquisition module includes the transmitting end signal acquisition after the power amplifier or the signal acquisition after the power amplifier nonlinear compensation;

[0047] An adaptive filtering module, used to estimate the self-interference channel of the collected signal by the underwater self-interference suppression method described in any one of claims 1 to 6, and generate a signal after cancellation;

[0048] The processing module is used to update the self-interference channel parameters in real time according to the dynamically adjusted forgetting factor to achieve signal cancellation.

[0049] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the underwater self-interference suppression method when executing the program.

[0050] The present invention also provides a computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the underwater self-interference suppression method are implemented.

[0051] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0052] 1. Reduce computational complexity and improve computational speed: Traditional RLS and DCD_RLS algorithms require global updates of the entire autocorrelation matrix, which has high computational complexity, especially in the case of high-order filters, where computational overhead and time consumption are high. The present invention introduces a circular index mechanism to update only a single element in the autocorrelation matrix, thereby avoiding the computational overhead of globally updating the entire matrix. This optimization effectively reduces the amount of computation, thereby significantly improving the computational speed and reducing the consumption of computing resources. In the application scenario of high-order filters, compared with traditional algorithms, the present invention can quickly adapt to dynamically changing underwater environments while ensuring high precision, thereby improving real-time processing capabilities.

[0053] 2. Adaptive performance optimization: The present invention introduces a variable forgetting factor mechanism to dynamically adjust the value of the forgetting factor according to the size of the instantaneous error.

[0054] When the channel changes rapidly, the algorithm can automatically reduce the forgetting factor, thereby accelerating convergence and quickly tracking the changing channel parameters; when the channel is stable, the forgetting factor increases, thereby improving the steady-state estimation accuracy and optimizing the suppression effect of self-interference signals.

[0055] This design achieves a dynamic balance between estimation accuracy and convergence speed, enabling the algorithm to provide excellent adaptive performance under various environmental conditions, especially adapting to the changeable underwater environment, and solving the problem that the fixed forgetting factor algorithm cannot take into account both accuracy and convergence speed.

[0056] 3. Improve channel estimation accuracy: By optimizing the weight update strategy and dynamically adjusting the variable forgetting factor, the present invention can accurately estimate the multipath channel in a complex underwater environment.

[0057] This algorithm makes the channel estimation result closer to the real channel parameters, providing a solid foundation for the subsequent reconstruction and elimination of interference signals. This advantage ensures the efficient removal of self-interference signals during the suppression process, improves the quality of received signals, and effectively enhances the reliability and stability of signals.

[0058] 4. Enhance environmental adaptability:

[0059] The underwater environment is usually highly dynamic, including changes in water flow, movement of target objects, and fluctuations in noise levels, which lead to rapid changes in channel parameters.

[0060] The present invention dynamically adjusts the forgetting factor so that the algorithm can respond quickly when the channel changes dramatically and maintain efficient interference suppression capabilities. When the environment is stable, the algorithm can also maintain a high estimation accuracy and enhance the adaptability to complex and changing environments. Compared with the traditional fixed forgetting factor algorithm, the present invention can provide more stable and accurate signal processing capabilities in more complex scenarios, meeting the needs of underwater communication and detection systems.

[0061] 5. Improve system stability and real-time performance:

[0062] Compared with the traditional DCD_RLS and FDCD_RLS algorithms, the VFF_FDCD_RLS algorithm of the present invention can quickly estimate the self-interference channel when the signal changes greatly, and eliminate the self-interference signal with high real-time performance and accuracy, avoiding the delay phenomenon that may occur in the traditional algorithm. The algorithm can maintain stable performance under different working conditions, improve the signal processing capability of the underwater transceiver system, and reduce the computational burden of the system.

[0063] 6. Strong system integration:

[0064] The algorithm of the present invention not only has high computational efficiency and strong adaptability, but also can be well integrated with the existing underwater transceiver system. Its design based on variable forgetting factor and binary coordinate descent method can adapt to different working environments and channel conditions, and provide efficient and reliable technical support for underwater communication, detection, monitoring and other fields.

[0065] In summary, the technical solution of the present invention has significant advantages over the prior art, can effectively reduce the computational complexity, improve the algorithm convergence speed and accuracy, enhance the system adaptability, and significantly improve the performance and real-time performance of the underwater self-interference suppression system. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 The implementation steps of the fast DCD_RLS algorithm (VFF_FDCD_RLS) based on the variable forgetting factor in this embodiment are shown.

[0067] Figure 2 The selection rule of the forgetting factor is shown, that is, the relationship function curve between the forgetting factor and the instantaneous error.

[0068] Figure 3 The figure shows the comparison of self-interference channel estimation results of VFF_DCD_RLS and FDCD_RLS algorithm with fixed forgetting factor under simulation test.

[0069] Figure 4 The application scenario of self-interference suppression of underwater transceiver is shown. DETAILED DESCRIPTION

[0070] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0071] Example 1

[0072] This embodiment provides a self-interference signal suppression method based on a fast DCD_RLS algorithm (VFF_FDCD_RLS) with a variable forgetting factor in underwater self-interference suppression. The algorithm (VFF_FDCD_RLS) can adjust the forgetting factor of the algorithm in real time according to the size of the instantaneous error, thereby controlling the self-interference channel estimation accuracy and convergence speed of the algorithm in iterative operation. When the underwater self-interference channel changes, the algorithm can achieve rapid convergence by reducing the forgetting factor and track the self-interference channel parameters that change in real time. When the environment remains stable, the algorithm can improve the channel estimation accuracy by increasing the forgetting factor and improve the effect of self-interference signal suppression. At the same time, the algorithm is an improved algorithm based on the FDCD_RLS algorithm. In each iteration, the entire autocorrelation matrix is ​​not updated, but only one element in the matrix is ​​updated by a ring index, and then the new autocorrelation matrix R is used to perform the binary coordinate descent method to solve the least squares solution. Compared with the DCD_RLS algorithm, the amount of calculation is further reduced and the calculation complexity is reduced.

[0073] For each time n, it is known that the objective functions of the traditional RLS algorithm and the DCD_RLS algorithm are

[0074]

[0075] Similarly, in the FDCD_RLS algorithm, the objective function is also this formula. Similar to the DCD_RLS algorithm, this embodiment introduces a pseudo error vector to iteratively update and solve the change vector of the filter coefficient, where the pseudo error vector is

[0076] β(n)=Rw(n) (2)

[0077] It is known that in the process of solving the least squares solution,

[0078] w(n)=w(n-1)+K(n)e(n) (3)

[0079] Where K(n) = R -1 (n)x(n) is the gain vector. Substituting it into the above formula, we can get

[0080] R(n)w(n)=R(n-1)w(n-1)+x(n)d(n) (4)

[0081] According to the recursive formula of the autocorrelation matrix, R(n)=λR(n-1)+x(n)x T (n) The recursive formula of the pseudo error vector can be obtained as follows:

[0082] β(n)=λβ(n-1)+e(n)x(n) (5)

[0083] Then, Δw is adjusted by the bisection coordinate descent method to reduce β(n) as much as possible, thereby obtaining the optimal weight change Δw for updating the estimated self-interference channel parameter (weight) w at the current moment, and obtaining a new pseudo error vector β(n).

[0084] The implementation process of the VFF_FDCD_RLS algorithm involved in this embodiment is similar to the basic principle of the DCD_RLS algorithm implementation process, but the update of the autocorrelation matrix R and the update of the pseudo error vector β(n) in the bisection coordinate descent process are different. The implementation steps are as follows: Figure 1 , as follows:

[0085] The first step is to initialize the algorithm parameters. Set the forgetting factor, step size, number of iterations, initial error, and the initial value of the autocorrelation matrix. The initial value of the autocorrelation matrix is:

[0086]

[0087] δ is the preset initial value.

[0088] The second step is to update the input signal vector x(n), which is an N-order one-dimensional vector

[0089] x(n)=[u(n),u(n-1),…,u(n-N+1)] T (7)

[0090] Calculate the instantaneous error value e(n) at the current moment,

[0091] e(n)=d(n)-w T (n-1)x(n) (8)

[0092] Where d(n) is the receiving end signal. In the scenario of a transceiver-in-one system, d(n) includes the self-interference signal from the transmitter, the far-end desired signal, and the environmental noise.

[0093] The third step is to calculate the forgetting factor value at the current moment based on the instantaneous error.

[0094] λ=λ mid -Aacrtan(B(e 2 -E)) (9)

[0095] The value of the forgetting factor λ is completely based on the instantaneous error e at the previous moment, and its value is related to the square of the instantaneous error e by an inverse tangent function, where λ mid is the middle value of the forgetting factor, A is the variable amplitude of the forgetting factor, parameter B controls the slope of the relationship curve between the forgetting factor and the instantaneous error e, that is, the rapidity of the change of the forgetting factor, and parameter E is the average value of the square of the instantaneous error, which also controls the left and right shift of the function curve. 2 When it is E, the forgetting factor λ is just λ mid The relationship between the forgetting factor and the instantaneous error e is as follows Figure 2 shown.

[0096] The fourth step is to update the ring index array, the autocorrelation matrix R, and the pseudo error vector. For each row i∈{1,2,…,N}, the position of the current row is determined by the ring index. The ring index sequence is set to I(N), and its update method is

[0097] I(i)=mod(I(i-1)-1,Ni) (10)

[0098] Where mod is the remainder function, and the value of mod(a,b) is the remainder after a is divided by b. If a=0 or a=b, its mod value is b.

[0099] Then the autocorrelation matrix is ​​updated according to the ring index sequence, and only the elements of the corresponding ring index sequence column are updated in each row. The update method is

[0100] R(i,I(i))=λR(i,I(i-1))+x(1)·x(i) (11)

[0101] Then the pseudo error vector update is performed,

[0102]

[0103] The fifth step is to solve the weight increment Δw and the new pseudo-error vector β(n) by the bisection coordinate descent method. The specific implementation method is to iteratively update the weight increment coefficient corresponding to the position with the largest absolute value in the pseudo-error vector according to the preset number of iterations. The implementation steps are:

[0104] (1) Initialize the step size and weight increment Δw.

[0105] (2) Find the position of the current maximum residual value and the position of the corresponding index sequence. Let the position be p (p∈{1,2…N}) and calculate p':

[0106] p'=mod(I(1)+p-1,N) (13)

[0107] I(1) represents the column index currently used by the first row of the autocorrelation matrix R. The value of p' is to find the update position index corresponding to the weight w(p) in the first row of the autocorrelation matrix R. Because R uses circular storage, the correct position needs to be found through the circular index when updating.

[0108] (3) Calculate the update step size h. Calculate the step size h by the binary coordinate descent method so that it can meet the following conditions. If not, divide h by 2 until the following formula is met or the preset binary upper limit is reached.

[0109] |β(p)|>=hR(1,p') (14)

[0110] (4) Update the weight increment Δw and the new pseudo error vector β(n).

[0111] Δw(p)=Δw(p)+sign(β(p))·h (15)

[0112] The sign function is a sign function, which represents the update of Δw. Then, the pseudo error vector β(n) is updated iteratively. Here, the update of β(n) is updated one by one in an iterative manner according to the ring index.

[0113] β(i)=β(i)-sign(β(p))R(j,mod(I(j)+k,Mj)) (16)

[0114] Here i∈{1,2,…,N}, j is the ring index coordinate when the position of the maximum residual error is p, and its change rule is to decrease from p to 0 and then continue to increase to Np, and the change rule of k is to increase from one to j to 0 and then remain unchanged. The above is the update method of the N-order pseudo-error vector β(n) under the ring index. At this point, the weight increment Δw and the new pseudo-error vector β(n) have been solved, and the bisection coordinate descent process is completed.

[0115] In the sixth step, the filter coefficients at this moment are updated according to the weight increment Δw obtained by the bisection coordinate descent method, that is, the estimated self-interference channel parameters. The estimated self-interference channel parameters w(n) are:

[0116] w(n)=w(n-1)+Δw (17)

[0117] At this point, the iteration ends, the input data vector continues to be updated, and the next iteration is entered to solve w so that it can be close to the actual self-interference channel.

[0118] Example 2

[0119] This embodiment provides a self-interference suppression system for an underwater transceiver, including:

[0120] The signal acquisition module is used to simultaneously acquire the transmitting end signal and the receiving end signal of the underwater transceiver; the signal acquisition module includes the transmitting end signal acquisition after the power amplifier or the signal acquisition after the power amplifier nonlinear compensation;

[0121] An adaptive filtering module, used to estimate the self-interference channel of the collected signal by the underwater self-interference suppression method described in Example 1, and generate a signal after cancellation;

[0122] The processing module is used to update the self-interference channel parameters in real time according to the dynamically adjusted forgetting factor to achieve signal cancellation.

[0123] Application scenarios of underwater transceiver self-interference suppression system Figure 4 In the application scenario of underwater self-interference suppression, since the self-interference signal of the transceiver will cause strong interference to the receiving end when it is communicating or detecting underwater, and the self-interference channel parameters will change due to changes in the underwater environment, the self-interference signal needs to be reconstructed and filtered out in the digital domain by adaptive filtering in the signal collected by the receiving end, so as to realize the extraction of useful signals.

[0124] The specific implementation steps are:

[0125] First, when the underwater transceiver is transmitting and receiving signals at the same time, the transmitting end signal and the receiving end signal are collected at the same time to the digital domain for the next step of processing. In order to avoid the nonlinear distortion caused by the power amplifier after amplifying the transmitting signal, the transmitting end signal collected by the signal acquisition module can collect the signal after the power amplifier transmits, or use the digital signal before transmission to compensate for the nonlinearity of the power amplifier.

[0126] Furthermore, at the digital domain end, the order of the digital filter is set by the channel length of the working environment, the number of iterations of the VFF_FDCD_RLS algorithm at a single moment is set by testing the sparsity of the channel, and the appropriate forgetting factor variation function λ=λ is determined by testing parameters such as the working environment noise level. mid -A·acrtan(B·(e 2 -E)) parameters, including the intermediate value of the forgetting factor λ mid , change range A, error square shift range value E, and forgetting factor change slope parameter B. At the same time, it is also necessary to set the algorithm initialization input vector, pseudo error vector β(n), autocorrelation matrix R, and step size and other parameters.

[0127] Furthermore, the collected transmitter signal and receiver signal are respectively used as the reference signal x(n) and the expected signal d(n) of the adaptive filter and input into the VFF_FDCD_RLS algorithm in the adaptive filtering module, and each time n is traversed. At each time, the N-order input signal vector x(n) = [u(n), u(n-1), ..., u(n-N+1)] is updated. T , taking x(n) and d(n) as the input signals of the adaptive filter, performing iterative operations at each moment.

[0128] In each iteration, the input signal vector x(n) is convolved with the latest estimated self-interference channel parameter w obtained at the previous moment, and then subtracted from the expected signal d(n) to obtain the instantaneous error value e at that moment. Then, the forgetting factor λ and pseudo-error vector β(n) at the current moment are updated according to the instantaneous error value.

[0129] Further, a circular index operation is performed to update the circular index vector I(n), and then the autocorrelation matrix is ​​updated through the circular index vector. The obtained new autocorrelation matrix R and the error vector β(n) are used as the input signals of the binary coordinate descent function, and the circular index vector I(n) is used as the auxiliary signal to calculate the weight increment Δw of the estimated self-interference channel parameter (weight) w at the current moment and the new pseudo-error vector β(n). The calculation method is the binary coordinate descent method in the technical solution, which obtains the optimal step size by binary step size, and updates the weight increment Δw at the position with the largest absolute value in the pseudo-error vector, and updates the pseudo-error vector β(n) one by one again through the circular index method.

[0130] Finally, the obtained weight increment Δw is used to correct the filter coefficient value at the current moment, that is, to estimate the self-interference channel parameter w, and update the new pseudo error vector β(n), and then enter the next iteration loop.

[0131] After the above steps, the w obtained at each moment is the estimated value of the underwater acoustic channel h at the current moment, and the accuracy of the estimation directly affects the elimination effect of the self-interference signal. The error signal e obtained is the signal after digital cancellation at the receiving end. After adaptive cancellation, e only contains the far-end desired signal, noise signal and residual self-interference signal. Most of the self-interference signals in the received signal are eliminated, and the signal after cancellation can be further analyzed to extract relevant information in the signal.

[0132] Preferably, the embodiments of the present application also provide a specific implementation of an electronic device capable of implementing all steps in the underwater self-interference suppression method in the above embodiments, and the electronic device specifically includes the following contents:

[0133] Processor, memory, communications interface and bus;

[0134] Among them, the processor, memory, and communication interface communicate with each other through the bus; the communication interface is used to realize information transmission between related devices such as server-side devices, metering devices, and user-side devices.

[0135] The processor is used to call the computer program in the memory, and when the processor executes the computer program, all the steps in the underwater self-interference suppression method in the above embodiment are implemented.

[0136] An embodiment of the present application also provides a computer-readable storage medium capable of implementing all the steps of the underwater self-interference suppression method in the above-mentioned embodiment. A computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, all the steps of the underwater self-interference suppression method in the above-mentioned embodiment are implemented.

[0137] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the hardware + program embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0138] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0139] Although the present application provides method operation steps such as embodiments or flow charts, more or fewer operation steps may be included based on conventional or non-creative labor. The order of steps listed in the embodiments is only one way of executing the order of many steps and does not represent the only execution order. When the actual device or client product is executed, it can be executed in the order of the method shown in the embodiments or the drawings or in parallel (for example, in a parallel processor or multi-threaded processing environment).

[0140] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0141] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0142] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0143] The present invention is not limited to the embodiments described above. The above description of the specific embodiments is intended to describe and illustrate the technical solution of the present invention. The above specific embodiments are merely illustrative and not restrictive. Without departing from the scope of the present invention and the scope of protection of the claims, a person of ordinary skill in the art can also make many forms of specific changes under the guidance of the present invention, which all fall within the scope of protection of the present invention.

Claims

1. An underwater self-interference suppression method based on a fast binary coordinate descent recursive least squares algorithm FDCD_RLS with a variable forgetting factor, characterized in that: include: Initialize algorithm parameters, including forgetting factor, step size, number of iterations, initial error, and initial value of autocorrelation matrix; By updating the input signal vector, the instantaneous error value at the current moment is calculated; The forgetting factor at the current moment is calculated according to the instantaneous error value. There is an inverse tangent function relationship between the value of the forgetting factor and the square of the instantaneous error. Performing a circular index sequence update, and updating an autocorrelation matrix and a pseudo error vector through the circular index sequence; Update the weight increment and pseudo error vector by bisection coordinate descent method; Update the estimated self-interference channel parameters at the current moment according to the weight increment; Repeat the above steps until convergence, and finally obtain the estimated value of the underwater self-interference channel.

2. The underwater self-interference suppression method according to claim 1, characterized in that: The initial values ​​of the autocorrelation matrix are: δ is the preset initial value; The input signal vector x(n) is a one-dimensional vector of order N x(n)=[u(n),u(n-1),…,u(n-N+1)] T Calculate the instantaneous error value e(n) at the current moment: e(n)=d(n)-w T (n-1)x(n) Where n is the time, w is the estimated self-interference channel parameter at the current time, T is the vector transpose; d(n) is the receiving end signal. In the scenario of the transceiver integrated system, d(n) includes the self-interference signal from the transmitter, the far-end desired signal, and the environmental noise.

3. The underwater self-interference suppression method according to claim 1, characterized in that: The calculation formula of the forgetting factor λ is as follows: λ=λ mid -Aacrtan(B(e 2 -E)) The value of the forgetting factor λ is dynamically adjusted according to the square of the instantaneous error e at the previous moment, where λ mid is the middle value of the forgetting factor, A is the variable amplitude of the forgetting factor, parameter B controls the slope of the relationship curve between the forgetting factor and the instantaneous error e, that is, the rapidity of the change of the forgetting factor, and parameter E is the average value of the square of the instantaneous error; when e 2 When E, the forgetting factor λ=λ mid .

4. The underwater self-interference suppression method according to claim 1, characterized in that: The circular index sequence is set to I(N), and its update method is: I(i)=mod(I(i-1)-1,Ni) For each row i∈{1,2,…,N}, mod is the remainder function, and the value of mod(a,b) is the remainder after a is divided by b. When a=0 or a=b, its mod value is b; Then the autocorrelation matrix R is updated according to the ring index sequence. In each row i, only the elements of the corresponding column of the ring index sequence are updated. The updating method is: R(i,I(i))=λR(i,I(i-1))+x(1)·x(i) Then the pseudo error vector β(n) is updated: β(n)=λβ(n-1)+e(n)x(n) λ is the forgetting factor, e(n) is the instantaneous error value, and x(n) is the input signal vector.

5. The underwater self-interference suppression method according to claim 1, characterized in that: Updating the weight increment and the pseudo error vector by the bisection coordinate descent method includes: adjusting the estimated self-interference channel parameters at each moment by iteratively calculating the step size h and the weight increment Δw to gradually reduce the error of the pseudo error vector.

6. The underwater self-interference suppression method according to claim 1, characterized in that: According to the dynamic changes of the underwater self-interference channel, the forgetting factor can be reduced to accelerate convergence when the channel changes rapidly, and the forgetting factor can be increased to improve the estimation accuracy when the channel is stable, thereby ensuring efficient self-interference signal suppression capability.

7. An application of an underwater self-interference suppression method, based on the underwater self-interference suppression method according to any one of claims 1 to 6, characterized in that: Applied to underwater transceiver systems, during underwater communication or detection, real-time self-interference signal suppression is performed through digital domain adaptive filtering to improve the reception quality of the desired signal at the far end.

8. An underwater transceiver self-interference suppression system, characterized in that: include: A signal acquisition module, used to simultaneously acquire the transmitting end signal and the receiving end signal of the underwater transceiver; The signal acquisition module includes the signal acquisition of the transmitting end after the power amplifier or the signal acquisition after the nonlinear compensation of the power amplifier; An adaptive filtering module, used to estimate the self-interference channel of the collected signal by the underwater self-interference suppression method described in any one of claims 1 to 6, and generate a signal after cancellation; The processing module is used to update the self-interference channel parameters in real time according to the dynamically adjusted forgetting factor to achieve signal cancellation.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the underwater self-interference suppression method according to any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the underwater self-interference suppression method according to any one of claims 1 to 6 are implemented.

Images