Self-interference cancellation method and system based on variable step size LMS algorithm of hyperbolic tangent function

Abstract

The present invention proposes a self-interference cancellation method and system based on a variable-step-size LMS algorithm using a hyperbolic tangent function. The method proposes a simultaneous and co-frequency self-interference cancellation model based on the digital domain. The method uses an improved hyperbolic tangent function to establish a nonlinear relationship between error and step size, sets a step size control parameter, and adjusts the value of the improved hyperbolic tangent function step size according to the error value, thereby achieving dynamic parameter adjustment: when the error is large in the initial stage of system convergence, a larger step size value is used to increase the convergence speed; when the error is small in the steady-state working range, a smaller step size value is used to reduce the steady-state error. The present invention can select different step size control parameters according to different engineering scenarios to effectively increase the convergence speed and achieve a smaller steady-state error, thereby achieving digital domain self-interference cancellation.

Description

Technical Field

[0001] The present invention relates to the field of digital signal processing technology and wireless communication transmission, and in particular to a self-interference cancellation method and system based on a hyperbolic tangent function variable step size LMS algorithm. Background Art

[0002] With the rapid development of wireless communication technology, ensuring the security and reliability of information transmission has become increasingly critical. Simultaneous co-channel communication (SCC) technology, with its significant spectrum efficiency advantages, can effectively alleviate the conflict between limited spectrum resources and the explosive growth of communication service demand. However, SCC implementation suffers from severe self-interference issues. Digital domain self-interference cancellation solutions achieve precise cancellation by constructing self-interference signal models. Common solutions are adaptive filtering techniques based on the Least Mean Square Error (LMS) algorithm and the Recursive Least Square (RLS) algorithm. The LMS algorithm minimizes the mean square error (MSE) as its optimization objective, combining Wiener filter theory with the steepest gradient descent optimization strategy to find the optimal value. With its advantages of low computational complexity and strong robustness, it has been widely used for digital domain self-interference cancellation in SCC systems.

[0003] During the iterative update of the LMS algorithm's weight coefficients, the filter weight coefficients are dynamically adjusted along the negative gradient of the error surface. The accuracy of the gradient value estimation directly affects the accuracy of the algorithm. Existing LMS algorithms use a fixed step size parameter design, which makes it difficult to effectively address the challenges posed by the time-varying characteristics of self-interference channels: while a larger step size setting can increase the convergence rate, it significantly reduces accuracy; while a smaller step size improves convergence accuracy, it prolongs the convergence process. In particular, when the step size parameter exceeds a reasonable range, the gradient estimation error will cause the weight coefficients to diverge, preventing the adaptive filter weight coefficients from converging. It can be seen that the existing LMS algorithm suffers from the drawback that two important performance characteristics, convergence rate and steady-state error, cannot be improved simultaneously. Therefore, when selecting the step size, the LMS algorithm must comprehensively consider both convergence accuracy and speed. Although many variable-step-size LMS algorithms have been proposed to improve the poor self-interference cancellation performance caused by the fixed step size of the LMS algorithm, most of these variable-step-size LMS algorithms currently have problems with unreliable performance or high computational complexity in actual application scenarios. The main reason is that these algorithms rely on many parameters that need to be manually adjusted, and manual parameter adjustment requires a lot of practice and has poor adaptability in different environments. Summary of the Invention

[0004] To overcome the drawbacks of existing LMS algorithms, which cannot simultaneously improve two important performance characteristics: convergence speed and steady-state error, and the inability to adaptively adjust the step-size control parameters of existing variable-step-size LMS algorithms, the present invention provides a self-interference cancellation method and system for a variable-step-size LMS algorithm based on a hyperbolic tangent function (HT-VSS-LMS). This method proposes a simultaneous and co-frequency self-interference cancellation model in the digital domain, utilizes an improved hyperbolic tangent function to establish a nonlinear relationship between error and step-size, and sets step-size control parameters and an adaptive update principle for the step-size control parameters, adjusting the value of the improved hyperbolic tangent function step-size based on the error value. This method allows for both manual selection of different step-size control parameters based on different engineering scenarios, and adaptive selection of appropriate step-size control parameters through an adaptive step-size control parameter update strategy, effectively improving convergence speed and minimizing steady-state error, thereby achieving digital-domain self-interference cancellation.

[0005] The technical solution of the present invention is:

[0006] A self-interference cancellation method based on a hyperbolic tangent function variable step size LMS algorithm comprises the following steps:

[0007] Step 1: Obtain the transmitted signal for simultaneous and co-frequency communication and receive signals , Indicates the points;

[0008] Step 2: According to the formula

[0009]

[0010] Calculating the self-interference signal ,in is the weight coefficient;

[0011] Step 3: According to the formula

[0012]

[0013] Calculate the error signal This is the received signal after self-interference cancellation;

[0014] Step 4: Update the formula based on the weight coefficient

[0015]

[0016] The weight coefficients are updated iteratively until the iteration termination condition is reached; For the The weight coefficient of each point, is the step length, according to the formula

[0017]

[0018] Get, among them 、 、 Both are control parameters that control the change of step size.

[0019] Furthermore, the step size formula is adjusted to:

[0020]

[0021] in For the The error signal of each point.

[0022] Furthermore, the step size formula is adjusted to:

[0023]

[0024] in is the step size feedback factor.

[0025] Furthermore, the step size control parameters are designed 、 、 Adaptive update strategy:

[0026] When the error is large when using the exponential moving average method Increase, speed up the convergence; when the error tends to be stable, Reduce and enhance steady-state performance:

[0027]

[0028] in, is the smoothing coefficient, ρ ∈ [ 0 , 1 ] , is the scaling factor;

[0029] Adaptively adjust the function curvature parameter using error variance :

[0030]

[0031] in, is the mapping ratio, , is the sliding window length, L ∈ [ 10 , 30 ] , is the average error within the sliding window, is the mapping ratio control parameter, Set to the minimum value to prevent ;

[0032] Optimize response speed to accelerate exponential contraction effect and improve dynamic adaptability. Set to:

[0033]

[0034] After introducing the above step size control parameter update adaptive update strategy, the step size formula expression can be obtained as follows:

[0035] .

[0036] Furthermore, the step size feedback factor is based on the formula

[0037]

[0038] Calculated, where is the setting coefficient, For the The step size feedback factor for each point.

[0039] Furthermore, the weight coefficient update formula is adjusted to

[0040]

[0041] in To set a constant.

[0042] In addition, the present invention also proposes a self-interference cancellation system based on a variable step-size LMS algorithm of a hyperbolic tangent function, wherein the adaptive filter weight coefficient is , The update formula is

[0043]

[0044] in For simultaneous and co-frequency communication, The receiving signal for simultaneous and same-frequency communication The self-interference signal output by the adaptive filter The error signal between them, that is, the received signal after self-interference cancellation, For the The weight coefficient of each point, is the step length, according to the formula

[0045]

[0046] Get, among them 、 、 Both are control parameters that control the change of step size.

[0047] Furthermore, the step size formula is adjusted to:

[0048]

[0049] in For the The error signal of each point.

[0050] Furthermore, the step size formula is adjusted to:

[0051]

[0052] in is the step size feedback factor.

[0053] Furthermore, the step size control parameters are designed 、 、 Adaptive update strategy:

[0054] When the error is large when using the exponential moving average method Increase, speed up the convergence; when the error tends to be stable, Reduce and enhance steady-state performance:

[0055]

[0056] in, is the smoothing coefficient, ρ ∈ [ 0 , 1 ] , is the scaling factor;

[0057] Adaptively adjust the function curvature parameter using error variance :

[0058]

[0059] in, is the mapping ratio, , is the sliding window length, L ∈ [ 10 , 30 ] , is the average error within the sliding window, is the mapping ratio control parameter, Set to the minimum value to prevent ;

[0060] Optimize response speed to accelerate exponential contraction effect and improve dynamic adaptability. Set to:

[0061]

[0062] After introducing the above step size control parameter update adaptive update strategy, the step size formula expression can be obtained as follows:

[0063] .

[0064] Furthermore, the step size feedback factor is based on the formula

[0065]

[0066] Calculated, where is the setting coefficient, For the The step size feedback factor for each point.

[0067] Furthermore, the weight coefficient update formula is adjusted to

[0068]

[0069] in To set a constant.

[0070] Beneficial effects:

[0071] Compared with the existing LMS algorithm self-interference cancellation technology, the present invention has the following advantages:

[0072] 1. First, the present invention utilizes an improved hyperbolic tangent function Constructing the error signal With step length The nonlinear relationship between the improved hyperbolic tangent function and Symmetrical relative to a certain value, when the error signal When the value of is large (that is, the initial stage of the algorithm), the corresponding step size is also large, which can speed up the convergence of the algorithm. The value tends to (That is, the algorithm converges and stably approaches the actual communication signal to be received ), the corresponding step value The improved hyperbolic tangent function complies with the step size selection principle of the optimized LMS algorithm, so that the present invention overcomes the defect that the convergence speed and steady-state error of the existing LMS algorithm cannot be improved simultaneously, and has a higher convergence speed and a smaller steady-state error.

[0073] 2. Secondly, the present invention uses a step-size control parameter to control the improved hyperbolic tangent function. This allows for better adjustment of the step-size of the improved hyperbolic tangent function based on the error value (i.e., the received signal after self-interference cancellation), ensuring both algorithm convergence and stability. Furthermore, in a further preferred embodiment, the present invention further optimizes the step-size, taking into account the influence of random variables and strengthening the relationship between the step-size and the input signal, thereby improving the tracking performance of the algorithm. Furthermore, a normalization algorithm is used to limit the step-size when updating the weight coefficients, minimizing the impact of sudden increases in the received signal on the algorithm.

[0074] 3. This invention aims to solve the problem that the step length parameters in the common variable step length self-interference cancellation algorithm need to be manually adjusted, which leads to the problem that the algorithm is unreliable or too complex in the self-interference cancellation. The step length control parameters are designed by using the exponential moving average method and the error variance method. 、 、 The adaptive update strategy reduces the complexity of manual participation and hyperparameter tuning, has stronger robustness and rapid adaptability, and enhances the algorithm's versatility and engineering feasibility.

[0075] The present invention provides strong theoretical support for the engineering implementation of self-interference cancellation and can be used to improve the self-interference cancellation capability of a simultaneous and co-frequency communication system.

[0076] Additional aspects and advantages of the present invention will be set forth in part in the description which follows and, in part, will be obvious from the description which follows, or may be learned by practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments with reference to the following drawings, in which:

[0078] Figure 1 :Block diagram of simultaneous and co-frequency self-interference cancellation system;

[0079] Figure 2 : Step size parameter in HT-VSS-LMS algorithm Impact on stride length;

[0080] Figure 3 : Step size parameter in HT-VSS-LMS algorithm Impact on stride length;

[0081] Figure 4 : Step size parameter in HT-VSS-LMS algorithm Impact on stride length;

[0082] Figure 5 : Step size parameter in HT-VSS-LMS algorithm Impact on stride length;

[0083] Figure 6 : Performance curve of simultaneous co-frequency self-interference cancellation based on HT-VSS-LMS. DETAILED DESCRIPTION

[0084] The following describes in detail embodiments of the present invention. The embodiments are exemplary and intended to explain the present invention, but are not to be construed as limiting the present invention.

[0085] Example 1:

[0086] In simultaneous and co-frequency communications, digital domain self-interference cancellation schemes based on the LMS algorithm have become relatively common, such as Figure 1 As shown, It is the transmission signal of the same frequency communication at the same time, and is also the input signal of the adaptive filter in the self-interference cancellation system, which is used for self-interference channel estimation. Output after processing , while the actual received signal of the same frequency communication This is the expected signal of the adaptive filter, which contains the communication signal to be received and self-interference signals, Output signal of the adaptive filter The actual received signal The error signal between them is used to adjust the adaptive filter, and is also the received signal after self-interference cancellation.

[0087] The principle of the simultaneous co-frequency self-interference cancellation scheme based on the LMS algorithm is: the adaptive filter weight coefficient and error After iteration, the adaptive filter outputs the reconstructed self-interference signal , and then reconstruct the self-interference signal From the actual receiving signal Subtracting the error signal That is, the received signal after self-interference cancellation.

[0088] The traditional LMS algorithm uses a fixed step size parameter to iteratively update the weight coefficients, which makes it difficult to effectively cope with the challenges brought by the time-varying characteristics of the self-interference channel. There are also schemes that use a variable step size design, but the variable step size adjustment method is difficult to simultaneously improve the two important performances of convergence speed and steady-state error. For this reason, this embodiment proposes a self-interference cancellation method based on the variable step size LMS algorithm of the hyperbolic tangent function. The hyperbolic tangent function is used to construct a nonlinear correlation function between the error signal and the iterative step size, and dynamic parameter adjustment is achieved: when the error is large in the initial stage of system convergence, a larger step size value is used to increase the convergence speed. When the error is small in the steady-state working range, a smaller step size value is used to reduce the steady-state error. The strategy of dynamically adjusting the step size in this embodiment overcomes the contradiction between the convergence speed and steady-state accuracy caused by the fixed step size of the traditional LMS algorithm. Simulation verification shows that the present invention uses different step size control parameters to have better control over the step size, and improves the performance of the algorithm when applied to self-interference cancellation.

[0089] The specific steps in this embodiment are:

[0090] Step 1: Obtain the transmitted signal for simultaneous and co-frequency communication and receive signals , because the present invention is based on the simultaneous same-frequency self-interference cancellation model in the digital domain, Indicates the points;

[0091] Step 2: According to the formula

[0092]

[0093] Calculating the self-interference signal ,in is the weight coefficient, and the superscript T indicates transposition;

[0094] Step 3: According to the formula

[0095]

[0096] Calculate the error signal This is the received signal after self-interference cancellation;

[0097] Step 4: Update the formula based on the weight coefficient

[0098]

[0099] The weight coefficients are updated iteratively until the iteration termination condition is reached; For the The weight coefficient of each point, is the step length, which is based on the improved hyperbolic tangent function formula

[0100]

[0101] Based on this, we introduce parameters 、 、 The change curve of the step length is jointly controlled. The step length value is adjusted according to the error value to better control the change of the step length value. The step length formula obtained is:

[0102]

[0103] The improved hyperbolic tangent function has the following characteristics: Symmetrical relative to a certain value, when the error signal When the value of is large (that is, the initial stage of the algorithm), the corresponding step size is also large, which can speed up the convergence of the algorithm. The value tends to (That is, the algorithm converges and stably approaches the actual communication signal to be received ), the corresponding step value The improved hyperbolic tangent function conforms to the step size selection principle of the optimized LMS algorithm and can overcome the defect that the convergence speed and steady-state error of the existing LMS algorithm cannot be improved at the same time. It has a higher convergence speed and a smaller steady-state error.

[0104] Example 2:

[0105] In this embodiment, in order to reduce the influence of random variables, the error signal in the step size formula is Use its related value Instead, the step size formula is adjusted to:

[0106]

[0107] in For the The error signal of each point.

[0108] Example 3:

[0109] In this embodiment, in order to further strengthen the relationship between the step length and the input signal and improve the tracking effect of the algorithm, a step length feedback factor is added to the step length formula. :

[0110]

[0111] in is the setting coefficient, For the The step length feedback factor of each point. In this embodiment, the initial value of the step length feedback factor is 1, and the coefficient is set Comprehensive consideration based on the step length curve.

[0112] Thus the step size formula is adjusted to:

[0113] .

[0114] The important difference between the HT-VSS-LMS algorithm and the LMS algorithm in this embodiment is the selection of the step size. 、 、 and Control the step size and adjust the step size according to the error value. and error signal relationship, such as Figures 2 to 5 shown.

[0115] Step size parameter Step length The value of has a great influence on The increase in step length The value of also increases accordingly; Figures 3 to 5 It can be seen that the step size parameter 、 、 Affects the shape of the step curve. The smaller, The bigger, The smaller the value, the smaller the corresponding step size under the same error signal value. Figures 2 to 5 Analysis of the impact of different parameters on the step size. In actual engineering applications, the step size is selected according to the state when the algorithm is close to convergence, so as to comprehensively consider the impact of various parameters on the step size and obtain better performance.

[0116] Example 4:

[0117] In this embodiment, in order to reduce the impact of a sudden increase in the received signal on the algorithm, a normalization algorithm is used to adjust the step size. Limit the step size to avoid large values, so that the weight coefficient update formula is adjusted to

[0118]

[0119] in To set a constant, the value is 1 in this embodiment.

[0120] Example 5:

[0121] This embodiment proposes a self-interference cancellation system based on a hyperbolic tangent function variable step size LMS algorithm, in which the adaptive filter weight coefficient is , The update formula is

[0122]

[0123] in To set a constant, in this embodiment, the value is 1; For simultaneous and co-frequency communication, The receiving signal for simultaneous and same-frequency communication The self-interference signal output by the adaptive filter The error signal between them, that is, the received signal after self-interference cancellation, is the step length, according to the formula

[0124]

[0125] Get, among them 、 、 are all control parameters that control the change of step size. is the step size feedback factor:

[0126]

[0127] in To set the coefficient, set the coefficient Comprehensive consideration based on the step length curve.

[0128] Example 6:

[0129] This embodiment proposes a step size control parameter adaptive update strategy, aiming to solve the problem that the step size parameter in the common variable step size self-interference cancellation algorithm needs to be manually adjusted, resulting in unreliable or excessively complex algorithms in the self-interference cancellation.

[0130] Design step size control parameters 、 、 Adaptive update strategy:

[0131] When the error is large when using the exponential moving average method Increase, speed up the convergence; when the error tends to be stable, Reduce and enhance steady-state performance:

[0132]

[0133] in, is the smoothing coefficient, ρ ∈ [ 0 , 1 ] , is the scaling factor;

[0134] Adaptively adjust the function curvature parameter using error variance :

[0135]

[0136] in, is the mapping ratio, , is the sliding window length, L ∈ [ 10 , 30 ] , is the average error within the sliding window, is the mapping ratio control parameter, Set to the minimum value to prevent , the value is ;

[0137] Optimize response speed to accelerate exponential contraction effect and improve dynamic adaptability. Set to:

[0138]

[0139] After introducing the above step size control parameter update adaptive update strategy, the step size formula expression can be obtained as follows:

[0140] .

[0141] By updating and iterating the adaptive filter weight coefficients, simultaneous and co-frequency self-interference cancellation is achieved. The following simulation analysis is performed:

[0142] The self-interference signal and communication signal in the simulation are generated by BPSK modulation, and the signal-to-interference ratio (SIR) is set to -50~20dB. Under different SIRs, the full-duplex self-interference cancellation performance of the algorithm is analyzed by comparing the residual self-interference signal power and the received self-interference signal power. For easy comparison, the communication signal power is compared with the residual self-interference signal power and the received self-interference signal power value. RES-SIR is defined to represent the ratio of the communication signal power to the residual self-interference signal power. The self-interference cancellation performance curve based on HT-VSS-LMS is shown below. Figure 6 shown.

[0143] according to Figure 6Analyzing the self-interference signal power from small to large, we find that when the SIR is greater than 0 dB, the residual self-interference signal's RES_SIR remains essentially unchanged as the SIR decreases, remaining near 0 dB. At this point, the received communication signal power is much greater than the received self-interference signal, and the algorithm's self-interference suppression effect is insignificant. When the SIR is between -40 and 0 dB, the difference between the residual self-interference signal's RES-SIR and SIR increases as the SIR decreases, indicating that the HT-VSS-LMS algorithm's self-interference cancellation performance improves with increasing self-interference signal strength. When the INR is less than -40 dB, the residual self-interference signal's RES-SIR and SIR exhibit a linear relationship. This is because the self-interference signal simulation process in the self-interference cancellation process converges, and this convergence is also reflected in the residual self-interference signal, resulting in a linear relationship between the two. The difference is approximately 33 dB, indicating that under the current parameter settings, the HT-VSS-LMS algorithm can achieve a self-interference cancellation effect of approximately 33 dB.

[0144] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative 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 purpose of the present invention.

Claims

1. A self-interference cancellation method based on a variable step-size LMS algorithm of a hyperbolic tangent function, characterized by: The following steps are involved: Step 1: Obtain the transmitted signal for simultaneous and co-frequency communication and receive signals , Indicates the points; Step 2: According to the formula Calculating the self-interference signal ,in is the weight coefficient; Step 3: According to the formula Calculate the error signal This is the received signal after self-interference cancellation; Step 4: Update the formula based on the weight coefficient The weight coefficients are updated iteratively until the iteration termination condition is reached; For the The weight coefficient of each point, is the step length, according to the formula Get, among them 、 、 are all control parameters for controlling step size changes; and the design step size control parameters 、 、 The adaptive update strategy is: Exponential moving average method is used to Perform adaptive updates: in, is the smoothing coefficient, , is the scaling factor; Using the error variance Perform adaptive adjustment: in, is the mapping ratio, , is the sliding window length, is the average error within the sliding window, is the mapping ratio control parameter, is the set minimum value; The adaptive adjustment formula is: 。 2. The method for self-interference cancellation based on the variable step size LMS algorithm of the hyperbolic tangent function according to claim 1, characterized in that: Adjust the step size formula to: in For the The error signal of each point.

3. The method for self-interference cancellation based on the variable step-size LMS algorithm of the hyperbolic tangent function according to claim 1, characterized in that: Adjust the step size formula to: in is the step size feedback factor.

4. The method for self-interference cancellation based on the variable step-size LMS algorithm of the hyperbolic tangent function according to claim 3, characterized in that: Adjust the step size formula to: 。 5. The self-interference cancellation method based on the hyperbolic tangent function variable step size LMS algorithm according to claim 3 or 4, characterized in that: The step size feedback factor is based on the formula Calculated, where is the setting coefficient, For the The step size feedback factor for each point.

6. The method for self-interference cancellation based on the variable step-size LMS algorithm of the hyperbolic tangent function according to claim 1, characterized in that: Adjust the weight coefficient update formula to in To set a constant.

Images