CN-120804504-B - Adaptive filtering noise reduction system based on variable step minimum mean square error algorithm

CN120804504BCN 120804504 BCN120804504 BCN 120804504BCN-120804504-B

Abstract

The invention discloses a self-adaptive filtering noise reduction system based on a variable step-size minimum mean square error algorithm, which comprises a signal main link and a closed-loop control module, wherein the signal main link is sequentially integrated with a low-noise amplifier, an analog-to-digital converter and a self-adaptive filter and is responsible for accurate amplification, high-efficiency analog-to-digital conversion and flexible filtering processing of signals, and the closed-loop control system captures and analyzes noise spectrum characteristics in real time through close coordination of a noise detection unit, a spectrum analysis unit and a parameter adjustment unit so as to dynamically optimize filtering parameters to realize the optimal noise reduction effect. The parameter adjusting unit adopts a variable step-size minimum mean square error algorithm, and combines the nonlinear characteristic of the hyperbolic tangent function by means of a unique error signal real-time amplitude feedback mechanism, thereby skillfully realizing the dynamic adjustment of the step size and further achieving ideal self-adaptive balance between rapid convergence and low steady-state error.

Inventors

LI CHAOYI
LUO SHAN
LIAO YONG
ZHANG BOWEN
ZHAO LI
LIU YONG
LIU XIAO
FAN ZHIYONG
XIE LEI
LI LICHUN
WANG SHIJUN
LIU SHUANG

Assignees

华能西藏雅鲁藏布江水电开发投资有限公司
成都希盟泰克科技发展有限公司

Dates

Publication Date: 20260505
Application Date: 20250526

Claims (10)

1. The adaptive filtering noise reduction system based on the variable step minimum mean square error algorithm is characterized by comprising a signal main link and a closed loop control module; the signal main link comprises a low noise amplifier, an analog-to-digital converter and an adaptive filter which are sequentially connected, wherein the low noise amplifier is used for amplifying an input signal to obtain an amplified signal; the closed-loop control module comprises a noise detection unit, a spectrum analysis unit and a parameter adjustment unit; the input end of the noise detection unit is connected with the output end of the analog-to-digital converter and is used for extracting a noise component from the digital signal and calculating the root mean square value of the noise component; the input end of the frequency spectrum analysis unit is connected with the output end of the noise detection unit, and is used for analyzing the noise power spectrum density by adopting the fast Fourier transform according to the root mean square value of the noise component, and adjusting the smoothing parameter and the scaling parameter of the variable step minimum mean square error algorithm according to the noise power spectrum density; The input end of the parameter adjusting unit is connected with the output end of the spectrum analyzing unit, the output end of the parameter adjusting unit is connected with the adaptive filter, and the parameter adjusting unit is used for dynamically adjusting the tap coefficient of the adaptive filter by adopting a variable step minimum mean square error algorithm to form closed loop control on a signal main link.
2. The adaptive filtering noise reduction system based on the variable step minimum mean square error algorithm according to claim 1, wherein the formula of extracting the noise component from the digital signal by the noise detection unit is: where n is an index representing a discrete time sequence, The noise component is represented by a representation of the noise component, Representing the ideal reference signal(s), Representing an output vector generated by inner-integrating the input signal vector with tap coefficients of the adaptive filter.
3. The adaptive filtering noise reduction system based on the variable step minimum mean square error algorithm according to claim 1, wherein the formula for calculating the root mean square value of the noise component in the noise detection unit is: Wherein the method comprises the steps of Represents the root mean square value of the noise component, N represents the number of sampling points for signal analysis, Representing the noise component.
4. The adaptive filtering noise reduction system based on a variable step minimum mean square error algorithm according to claim 1, wherein the spectral analysis unit is responsive to a root mean square value of a noise component The noise power spectral density is resolved using a fast fourier transform.
5. The adaptive filtering noise reduction system based on a variable step minimum mean square error algorithm according to claim 4, wherein the formula for resolving the noise power spectral density is: Wherein the method comprises the steps of Representing the noise power density at the frequency k, Representing a single time domain sample point, N is an index representing a discrete time sequence, N represents the number of sample points for signal analysis, and j represents the imaginary unit.
6. The adaptive filtering noise reduction system based on the variable-step minimum mean square error algorithm according to claim 1, wherein the method for adjusting the smoothing parameter and the scaling parameter of the variable-step minimum mean square error algorithm according to the noise power spectral density in the spectrum analysis unit specifically comprises: If the variance sigma 2 of the noise power spectral density on each frequency point is less than or equal to 0.1, judging the noise as flat noise, and changing the smoothing parameter of the step-size minimum mean square error algorithm Increasing the initial value from 0.5 to 0.8 to accelerate convergence, if the variance sigma 2>0.1 of the noise power spectral density at each frequency point, judging the noise as uneven noise, and changing the scaling parameter of the step-size minimum mean square error algorithm From an initial value of 0.05 to 0.02 to improve steady state accuracy.
7. The adaptive filtering noise reduction system based on the variable step minimum mean square error algorithm according to claim 1, wherein the variable step minimum mean square error algorithm in the parameter adjustment unit comprises the steps of: s1, initializing parameters of a variable step minimum mean square error algorithm; s2, according to the noise component Calculating filter step size parameters ; S3, according to the step length parameter of the filter And dynamically adjusting the tap coefficient of the adaptive filter.
8. The adaptive filtering noise reduction system based on the variable step minimum mean square error algorithm of claim 7, wherein the step S1 is specifically that a maximum value mu max =0.1 of a filter step parameter is initialized, a minimum value mu min =0.001 of the filter step parameter is initialized, and a smoothing parameter is set Initial value of 0.5, scaling parameter The initial value of (2) is 0.05 and the initial value of the tap coefficient vector of the adaptive filter is 0.
9. The adaptive filtering noise reduction system based on the variable step minimum mean square error algorithm according to claim 7, wherein the step S2 is a filter step parameter The calculation formula of (2) is as follows: Wherein the method comprises the steps of The smoothing parameter is represented by a value of the smoothing parameter, The scaling parameter is indicated as such, Representing a hyperbolic tangent function.
10. The adaptive filtering noise reduction system based on the variable step minimum mean square error algorithm as claimed in claim 7, wherein the formula for dynamically adjusting the tap coefficient of the adaptive filter in step S3 is: Wherein the method comprises the steps of Representing the filter tap coefficient vector before adjustment, Representing the adjusted filter tap coefficient vector, Representing the input signal vector.

Description

Adaptive filtering noise reduction system based on variable step minimum mean square error algorithm Technical Field The invention belongs to the technical field of communication, and particularly relates to a design of a self-adaptive filtering noise reduction system based on a variable step minimum mean square error algorithm. Background In a wireless communication system, the quality of the receiver front-end signal processing directly affects the reliability and sensitivity of the communication link. With the rapid development of 5G, internet of things and high-frequency communication technologies, a receiver needs to process multi-frequency band and time-varying noise interference in a complex electromagnetic environment, including white noise, narrowband interference, impulse noise and the like. The traditional fixed parameter filter is difficult to effectively inhibit non-stationary noise due to lack of dynamic adaptability, so that signal distortion, increase of error rate and reduction of overall system performance are caused. Currently, the adaptive filtering technology is widely used for dynamic noise suppression, but has the core defect that the traditional Least Mean Square (LMS) algorithm adopts fixed step parameters, and is difficult to simultaneously meet the balance requirement of convergence speed and steady-state error. In addition, the existing adaptive filtering scheme is mostly dependent on a single noise characteristic adjustment parameter, and has insufficient adaptability to scenes with complex spectrum distribution, so that the noise suppression effect is limited. In recent years, in order to solve the above problems, researchers have proposed an improved algorithm based on LMS, but there still exist problems such as low dynamic adjustment efficiency, imperfect closed-loop control, and high complexity of hardware implementation. Therefore, there is a need for an adaptive filtering noise reduction circuit that combines convergence speed, steady-state accuracy and hardware efficiency to meet the stringent requirements of modern wireless communication systems for high sensitivity, low bit error rate and complex noise suppression capability. Disclosure of Invention The invention aims to solve the problem of low noise reduction performance of the existing self-adaptive filtering noise reduction technology based on an LMS algorithm, and provides a self-adaptive filtering noise reduction system based on a variable step minimum mean square error algorithm. The technical scheme of the invention is that the adaptive filtering noise reduction system based on the variable step minimum mean square error algorithm comprises a signal main link and a closed loop control module. The signal main circuit comprises a low-noise amplifier, an analog-to-digital converter and an adaptive filter which are sequentially connected, wherein the low-noise amplifier is used for amplifying an input signal to obtain an amplified signal, the analog-to-digital converter is used for performing analog-to-digital conversion on the amplified signal to obtain a digital signal, and the adaptive filter is used for filtering the digital signal to obtain an output signal. The closed loop control module comprises a noise detection unit, a frequency spectrum analysis unit and a parameter adjustment unit, wherein the input end of the noise detection unit is connected with the output end of the analog-to-digital converter and used for extracting noise components from digital signals and calculating root mean square values of the noise components, the input end of the frequency spectrum analysis unit is connected with the output end of the noise detection unit and used for analyzing noise power spectral density by adopting fast Fourier transformation according to the root mean square values of the noise components and adjusting smooth parameters and scaling parameters of a variable step minimum mean square error algorithm according to the noise power spectral density, and the input end of the parameter adjustment unit is connected with the output end of the frequency spectrum analysis unit and connected with the adaptive filter and used for dynamically adjusting tap coefficients of the adaptive filter by adopting the variable step minimum mean square error algorithm to form closed loop control of a signal main link. Further, the formula of extracting the noise component from the digital signal by the noise detection unit is: where n is an index representing a discrete time sequence, The noise component is represented by a representation of the noise component,Representing the ideal reference signal(s),Representing an output vector generated by inner-integrating the input signal vector with tap coefficients of the adaptive filter. Further, the formula for calculating the root mean square value of the noise component in the noise detection unit is: Wherein the method comprises the steps of Represents the root mean square value of t