WO2012139357A1 - Procédé destiné à réaliser un filtre auto-adaptatif de sous-bande - Google Patents

Procédé destiné à réaliser un filtre auto-adaptatif de sous-bande Download PDF

Info

Publication number
WO2012139357A1
WO2012139357A1 PCT/CN2011/079024 CN2011079024W WO2012139357A1 WO 2012139357 A1 WO2012139357 A1 WO 2012139357A1 CN 2011079024 W CN2011079024 W CN 2011079024W WO 2012139357 A1 WO2012139357 A1 WO 2012139357A1
Authority
WO
WIPO (PCT)
Prior art keywords
stage
signal
adaptive
filter
decomposition
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2011/079024
Other languages
English (en)
Chinese (zh)
Inventor
谢宁
王晖
凌均跃
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenzhen University
Original Assignee
Shenzhen University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shenzhen University filed Critical Shenzhen University
Publication of WO2012139357A1 publication Critical patent/WO2012139357A1/fr
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H03—ELECTRONIC CIRCUITRY
    • H03H—IMPEDANCE NETWORKS, e.g. RESONANT CIRCUITS; RESONATORS
    • H03H21/00—Adaptive networks
    • H03H21/0012—Digital adaptive filters
    • H—ELECTRICITY
    • H03—ELECTRONIC CIRCUITRY
    • H03H—IMPEDANCE NETWORKS, e.g. RESONANT CIRCUITS; RESONATORS
    • H03H21/00—Adaptive networks
    • H03H21/0012—Digital adaptive filters
    • H03H21/0025—Particular filtering methods
    • H03H2021/0041—Subband decomposition

Definitions

  • the invention belongs to the field of signal processing, and in particular relates to a method for constructing a subband adaptive filter.
  • Adaptive filters can be applied in many fields, such as system identification, channel equalization, echo cancellation, beamforming, and the like.
  • the general method of constructing an adaptive filter is to first set the initial filter coefficients to arbitrary conditions, and then update the filter coefficients step by step according to the input signal and the desired signal to obtain Optimal filter coefficient setting.
  • the least mean square is often used to construct adaptive filters ( The Least mean square (LMS) algorithm updates the filter coefficients.
  • LMS Least mean square
  • the algorithm's further filter coefficients produce a problem of slow convergence. For colored inputs, such as speech signals, especially when the order of the adaptive filter to be constructed is very long, higher computational overhead is required.
  • a good way to improve the convergence speed and reduce the computational overhead is to construct a subband adaptive filter, which decomposes the input signal into multiple subband signals and performs adaptive filtering on each subband, so that the input colored signal can be whitened. , improve the convergence speed.
  • the structure of the existing constructed polyphase decomposition subband adaptive filter is shown in Fig. 1.
  • the input signals N 'times the subband decomposition the colored band by dividing the input signal, the input signal corresponding to a whitened, can reduce the correlation of the input signals, to improve the convergence of the adaptive algorithm Sex.
  • the input signal N 'extraction times the data rate can be reduced subband, thereby reducing the computational overhead of updating the adaptive filter coefficients.
  • the polyphase decomposition of the adaptive filter to be constructed can reduce the order of the adaptive filter and improve the convergence speed. For the finite impulse response, it can also completely reconstruct the finite impulse response of any order ( Finite Impulse Response, FIR) system.
  • FIR Finite Impulse Response
  • the subband adaptive filter of this structure only decomposes the subbands first, in order to obtain a faster convergence speed, the number of subbands needs to be increased, and the filter length in the filter bank is correspondingly Growth.
  • the computational overhead required for subband decomposition is greatly increased, which necessitates a compromise between computational complexity and convergence speed.
  • the purpose of embodiments of the present invention is to solve the existing sub-bands for constructing polyphase decomposition.
  • the adaptive filter needs to make a compromise between computational complexity and convergence speed, and provides a method for constructing a subband adaptive filter, which can further improve the convergence speed and reduce the computational complexity.
  • a method for constructing a subband adaptive filter includes the following steps:
  • the filter adjustment coefficient that is adaptively filtered at the next time is updated according to the filter adjustment coefficient that is currently adaptively filtering the input signal, and the difference between the obtained adaptive filtered output signal and the two-stage decomposed desired signal.
  • the filter adjustment coefficient for adaptive filtering is updated at the next moment, thereby realizing Species
  • the method of constructing a subband adaptive filter can further improve the convergence speed and reduce the computational complexity.
  • FIG. 1 is a schematic structural diagram of a multi-phase decomposed sub-band adaptive filter provided by the prior art
  • FIG. 2 is a flowchart of an implementation of a method for constructing a subband adaptive filter according to an embodiment of the present invention
  • FIG. 3 is a schematic structural diagram of performing two-stage subband decomposition and adaptive filtering on an input signal according to an embodiment of the present invention
  • FIG. 4 is a structural diagram of two-stage subband decomposition of a desired signal according to an embodiment of the present invention.
  • FIG. 5 is a block diagram of the entire model of system identification provided by an embodiment of the present invention.
  • the filter adjustment coefficient that is currently adaptively filtered on the input signal, and the difference between the obtained adaptive filtered output signal and the two-stage decomposed desired signal are updated to the next time for adaptive filtering.
  • Filter adjustment factor a kind of A method of constructing a subband adaptive filter.
  • FIG. 2 is a flowchart showing an implementation process of constructing a subband adaptive filter method according to an embodiment of the present invention, which is described in detail as follows:
  • step S201 the input signal is subjected to two-stage decomposition and adaptive filtering
  • step S202 the desired signal is subjected to two-stage sub-band decomposition
  • step S203 The filter adjustment coefficient that is adaptively filtered at the next time is updated according to the filter adjustment coefficient that is currently adaptively filtering the input signal, and the difference between the obtained adaptive filtered output signal and the two-stage decomposed desired signal.
  • the input signal is subjected to two-stage sub-band decomposition and adaptive filtering using the structure shown in FIG. Specifically, the steps S201 includes:
  • Step S2011 performing N-times first-order sub-band decomposition and N-times extraction on the input signal
  • Step S2012 respectively performing the second-level sub-band decomposition of the obtained first-level sub-band signal by M times;
  • Step S2013 performing adaptive filtering and M-time extraction on each second-level sub-band signal to obtain a second-stage output signal.
  • Step S202 is:
  • step S203 a filter adjustment coefficient for adaptively filtering each of the second-level sub-band signals, and a step The difference between each second-stage output signal obtained in S2013 and the corresponding two-stage decomposed second-order expected signal obtained in step S202 is updated with the filter adjustment coefficient that is adaptively filtered at the next time.
  • step S2011 the input signal is first Pass through the first stage N-channel filter bank , , ..., Filtering, and performing a time unit z -1 delay; then, respectively, N-times extraction of the obtained first-level sub-band signal.
  • step S2012 the first-level sub-band signals after the N-time extraction are respectively passed through the second-stage M-channel filter bank. , , ..., Filtering, can get N ⁇ N ⁇ M second-level sub-band signals , , « , , , .... , ; ; accompanied that is, the input signal loss of N ⁇ N ⁇ M adaptive filters.
  • each second-level sub-band signal is respectively passed through an adaptive filter bank.
  • step S202 the desired signal is applied to the structure shown in FIG. Perform two-stage subband decomposition: first expect the signal Pass through the first stage N-channel filter bank , , ..., Filtering, then N is extracted; then, the obtained first-stage desired signal is passed through the second-stage M-channel filter bank , , ..., Filtering, and then M is extracted to obtain the second-order expected signal after two-stage decomposition , , ..., .
  • the filter adjustment coefficient of the adaptive filter bank adaptively filtered according to the current time, that is, the time k , , ..., And the error signal obtained , , ..., Update the filtering adjustment coefficient of the adaptive filter bank for adaptive filtering at the next moment, ie, k+1 time , , ..., .
  • n represents the data rate of the input signal
  • k represents the data rate of the input signal and the output signal corresponding to each sub-band
  • k n/N.
  • the N-phase decomposition is performed as follows:
  • the second-stage output signal y i (k) obtained by adaptively filtering each second-level sub-band signal is:
  • the superscript T indicates that the adaptive filter adjustment coefficient matrix wj(k) is transposed.
  • Subband error signal for:
  • an update algorithm of the filter adjustment coefficient of the adaptive filter is calculated based on the principle of minimizing interference.
  • the principle of minimizing interference is to ensure the total amount of change in the adaptive filter adjustment coefficient when the desired signal constraint is met at two iterations. f is the smallest.
  • the amount of change f of the total adaptive filter adjustment coefficient is defined as:
  • the principle of minimizing interference can be expressed as: (5) When the formula is satisfied, the equation (4) is minimized.
  • the cost function can be constructed using the Lagrangian multiplier method. ,which is:
  • the adaptive filter bank does not have a large aliasing, the cross-correlation of its output signal is much smaller than its autocorrelation, so when i ⁇ l Time,
  • the formula (12) is the step S203 The calculation formula of the filter adjustment coefficient for adaptive filtering according to the current filter adjustment coefficient for adaptively filtering the input signal and the difference between the obtained adaptive filtered output signal and the expected signal of the two-stage decomposition .
  • the block diagram of the model for implementing system identification is shown in Figure 5.
  • the input signal is , will input signal Through unknown systems After getting the desired signal , For observation The effect of noise received.
  • the desired sub-band adaptive filter method is provided by the embodiment of the present invention to the desired signal Perform two-stage subband decomposition and input signal Subband adaptive filtering is performed.
  • the polyphase decomposition is as follows, the unknown system It is also decomposed into N parts s j . If the length of the unknown system is set to L, the length of the polyphase decomposition component s j is L/N.
  • the error vector for the filter adjustment coefficient that defines the sub-band adaptive filtering at time k is:
  • the mean mean square deviation (MSD) c(k) is ,
  • E Indicates the expectation, here is the sum of the squares of the inner products of the error vectors of all subband adaptive filter filter adjustment coefficients, used to represent the statistical value of the magnitude of the entire coefficient error vector.
  • the coefficient update equation (12) is substituted into (14) and then substituted into (15) to obtain the step size parameter.
  • the range of values is:
  • the coefficient vector of the full band can also be obtained.
  • N The multiphase component can in turn find the full band coefficient vector.
  • the complexity of the adaptive filtering structure using the two-stage decomposition can be decomposed, considering the number of multiplications of each input sample, including subband decomposition of the input signal and the desired signal, filtering of the input signal, and the second level of each
  • the integration with the output signal, as well as the filter adjustment factor of the adaptive filter is calculated in five parts.
  • the length of the filter is a multiple of the number of subbands. If the number of subbands is m, the length of the filter is . In this way, the computational complexity of the entire structure can be obtained as:
  • L is the length of the full-band adaptive filter; (4, 1) represents the use of level one 4
  • the filter adjustment coefficient for adaptive filtering is updated at the next moment, thereby realizing Species
  • the method of constructing a subband adaptive filter can further improve the convergence speed and reduce the computational complexity.
  • the filter adjustment coefficient that is adaptively filtered at the next time is updated according to the filter adjustment coefficient that is currently adaptively filtering the input signal, and the difference between the obtained adaptive filtered output signal and the two-stage decomposed desired signal.

Landscapes

  • Filters That Use Time-Delay Elements (AREA)

Abstract

L'invention se rapporte au domaine de l'élimination d'un signal et à un procédé destiné à réaliser un filtre auto-adaptatif de sous-bande. Le procédé comprend les étapes consistant à : exécuter une décomposition à deux étapes et un filtrage auto-adaptatif sur le signal d'entrée; exécuter une décomposition de sous-bande à deux étapes sur le signal souhaité; et mettre à jour le coefficient de réglage du filtre soumis au filtrage auto-adaptatif la fois suivante selon les coefficients de réglage du filtre actuels soumis au filtrage auto-adaptatif exécuté sur le signal d'entrée et la différence obtenue entre le signal de sortie soumis au filtrage auto-adaptatif et le signal souhaité soumis à la décomposition à deux étapes. Dans le mode de réalisation de l'invention, un procédé destiné à réaliser un filtre auto-adaptatif de sous-bande est réalisé en mettant à jour le coefficient de réglage du filtre soumis au filtrage auto-adaptatif la fois suivante selon les coefficients de réglage du filtre actuels soumis au filtrage auto-adaptatif exécuté sur le signal d'entrée et la différence obtenue entre le signal de sortie soumis au filtrage auto-adaptatif et le signal souhaité soumis à la décomposition à deux étapes, ce qui permet d'améliorer le taux de convergence et de réduire la complexité de calcul.
PCT/CN2011/079024 2011-04-15 2011-08-28 Procédé destiné à réaliser un filtre auto-adaptatif de sous-bande Ceased WO2012139357A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201110096158XA CN102281045A (zh) 2011-04-15 2011-04-15 一种构建子带自适应滤波器方法
CN201110096158.X 2011-04-15

Publications (1)

Publication Number Publication Date
WO2012139357A1 true WO2012139357A1 (fr) 2012-10-18

Family

ID=45106266

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2011/079024 Ceased WO2012139357A1 (fr) 2011-04-15 2011-08-28 Procédé destiné à réaliser un filtre auto-adaptatif de sous-bande

Country Status (2)

Country Link
CN (1) CN102281045A (fr)
WO (1) WO2012139357A1 (fr)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115842532A (zh) * 2022-12-21 2023-03-24 哲库科技(上海)有限公司 信号滤波方法、装置以及电子设备
CN121077542A (zh) * 2025-10-30 2025-12-05 北京理工大学 多波束信号动态截位方法、装置、电子设备及存储介质

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104427144B (zh) * 2013-09-11 2017-06-13 联芯科技有限公司 一种线性回声消除方法及其装置
CN103762958B (zh) * 2014-01-07 2016-09-28 南京信息工程大学 一种改进的仿射组合自适应滤波方法
CN103929150B (zh) * 2014-03-27 2017-02-01 苏州大学 一种子带自适应滤波器的权值向量更新方法
CN108574459B (zh) * 2017-03-14 2022-04-01 南京理工大学 一种高效时域宽带波束形成电路及方法
CN111211759B (zh) * 2019-12-31 2022-03-25 京信网络系统股份有限公司 滤波器系数确定方法、装置和数字das系统
CN112803921B (zh) * 2021-04-13 2021-09-07 浙江华创视讯科技有限公司 自适应滤波器、方法、介质及电子设备
CN119995346B (zh) * 2025-02-18 2026-04-10 电子科技大学 一种具有高响应速度特性的开关直流变换器控制方法
CN120185583B (zh) * 2025-05-22 2025-08-19 成都必控科技有限责任公司 基于多频段融合的自适应信号滤波方法及系统

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0903899A2 (fr) * 1997-09-12 1999-03-24 Siemens Aktiengesellschaft Procédé et dispositif d'évaluation des signaux de tonalités multifréquences utilisant un filtre d'ondes numérique adaptatif à réjection
CN1790902A (zh) * 2004-12-13 2006-06-21 上海无线通信研究中心 一种自适应滤波方法及装置

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0903899A2 (fr) * 1997-09-12 1999-03-24 Siemens Aktiengesellschaft Procédé et dispositif d'évaluation des signaux de tonalités multifréquences utilisant un filtre d'ondes numérique adaptatif à réjection
CN1790902A (zh) * 2004-12-13 2006-06-21 上海无线通信研究中心 一种自适应滤波方法及装置

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
ZHANG, YINHANG: "FPGA Realization of Self-adaptive Filter Based on Subband Decomposition", SOUTHWEST JIAOTONG UNIVERSITY MASTER DEGREE THESIS, 18 May 2010 (2010-05-18) *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115842532A (zh) * 2022-12-21 2023-03-24 哲库科技(上海)有限公司 信号滤波方法、装置以及电子设备
CN121077542A (zh) * 2025-10-30 2025-12-05 北京理工大学 多波束信号动态截位方法、装置、电子设备及存储介质

Also Published As

Publication number Publication date
CN102281045A (zh) 2011-12-14

Similar Documents

Publication Publication Date Title
WO2012139357A1 (fr) Procédé destiné à réaliser un filtre auto-adaptatif de sous-bande
Arce A general weighted median filter structure admitting negative weights
Yoo et al. Adaptive dynamic surface control for stabilization of parametric strict-feedback nonlinear systems with unknown time delays
Leahy et al. Adaptive filtering of stable processes for active attenuation of impulsive noise
Aslam et al. A new adaptive strategy to improve online secondary path modeling in active noise control systems using fractional signal processing approach
Ngwompo et al. Inversion of linear time-invariant SISO systems modelled by bond graph
CN112017679A (zh) 用于自适应滤波器系数更新的方法及装置、设备
CN113078884B (zh) 添加非线性拟合的自适应算法
CN102811267A (zh) 近端语音干扰消除系统及移动通信终端
CN113347536B (zh) 基于线性预测与子带自适应滤波的声反馈抑制算法
CN110491405B (zh) 基于协同非线性自适应滤波的麦克风阵列语音增强方法
Bose et al. The Euclidean direction search algorithm in adaptive filtering
Ni et al. Efficient implementation of the affine projection sign algorithm
CN109658947A (zh) 一种同步建模和控制的主动噪声控制方法
CN111755020A (zh) 一种立体声回声消除方法
Mikhael et al. A fast frequency-domain adaptive algorithm
CN119232119A (zh) 基于自适应阈值的抗噪声方法、系统、介质、设备及程序
Bouchard et al. The multichannel gauss-seidel fast affine projection algorithm for active noise control
CN113972912A (zh) 一种直流捕获及直流消除电路、高通滤波器及芯片
Sulyman et al. Echo cancellation using a variable step-size NLMS algorithm
WO2022031061A1 (fr) Appareil d'élimination de réverbération à base de wpe faisant appel à une extension de canal virtuel basée sur un réseau neuronal profond
Treichler γ-LMS and its use in a noise-compensating adaptive spectral analysis technique
Goodwin et al. Uncertainty, information and estimation
WO2025254497A1 (fr) Système et procédé d'annulation de rétroaction acoustique adaptative
Wang et al. Adaptive combination of filtered-X NLMS and affine projection algorithms for active noise control

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 11863479

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 11863479

Country of ref document: EP

Kind code of ref document: A1