改进子带自适应滤波算法及其在回波抵消中的应用
|
摘要
本文提出了两种基于多带结构的仿射投影符号子带自适应滤波器(Affine Projection Sign Subband Adaptive Filter,AP-SSAF)的改进方法。针对稀疏系统的系统识别,设计了两种子带自适应滤波器。首先给出了AP-SSAF的变正则化参数更新方程,文中采用随机梯度下降法来更新正则化参数,来使系统的均方偏差最小化,该方法能同时兼顾快速收敛及低稳态失调。其次将权重分布矩阵引入AP-SSAF得到系数比例AP-SSAF,该方法能够利用系统的稀疏性提高AP-SSAF的收敛性能。仿真中将本文所提算法用于一般系统识别以及回波抵消,实验结果验证了本文的算法对脉冲噪声具有稳健性,具有较好的跟踪性能,并具有较快的收敛速度及低稳态失调。
This paper proposes two improved affine projection sign subband adaptive filters( AP-SSAF) based on the multiband structure,which called variable regularization parameter AP-SSAF( VRP-AP-SSAF) and proportionate AP-SSAF( PAP-SSAF). For sparse system identification,two AP-SSAFs are designed. On the one hand,VRP-AP-SSAF is presented to get a trade-off between fast convergence rate and low steady-state misalignment by minimizing the mean-square deviation( MSD) of the system weight vector. In addition,the update of regularization parameter is achieved by the normalized stochastic gradient of MSD. On the other hand,P-AP-SSAF incorporates a gain distribution matrix into the AP-SSAF to proportionately adapt the tap-weight vector of adaptive filter,which can both utilize the sparsity of system and raise convergence performance for sparse system identification. Simulation results show that the proposed VRP-AP-SSAF and P-APSSAF used in general system identification and echo cancellation,not only maintain the robustness against impulsive noise and have a good result for tracking capacity,but also have an improved performance in convergence rate and steady-state misalignment in their proper context.
This paper proposes two improved affine projection sign subband adaptive filters( AP-SSAF) based on the multiband structure,which called variable regularization parameter AP-SSAF( VRP-AP-SSAF) and proportionate AP-SSAF( PAP-SSAF). For sparse system identification,two AP-SSAFs are designed. On the one hand,VRP-AP-SSAF is presented to get a trade-off between fast convergence rate and low steady-state misalignment by minimizing the mean-square deviation( MSD) of the system weight vector. In addition,the update of regularization parameter is achieved by the normalized stochastic gradient of MSD. On the other hand,P-AP-SSAF incorporates a gain distribution matrix into the AP-SSAF to proportionately adapt the tap-weight vector of adaptive filter,which can both utilize the sparsity of system and raise convergence performance for sparse system identification. Simulation results show that the proposed VRP-AP-SSAF and P-APSSAF used in general system identification and echo cancellation,not only maintain the robustness against impulsive noise and have a good result for tracking capacity,but also have an improved performance in convergence rate and steady-state misalignment in their proper context.
引文
[1]Haykin S.Adaptive Filter Theory[M].Upper Saddle River,NJ:Prentice-Hall,2002.
[2]Lee K A,Gan W S,Kuo S M.Subband Aadaptive Filtering:Theory and Implementation[M].Chichester:John Wiley&Sons,UK,2009.
[3]Shao T,Zheng Y R,Benesty J.An affine projection sign algorithm robust against impulsive interferences[J].IEEE Transactions Signal Processing Letters,2010,17(4):327-330.
[4]Ni J,Li F.Variable regularization parameter sign subband adaptive filter[J].Electronics Letter,2010,46(24):1605-1607.
[5]Shin J,Yoo J,Park P.Variable step-size affine projection sign algorithm[J].Electronics Letter,2012,48(9):483-485.
[6]Yoo J W,Shin J W,Park P G.Variable step-size affine projection sign algorithm[J].IEEE Transactions on Circuits and Systems-II Aanalog and Digital Signal Processing,2014,61(4):274-278.
[7]Shin J,Yoo J,Park P.Variable step-size sign subband adaptive filter[J].IEEE Signal Processing Letter,2013,20(2):173-176.
[8]Kim J H,Chang J H,Nam S W.Sign subband adaptive filter with1-norm minimisation-based variable step-size[J].Electronics Letter,2013,49(21):1325-1326.
[9]Chng M,Kong N,Park P.Variable regularization leastsquare algorithm:one-step-ahead cost function with equivalent optimality[J].Signal Processing,2011,91(5):1224-1228.
[10]Se Choi Y,Shin H C,Song W J.Adaptive regularization matrix for affine projection algorithm[J].IEEE Transactions on Circuits and Systems-II:Express Briefs,2007,54(12):1087-1091.
[11]Yin W,Meh A S.A variable regularization method for affine projection algorithm[J].IEEE Transactions on Circuits and Systems-II:Express Briefs,2010,57(6):476-480.
[12]Ni J,Li F.A variable regularization matrix normalized subband adaptive filter[J].IEEE Signal Processing Letters,2009,16(2):105-108.
[13]Ni J,Li F.Normalised subband adaptive filter with new variable regularisation matrix[J].Electronicsics Letters,2010,46(18):1293-1295.
[14]Jeong J J,Koo K,Koo G,et al.Variable regularization for normalized subband adaptive filter[J].Signal Processing,2014,104:432-436.
[15]Paleologu C,Benesty J,CiochinS.Regularization of the affine projection algorithm[J].IEEE Transactions on Circuits and Systems-II:Express Briefs,2011,58(6):366-370.
[16]Nin J,Chen X,Yang J.Two variants of the sign subband adaptive filter with improved convergence rate[J].Signal Processing,2014,96:325-331.
[17]Duttweiler D L.Proportionate normalized least-mean-squares adaptation in echo cancellers[J].IEEE Transactions on Audio,Speech,and Language Processing,2000,8(5):508-518.
[18]Benesty J,Gay S L.An improved PNLMS algorithm[C]∥Proceedings of the ICASSP-02,USA,2002:1881-1884.
[19]See X Y,Lee K A,Gan W S.Proportionate subband adaptive filtering[C]∥Proceedings of ICALIP,2008:128-132.
[20]Yang Z,Zheng Y R,Grant S L.Proportionate affine projection sign algorithms for network echo cancellation[J].IEEE Transactions on Audio,Speech,and Language Processing,2011,19(8):2273-2284.
[2]Lee K A,Gan W S,Kuo S M.Subband Aadaptive Filtering:Theory and Implementation[M].Chichester:John Wiley&Sons,UK,2009.
[3]Shao T,Zheng Y R,Benesty J.An affine projection sign algorithm robust against impulsive interferences[J].IEEE Transactions Signal Processing Letters,2010,17(4):327-330.
[4]Ni J,Li F.Variable regularization parameter sign subband adaptive filter[J].Electronics Letter,2010,46(24):1605-1607.
[5]Shin J,Yoo J,Park P.Variable step-size affine projection sign algorithm[J].Electronics Letter,2012,48(9):483-485.
[6]Yoo J W,Shin J W,Park P G.Variable step-size affine projection sign algorithm[J].IEEE Transactions on Circuits and Systems-II Aanalog and Digital Signal Processing,2014,61(4):274-278.
[7]Shin J,Yoo J,Park P.Variable step-size sign subband adaptive filter[J].IEEE Signal Processing Letter,2013,20(2):173-176.
[8]Kim J H,Chang J H,Nam S W.Sign subband adaptive filter with1-norm minimisation-based variable step-size[J].Electronics Letter,2013,49(21):1325-1326.
[9]Chng M,Kong N,Park P.Variable regularization leastsquare algorithm:one-step-ahead cost function with equivalent optimality[J].Signal Processing,2011,91(5):1224-1228.
[10]Se Choi Y,Shin H C,Song W J.Adaptive regularization matrix for affine projection algorithm[J].IEEE Transactions on Circuits and Systems-II:Express Briefs,2007,54(12):1087-1091.
[11]Yin W,Meh A S.A variable regularization method for affine projection algorithm[J].IEEE Transactions on Circuits and Systems-II:Express Briefs,2010,57(6):476-480.
[12]Ni J,Li F.A variable regularization matrix normalized subband adaptive filter[J].IEEE Signal Processing Letters,2009,16(2):105-108.
[13]Ni J,Li F.Normalised subband adaptive filter with new variable regularisation matrix[J].Electronicsics Letters,2010,46(18):1293-1295.
[14]Jeong J J,Koo K,Koo G,et al.Variable regularization for normalized subband adaptive filter[J].Signal Processing,2014,104:432-436.
[15]Paleologu C,Benesty J,CiochinS.Regularization of the affine projection algorithm[J].IEEE Transactions on Circuits and Systems-II:Express Briefs,2011,58(6):366-370.
[16]Nin J,Chen X,Yang J.Two variants of the sign subband adaptive filter with improved convergence rate[J].Signal Processing,2014,96:325-331.
[17]Duttweiler D L.Proportionate normalized least-mean-squares adaptation in echo cancellers[J].IEEE Transactions on Audio,Speech,and Language Processing,2000,8(5):508-518.
[18]Benesty J,Gay S L.An improved PNLMS algorithm[C]∥Proceedings of the ICASSP-02,USA,2002:1881-1884.
[19]See X Y,Lee K A,Gan W S.Proportionate subband adaptive filtering[C]∥Proceedings of ICALIP,2008:128-132.
[20]Yang Z,Zheng Y R,Grant S L.Proportionate affine projection sign algorithms for network echo cancellation[J].IEEE Transactions on Audio,Speech,and Language Processing,2011,19(8):2273-2284.