摘要
针对脉冲噪声下恒模算法(Constant Modulus Algorithm,CMA)失败的问题,通过分析脉冲噪声的影响,提出了一种基于最小均方(Least Mean Square,LMS)准则的对数型恒模算法(Logarithmic-type CMA,LT-CMA)。LT-CMA利用对数函数的非线性变换特性自适应地抑制强脉冲噪声对误差函数的影响,并且利用l2-范数进行信号归一化处理以增强算法的稳健性。仿真结果表明,所提出的LT-CMA可以适应于高斯噪声环境和脉冲噪声环境;与经典自适应均衡算法相比,在收敛速度和稳健性两方面上,所提出的LT-CMA都有显著的提升。
For the failure of constant modulus algorithm(CMA) for impulse noise,a logarithmic-type CMA(LT-CMA) is proposed based on least mean square(LMS) error criterion after analyzing influence of the impulse noise. It takes advantage of the characteristic of nonlinear transformation of the logarithmic function to adaptively suppress impact of the impulse noise on error function. Also,it uses the l2-norm to normalize the signals in order to enhance its robustness. Simulation results show that the proposed LT-CMA not only adapts to Gaussian noise environment but also impulse noise environment. Compared with some classical adaptive algorithms,the proposed LT-CMA can achieve a great improvement in convergence speed and robustness.
引文
[1] LI M J.Mixture filtering approaches to blind equalizationbased on estimation of time-varying and multi-path chan-nels[J]. IEEE Journal of Communications&Networks,2016,18(1):8-18.
[2] KHAN Q,VIQAR S,KH S.Two novel blind equalizationalgorithms for rectangular quadrature amplitude modula-tion constellations[J]. IEEE Access,2016(4):9512-9519.
[3]刘峰,葛临东,吴业进,等.一种基于实数编码遗传算法的常模盲均衡[J].电讯技术,2008,48(10):5-8.
[4] PRIYADARSHI P,RAI C.Blind channel equalization u-sing modified constant modulus algorithm[C]//Proceed-ings of IEEE International Conference on Computing,Communication and Automation(ICCCA). Noida:IEEE,2016:1020-1024.
[5] WU Z,KASSAM S,KOIVUNEN V. Relative-gradientbussgang-type blind equalization algorithms[C]//Pro-ceedings of IEEE International Conference on Acoustics,Speech and Signal Processing(ICASSP).Shanghai:IEEE,2016:2459-2463.
[6] BLOM K C,DOL H,KOKKELER A,et al.Blind equaliza-tion of underwater acoustic channels using implicit higher-order statistics[C]//Proceedings of IEEE UnderwaterCommunications and Networking Conference(UComms).Lerici:IEEE,2016:1-5.
[7] LUO J,WANG S,ZHANG W.Stochastic signal detectionbased on changeable fractional lower order moments in al-pha stable noise[C]//Proceedings of IEEE InternationalConference on Signal Processing,Communications andComputing(ICSPCC).Hong Kong:IEEE,2016:1-5.
[8] SHEN X,ZHANG H,XU Y,et al.Observation of alpha-stable noise in the laser gyroscope data[J].IEEE SensorsJournal,2016,16(7):1998-2003.
[9] MITRA R,BHATIA V. Normalised multi-stage clusteringequaliser for underwater acoustic channels[C]//Proceed-ings of IEEE Sensor Signal Processing for Defence(SSPD). Edinburgh:IEEE,2015:1-5.
[10] MITRA R,BHATIA V.Improved multi-stage clusteringbased blind equalization in distributed environments[C]//Proceedings of IEEE International Conference onInformation Technology(ICIT). Bhubaneswar:IEEE,2014:1-5.
[11] RUPI M,TSAKALIDES P,NIKIAS C,et al. Robust con-stant modulus arrays based on fractional lower-order sta-tistics[C]//Proceedings of IEEE Acoustics,Speech andSignal Processing(ICASSP).Phoenix:IEEE,1999:2945-2948.
[12] LI S,WANG Y,LIN B.Concurrent blind channel equali-zation in impulsive noise environments[J].Chinese Jour-nal of Electronic,2013,22(4):741-746.
[13]王大磊,杨宾,吴瑛,等.一种利用范数性质的无约束盲均衡准则[J].西安电子科技大学学报,2014,41(4):123-129.
[14]韩仿仿,冯宝强.一种改进的变动量项归一化p-范数短波信道盲均衡算法[J].重庆邮电大学学报,2015,27(6):717-721.
[15]张红梅,韩万刚.一种新的变步长LMS自适应滤波算法研究及其应用[J].仪器仪表学报,2015,36(8):1822-1830.