A Generalization of the Normalized LMS Algorithm

As previously shown, the normalized least mean square (NLMS)algorithm has superior convergence properties than the least meansquare(LMS) algorithm. However, the weight noise effect of the NLMS algorithm is large so that the steady state residue power islarger than that for the LMS algorithm. A generalized NLMSalgorithm is developed based upon the pseudoinverse of anestimated covariance matrix. A preliminary evaluation indicatesimproved performance can be attained but the implementationcomplexity might be high.     

一种变步长LMS自适应滤波算法研究

为了解决最小均方(LMS)算法的稳定性以及收敛速度和稳态误差之间的矛盾,提出了一种复数型自适应变步长的LMS算法,它的权系数的调整取决于误差曲面在新权值点上的梯度。分析了新算法的收敛特性以及参数选择对算法性能的影响。算法具有较快的收敛速度、鲁棒稳定性且运算小易于实现的特点。计算机仿真的结果与理论分析相一致,证实了算法的收敛性能优于标准的LMS算法并且具有较好的实用性。     

An Lms Adaptive Array Using a Pilot Signal

A least mean square (LMS) adaptive array requires a reference signal. When the desired signal contains a pilot signal, it may be used as the reference signal. In this paper the steady-state performance of an LMS adaptive array in which the pilot signal is used as the reference signal is examined. It is shown that the LMS adaptive array occasionally suppresses the desired signal. The loop gain, which is an important parameter, is also considered.     

An Adaptive Array Signal Processing Algorithm for Communications

An algorithm is described for initial synchronization in a communication system with a digital adaptive array. This algorithm can also be used for message extraction. A set of consecutive complex video samples of the array output is processed to obtain optimum adaptive array weights, based on a least mean square (LMS) error criterion. This computation is performed for each of the possible alternative signals which may be present during an observation interval. The correct synchronization time or message symbol is selected as the one which yields the minimum LMS error. Assuming orthogonality of the alternative codes, a probability distribution for the output of this processor has been derived.     

一种改进的LMS算法及其在自适应消噪中的应用

介绍了一种改进的LMS自适应滤波算法及其在自适应噪声对消中的应用。该算法利用误差信号的相关值去调节步长,解决了算法收敛时间与稳态误差间的矛盾,为实际应用提供了更大的灵活性。它同时实现了均方误差小和收敛速度快,并且降低了LMS算法对噪声的敏感性。文中最后给出了仿真结果,仿真结果与理论分析是一致的。     

A fast adaptive null-steering algorithm based on output powermeasurements

Presented and investigated here is a simple and fast adaptive algorithm for linear power inversion arrays whose nulls can be accurately steered by controlling the element weights. Based on measurements of the powers of the three signals derived from the array output and the output of an auxiliary beamformer, the algorithm tracks unknown jammers in the environment by steering the nulls of the array one by one in a cyclical time-multiplexed manner. When compared with the least mean square (LMS) algorithm, the proposed algorithm has about the same implementation complexity, a better convergence behavior and the advantage that nulls are directly available     

基于比例归一化LMS算法次级通道模型辨识

将主动噪声控制技术应用于地铁通风隧道的噪声控制上,其应用效果在很大程度上取决于信号处理的实时性。在保证控制算法收敛的基础上,要求算法收敛速度尽可能快,以确保对声波变化的准确跟踪,据此产生匹配的振幅相等、相位相反的次级声信号,进而对噪声进行控制。利用比例思想对LMS算法进行改进,为小系数配上小的步长因子,为大系数配上较大的步长因子。最后,将其应用在地铁通风隧道的次级声通道模型的辨识上,并与常用的LMS算法在收敛速度、稳态失调性能上做出比较。数值分析结果表明,采用改进后的LMS算法能够有效地加快算法收敛,改善算法的稳态失调性能。     

A method for aircraft magnetic interference compensation based on small signal model and LMS algorithm

Aeromagnetic interference could not be compensated effectively if the precision of parameters which are solved by the aircraft magnetic field model is low. In order to improve the compensation effect under this condition, a method based on small signal model and least mean square（LMS） algorithm is proposed. According to the method, the initial values of adaptive filter＇s weight vector are calculated with the solved model parameters through small signal model at first,then the small amount of direction cosine and its derivative are set as the input of the filter, and the small amount of the interference is set as the filter＇s expected vector. After that, the aircraft magnetic interference is compensated by LMS algorithm. Finally, the method is verified by simulation and experiment. The result shows that the compensation effect can be improved obviously by the LMS algorithm when original solved parameters have low precision. The method can further improve the compensation effect even if the solved parameters have high precision.     

An Improved Algorithm for Adaptive Processing

The Widrow-Hoff least mean square (LMS) algorithm based on the method of steepest descent is conditionally stable. A modified algorithm is given which is unconditionally stable, capable of better performance when used in adaptive filter processing, and provides a more realistic means for simulating the Applebaum-Howells adaptive loop.     

空间非合作目标惯性参数的Adaline网络辨识方法

空间在轨操作中,航天器在对空间非合作目标的抓捕行动常常导致航天器本体的姿态和空间轨迹发生变化。为克服空间非合作目标对航天器本体动力学、运动学的影响,使控制系统做出精准及时的姿控策略调整,确保航天器正常在轨工作和轨迹姿态的高精度,需对抓捕的非合作目标的惯性参数进行辨识。针对传统辨识方法依赖广义逆求解导致的辨识过程运算量大,且数值容易产生剧烈振荡,造成辨识结果不稳定等不足,采用基于归一化最小均方（NLMS）准则的Adaline神经网络方法进行空间非合作目标惯性参数的辨识。首先,基于动量守恒理论建立抓捕后的航天器-机械臂-空间非合作目标系统模型;然后将辨识方程的系数矩阵作为网络的输入和输出,空间非合作目标的惯性参数作为神经网络的训练权重,基于迭代步长可变的NLMS准则实现对目标惯量参数的快速、准确辨识;最后,在构造的ADAMS/MATLAB联合仿真平台上进行了验证。仿真结果表明,基于NLMS准则的Adaline神经网络是一种快速、准确辨识目标惯量参数的有效方法。     

Adaptive Array Behavior with Periodic Envelope Modulated Interference

A method is presented for determining the effects of envelope modulated interference on a least mean square (LMS) adaptive array. The interference is assumed to have periodic envelope modulation with a bandwidth that is small compared with the carrier frequency. For such interference, the method allows one to calculate the periodic steady-state behavior of the array weights and the resulting array performance. As an example, we compute the effects of an ordinary amplitude modulated (AM) interference signal on the array. It is shown that such interference causes the array to modulate the desired signal envelope but not its phase. With a differential phase-shift-keyed (DPSK) desired signal, AM interference is found to have about the same effect on bit error probability as CW interference.     

Results from a Four-Element Adaptive Antenna Experiment

Experimental results from a four-element, linear, half-wavelength spacing, adaptive-array antenna under the control of the least mean square (LMS) algorithm are presented. The array is found to be capable of nulling a 70-MHz signal to -35 dB below a desired signal over a 5-MHz bandwidth. The antenna processing gain is constant over a desired signal-to-jammer signal power ratio range from -20 dB to 5 dB. A sharp reduction in processing gain is observed for angular separations between jammer and desired signal of less than 10°. Antenna patterns taken with weights set in 300 iterations of the LMS algorithm show that the one strong, one weak jammer combination has a longer weight convergence time and reduced processing gain compared with a two strong jammers combination. Contours of constant desired signal-to-jammer signal power ratio, after adaptive antenna processing, reveal a complex shape for communication between air and ground due to the finite angular resolution of the adaptive antenna.     

航空声纳浮标定位的数据融合方法研究

声纳浮标由于受到多种随机因素的干扰,定位精度相差很大,在对具有一定散布海域的潜艇目标进行探测时,往往会投放多枚定位浮标,出现数据冗余的现象。针对这一问题,本文提出了浮标定位算法中的一种数据融合算法;并分析了加权融合和非线性最小二乘数据融合的方法,最后通过仿真验证了这种算法的有效性和可行性。     

总体最小二乘算法在自适应信道均衡中的应用研究

分析了最小二乘解产生偏差的原因,推导出了最小二乘解偏差项的数学表达式。提出了一种基于瑞利商的变步长总体最小二乘求解方法。仿真试验表明,总体最小二乘算法的性能优于加权递归最小二乘算法,是一种适用于输入信号端含有噪声的自适应信道均衡算法。     

基于变步长LMS算法的线性调频引信噪声抑制

利用自适应滤波技术,研究了线性调频引信通带内噪声的抑制问题。根据线性调频信号与噪声可预测性的差异进行噪声抑制,分析了噪声抑制原理。仿真结果表明,在SNB=-5dB时仍然可以达到很好的噪声抑制效果;在算法收敛的条件下,信噪比越小,噪声抑制对信噪比的相对改善越大。     

新的变步长LMS算法及其在自适应噪声对消中的应用

介绍了一种基于均方误差估计的新变步长LMS自适应滤波算法。与传统的LMS算法相比,该算法具有较快的收敛速度和较小的稳态失调。将该算法应用于自适应噪声对消系统的仿真中,给出了计算机仿真结果,仿真结果与理论分析一致,证明了新算法的优越性。     

Robust algorithm for combating look direction error problems

A simple and fast zero tracking algorithm for adaptive arrays with large look direction errors is presented and investigated. Basically, the algorithm is based on adjusting the complex zeroes of a power inversion array in a time-multiplexed manner to track all the sources in the environment. To preserve the desired signal which is supposed to be closest to the look direction of 0°, the algorithm removes the zero with the shortest distance to e^j0 so that the directional response consists of only nulls steered at the jammers. When compared with the least mean square (LMS) algorithm employing zeroth and first-order look direction constraints, the new algorithm has about the same implementation complexity, is considerably faster, and possesses a much better signal-to-noise ratio (SNR) performance when the look direction is erroneous     

Improved Feedback Loop for Adaptive Arrays

This paper is concerned with the problem of time constants in adaptive arrays. The paper presents a improved form of an adaptive array feedback loop, which has the property that its time constants are fixed. This property is an advantage over the well-known least mean square (LMS) loop, for which time constants depend on received signal power. Fixed time constants are of interest because they simplify dynamic range problems for adaptive arrays in communication and radar systems.     

Phase-only LMS and perturbation adaptive algorithms

The author defines a phase-only gradient-based adaptive algorithm analogous to the least mean square (LMS) algorithm. Two phase-only perturbation algorithms are then defined. It is shown that the gradient for the perturbation algorithms can be obtained by cross-correlating binary perturbation sequences, that are applied to the adaptive phases, with the resulting instantaneous output power or voltage. It is also shown that a single set of phase-only adaptive weights can be used to simultaneously null interference in multiple output beams. Simulation results are presented for all of the new algorithms. The phase-only perturbation techniques eliminate the need for element level receivers and support low cost retrofitting of adaptive nulling on phased arrays by using conventional beamsteering circuits to apply the adaptive weights     

自适应消噪与轴承故障诊断

本文采用自适应消噪技术来提高轴承振动信号的信噪比,在常用的 LMS (widrow)算法的基础上,讨论收敛快而计算量增加又较少的 VS (Variable Step)算法,分析表明 VS算法较LMS算法更适于自适应消噪在机械故障诊断中的应用。对故障轴承的振动信号进行同态自适应消噪处理和异态自适应消噪处理,都有助于发现轴承的故障。