哈默斯坦非线性时变系统的加权学习辨识方法

针对有限区间哈默斯坦(Hammerstein)非线性时变系统,该文提出一种加权迭代学习算法用以估计系统时变参数。首先将Hammerstein系统输入非线性部分进行多项式展开,采用迭代学习最小二乘算法辨识系统的时变参数。为了防止数据饱和,采用带遗忘因子的迭代学习最小二乘算法,进而引入权矩阵,采用加权迭代学习最小二乘算法改进系统跟踪误差,以提高辨识精度。该文分别给出3种算法的推导过程并进行仿真验证。结果表明,与迭代学习最小二乘算法和带遗忘因子迭代学习最小二乘算法相比,加权迭代学习最小二乘算法具有辨识精度高、跟踪误差小以及迭代次数少等优点。     

扩频系统中最优干扰抑制技术研究——盲自适应算法及并行实现

为解决扩频系统在动态环境下的干扰抑制问题,分析了从最小均方误差(MMSE)准则和约束最小均值输出能量(MMOE)准则导出的递推最小二乘(RLS)算法和盲递推最小二乘(BRLS)算法的性能。采用正交三角分解克服两算法数值稳定性差,运算量大,很难并行实现的缺点,讨论了正交三角分解——递推最小二乘(QR-RLS)算法与正交三角分解——盲递推最小二乘(QR-BRLS)算法的配合使用,并给出实现QR-RLS算法和QR-BRLS算法的脉动阵列(systolic array)。理论分析和仿真结果均表明QR-RLS与QR-BRLS算法的合理配合能较好的解决动态环境下的干扰抑制问题。     

基于可变遗忘因子RLS的Boost转换器在线多参数辨识北大核心CSCD

针对Boost转换器控制性能受电感和电容变化影响的问题,提出了一种基于可变遗忘因子递推最小二乘法(recursive least squares method,RLS)的在线多参数辨识算法.考虑电感电流纹波,推导了精确的电感和电容辨识模型.在此基础上,研究了RLS算法中遗忘因子动态取值问题.通过在算法的误差信号中恢复系统噪声的方法,动态计算遗忘因子的取值,解决了传统RLS算法难以兼顾稳态精度和参数跟踪能力的问题.仿真结果表明,该算法可以在动态条件下,精确且快速地跟踪电感和电容值的变化,且具有良好的鲁棒性.     

水下机械手柔顺抓取的阻抗控制方法

针对具有较多不确定性和随机干扰的水下机械手系统，文中分析了水下环境对动力学模型和阻抗控制器的影响，构建了基于环境参数辨识的阻抗控制器来解决水下柔顺抓取的力跟踪问题。通过引入遗忘因子的递推最小二乘法在线辨识环境刚度并修正参考位置，从而减小接触力的稳态误差，实现实时性和准确性的力跟踪。利用MATLAB/Simulink软件平台进行仿真实验，仿真结果表明该控制器在引入持续外部干扰力和期望力调整的情况下，能快速实现对期望力的跟踪和实时估计接触刚度，具有较好的鲁棒性。最后，通过实验验证了该控制器在真实的水下柔顺抓取任务中具有快速稳定的力跟踪性能，其平均稳态力跟踪偏差约为0.014 N,最大力跟踪偏差约为0.138 N。     

RLS算法及其改进形式在信号分离中的应用分析

本文着重研究了自适应滤波器的重要实现形式——递推最小二乘算法（RLS）的原理,分析了RLS算法在应用中的优点及存在问题。为解决RLS算法收敛速度和稳态误差的矛盾及系统在趋于平稳时跟踪效果差的问题,本文从实现可变遗忘因子和增加自扰动项两个方面介绍了RLS算法的几种改进方法。并将它们应用于复杂电磁环境、强干扰背景下的信号分离中去。通过仿真实验,对RLS算法及其两种改进方法在信号分离中的效果进行了比较,得出可变遗忘因子RLS算法在收敛速度和分离信号的准确性上都具有较好的性能。     

一种用于窄带主动噪声控制系统的性能优化算法

针对窄带主动噪声控制(NANC)系统的收敛问题,提出一种变遗忘因子变步长的滤波-X加权累加最小均方算法.本文在滤波-X加权累加最小均方算法基础上,利用互相关的误差信号构建变遗忘因子策略,并通过遗忘因子构造了变步长策略使系统获得更优的参数值,更好的平衡算法的收敛速度、跟踪能力及稳态误差之间的矛盾,同时增强了抗干扰能力,有效提升算法的整体性能.仿真实验表明本文算法在平稳和非平稳环境下具有更好的性能.     

基于迭代变步长LMS的数字域自干扰对消

针对同时同频全双工(Co-frequency and Co-time Full Duplex,CCFD)系统已有的数字域干扰对消方法收敛速度慢和对消比低的问题,本文提出了迭代变步长最小均方(Least Mean Square,LMS)算法,利用该算法实现了快速收敛的高对消比数字域干扰对消.首先,改进Logistic函数,缩短其函数值由大至小的变化区间,再利用该非线性函数计算随迭代次数变化的步长因子值,从而加快干扰对消的收敛速度,高精度递推估计自干扰信道参数,即获得高的对消比.最后,理论分析了该对消方法收敛性和计算复杂度,得到了稳态条件下对消比的闭合表达式.仿真表明,该方法与已有变步长LMS对消方法相比,对消比可增加6dB以上,收敛速度可提高1倍,与最小二乘信道估计干扰对消方法相比,对消比提高了至少10dB.     

一类Wiener非线性时变系统的迭代学习辨识

针对Wiener非线性时变系统的参数辨识问题,该文提出一种基于重复轴的迭代学习算法来实现对时变甚至突变参数的估计.文中将维纳系统输出非线性部分的反函数进行多项式展开,进而构造了回归模型,未知参数及中间变量用其估计替代,分别给出了采用迭代学习梯度算法和迭代学习最小二乘算法实现时变参数辨识的方法.仿真结果表明,与带遗忘因子的递推算法和迭代学习梯度算法相比,迭代学习最小二乘算法更具有参数估计收敛速度快,辨识精度高,系统输出误差小等优势,验证了所提学习算法的有效性.     

严格采样FMT系统频域均衡算法

根据滤波多音调制(FMT)系统基本原理推导了严格采样FMT系统的矩阵表示式,并针对严格采样FMT系统符号间干扰(ISI)严重的问题,根据最小均方误差准则提出了一种子信道频域均衡算法,给出了新算法的最小二乘迭代实现.仿真结果表明,新算法能有效均衡FMT信号,相比较于判决反馈均衡(DFE)算法,新算法的实现结构更简单并且误码率性能更好.     

UFMC系统在多径衰落信道中的干扰消除

通用滤波多载波(UFMC)是5G通信系统中的关键技术,能够降低带外泄露.但是在多径衰落信道下UFMC系统会受到符号间干扰(ISI)和多普勒效应产生的载波频率偏差的影响,从而使系统的性能下降.为了消除系统中的干扰,提出了一种迭代最大似然算法.该方法主要通过迭代最大似然算法(ML)计算出载波频率偏差,把估计出的结果作为初始值并运用迭代的方法得到最终的载波频率偏差,当达到收敛区间时,迭代结束;最后利用相位旋转的概念补偿载波频率偏差并运用最小二乘算法更新信道响应信息,减少该系统干扰.仿真结果表明,在信噪比大于10 dB时,随着信噪比的增大,算法能够有效地抑制系统中的干扰,提高UFMC系统的性能.     

RLS Algorithm with Variable Fogetting Factor for Decision Feedback Equalizer over Time-Variant Fading Channels

In a high-rate indoor wireless personal communication system, the delay spread due to multipath propagation results in intersymbol interference (ISI) which can significantly increase the transmission bit error rate (BER). Decision feedback equalizer (DFE) is an efficient approach to combating the ISI. Recursive least squares (RLS) algorithm with a constant forgetting factor is often used to update the tap-coefficient vector of the DFE for ISI-free transmission. However, using a constant forgetting factor may not yield the optimal performance in a nonstationary environment. In this paper, an adaptive algorithm is developed to obtain a time-varying forgetting factor. The forgetting factor is used with the RLS algorithm in a DFE for calculating the tap-coefficient vector in order to minimize the squared equalization error due to input noise and due to channel dynamics. The algorithm is derived based on the argument that, for optimal filtering, the equalization errors should be uncorrelated. The adaptive forgetting factor can be obtained based on on-line equalization error measurements. Computer simulation results demonstrate that better transmission performance can be achieved by using the RLS algorithm with the adaptive forgetting factor than that with a constant forgetting factor previously proposed for optimal steady-state performance or a variable forgetting factor for a near deterministic system.     

Code-aided interference suppression for DS/CDMA communications. II.Parallel blind adaptive implementations

For pt.I see ibid., vol.45, no.9, p.1101-11 (1997). An adaptive code-aided technique for the simultaneous suppression of narrow-band interference (NBI) and multiple-access interference (MAI) in direct-sequence code-division multiple-access (DS/CDMA) networks is proposed. This technique is based on the recursive least-squares (RLS) version of the minimum mean-square error (MMSE) algorithm for multiuser detection. The convergence dynamics of the RLS blind adaptive algorithm for suppressing the combined NBI and MAI are analyzed. The steady-state performance of this algorithm in terms of the signal-to-interference ratio (SIR) is also derived. Systolic array structures for parallel implementations of the RLS adaptive interference suppression algorithms are then proposed. Versions of the rotation-based QR-RLS algorithms for both the blind adaptation mode and the decision-directed adaptation mode are derived. These algorithms exhibit high degrees of parallelism, and can be mapped to VLSI systolic arrays to exploit massively parallel signal processing computation     

Performance analysis of adaptive receivers for DS/CDMA communication systems

Both intersymbol interference (ISI) and multiple-access interference (MAI) are the important issues in code division multiple access (CDMA) multiuser communication systems. The step size of the traditional least-mean-square (LMS) adaptive receivers must be substantially adjusted to overcome the effect due to different channel responses or different numbers of active users. The normalized-LMS (NLMS) algorithm can automatically manipulate the adjustment to avoid the problem just discussed. This paper proposes the symbol-based interference rejection filter with NLMS algorithm and derives both finite impulse response (FIR)- and infinite impulse response (IIR)-type algorithms. We also derive the optimal step sizes and minimum mean-square errors (MSEs) for both the FIR and IIR symbol-based receivers. The complexity of our proposed FIR receiver is lower than that of the conventional chip-based receivers. In addition, it is shown that the bit error rate performance of our proposed symbol-based receivers is superior to conventional one in the simulations. Simulations also show the correctness of our theoretical analysis of minimum MSE. Copyright © 2008 John Wiley & Sons, Ltd.     

Adaptive step-size algorithms for blind interference suppression inDS/CDMA systems

This paper develops adaptive step-size blind LMS algorithms and adaptive forgetting factor blind RLS algorithms for code-aided suppression of multiple access interference (MAI) and narrowband interference (NBI) in DS/CDMA systems. These algorithms optimally adapt both the step size (forgetting factor) and the weight vector of the blind linear multiuser detector using the received measurements. Simulations are provided to compare the proposed algorithms with previously studied blind RLS and blind LMS algorithms. They show that the adaptive step-size blind LMS algorithm and adaptive forgetting factor blind RLS algorithm field significant improvements over the standard blind LMS algorithm and blind RLS algorithm in dynamic environments where the number of interferers are time-varying     

Performance of multibeam CDMA-based LEO satellite systems with imperfect power control

Bit error probability of a DS-CDMA multispot beam LEO system with imperfect power control is analysed in this paper. The power control error is assumed to be a lognormal random variable. The channel is a combination of Rician and Suzuki distributions. For more realistic applications the multiple-access interference due to all users seen by the serving satellite for the user of interest is taken into account. This differs from other reported work in which only part of all possible interference was considered. The numerical results with different numbers of interferers show that the interference coming from the users located outside the footprint but seen by the reference satellite has a non-trivial effect on the system capacity. However, the multiple-access interference can be reduced by increasing the elevation restraint of each satellite footprint. Estimates on this reduction are given. It is also shown that the performance of shadowed users is highly sensitive to the power control error. © 1998 John Wiley & Sons, Ltd.     

频谱相位编码OCDMA系统的误码特性研究

本文利用中心极限定理和雅可比变换的方法，对频谱相位编码(spectral phase encoding)OCDMA系统中多址干扰和解码器输出信号的统计特性进行了研究，并据此推导出多址干扰和背景噪声共同作用下系统新的误码率公式．本文还分析了系统性能与判决门限选择的关系，数值结果表明：系统最佳的判决门限Rth=√P0/2，在此判决门限的条件下，系统的性能满足高速全光通信的需要．     

同步CDMA系统中的MMSE检测器研究

分析了基于实域估计的最小均方误差(MMSE)检测器及其在同步CDMA系统的应用。首先对同步CDMA系统进行了描述,并在此基础上对这种新的MMSE检测器在同步CDMA系统中的设计和近似实现方法进行了探讨。     

基于遗传算法的RLS自适应算法研究

在通信系统中采用信道均衡技术是改善信道特性行之有效的方法,但研究算法的同时,往往需要通过大量的仿真实验取平均值来选取最优的参数值,本文首先利用MATLAB仿真软件对线性调制下RLS自适应算法进行仿真分析,然后引入遗传算法的寻优特性及其优点,对RLS最佳遗忘因子λ的选取进行了寻优,得出最佳遗忘因子λ的取值,提供了参数选择的一条捷径,最后通过对比最优λ与参照λ,计算RLS算法均衡已知信号的均方误差值,证明了该方法的可行性。     

扩频系统中最小误码率意义下的最优干扰抑制技术

通过论证最小化误码率(MBER),最小均方误差(MMSE)以及约束最小均值输出能量(MMOE)之间的关系,将MBER准则下最优干扰抑制器的设计转化为后两种准则下最优干扰抑制器的设计,并分别导出两种自适应算法:递推最小二乘(RLS)和盲递推最小二乘(BRLS).前者抑制干扰效果好,但需要期望信号;后者无需期望信号,但抑制效果较差.本文将两种算法合理配合,给出了动态环境下的干扰抑制方法.