移动通信信号自动调制识别的谱相关方法

一些不同的数字调制信号有着相同或相近的功率谱密度,但它们的谱相关函数却有明显区别。平稳的噪声和干扰的谱相关为零。利用谱相关函数的这些性质可以识别出噪声中的调制信号。本文提出了一种用于移动通信信号自动调制识别的谱相关算法,该方法可以识别各种主要无线系统和标准所采用的调制类型。     

基于谱相关分析的通信信号监测方法研究

在介绍谱相关函数原理、分析谱相关识别特性基础上，利用谱相关理论对常规通信信号调制模式进行分析识别，并仿真观察了其性能的优点与不足。     

基于谱相关函数截面特征的调制方式识别

谱相关函数在双频平面上全面反映了循环平稳信号的二阶周期性。调制信号是循环平稳信号,不同调制信号的谱相关函数具有不同的分布特性,且对平稳噪声、干扰不敏感,可用于低信噪比下的调制方式识别。采用时域平滑FFT积累算法计算出调制信号的谱相关函数,对结果进行了对比分析,提取了多种调制信号的谱相关函数截面特征,设计了AM、MSK、PAM、2ASK、2FSK、4FSK、QPSK七种调制信号的树形识别流程。仿真结果表明在低信噪比下,七种调制方式具有90%以上的识别率。     

伪码-载波调频侦察信号识别的谱相关方法(Ⅱ)伪码-载波调频信号的调制识别和参数估计

研究了伪码-载波调频侦察信号识别的谱相关方法,具体针对伪随机二相编码(PRBC)信号及其与正弦调频和线性调频复合信号作了研究。通过3种复合信号循环谱截面谱相关特征的分析和提取,完成了调制类型的自动识别,估计了载频和载波调频参数,对伪随机码信号的各项参数进行了估计。分析了噪声环境和有限采集时间对参数估计的影响,最后对调制识别和参数估计进行仿真分析,表明了谱相关方法在伪码-载波调频信号识别中的优势。     

一种数字信号调制方式识别方法

数字信号调制方式的识别是近年来迅速发展起来的一门新技术,它是信号分析领域的重要组成部分。调制识别技术具有很高的实用价值。在军用和民用方面都有重要应用,如通过调制识别技术查清相关信号的参数和性质。这里针对短波信道条件下的MPSK信号调制识别问题,研究了一种基于功率谱和谱相关为特征的调制方式的分类与识别方案。该方案中提取了以信号功率谱的各种特征为主要的特征参数,通过对BPSK、QPSK、8PSK几种常用的数字信号进行仿真识别表明,该方法简单便于工程实现,具有很好的实用性。     

改进的通信调制信号谱相关识别方法

传统谱相关分类法利用不同信号在a截面上所表现的不同谱相关特征可对通信信号进行分类,需要提取六种特征参数。文章结合小波分析法估计信号的码元宽度Tt,利用谱相关提取的三种特征参数,可对常见八类调制信号进识别。仿真表明,该方法具有很好的识别效果。     

基于谱域联合特征的信号调制类型识别

为了在低信噪比条件下识别信号的调制类型,提出了一种由信号瞬时包络的高阶组合矩、零中频信号的高阶组合累积量和谱相关函数共同构建的联合特征,用于数字调制信号调制类型识别的方法。新方法分析了多种常规数字调制信号的瞬时包络的幅度分布特性,利用了高斯过程的高阶累积量当阶次大于2时恒等于零的特性和谱相关函数突出周期平稳信号的特性,对高斯噪声具有较好的抑制作用。仿真试验结果表明,该方法运算简单,适应性强,在低信噪比下正确识别率高。     

基于信号谱线特征的调制方式识别

针对通信信号调制识别,提出了一种基于信号谱线特征的调制识别方法。该方法采用信号平方谱和四次方谱的强度与位置作为信号的特征参数。信号的这种谱线特征与信号调制样式有着密切的内在联系,较好地反映了信号调制特性。相比其它调制样式识别方法,这种方法计算复杂度低,在非理想环境下对信号特征提取较为容易并且这种特征有较强的鲁棒性。对于调制样式多变的软件无线电和认知无线电有较好的应用效果。仿真表明,在低信噪的情况下,对不同的调制样式识别概率较高。     

数字调制信号通过循环谱与参数统计进行识别的研究

数字调制信号通过运用循环谱与参数统计的方法进行识别可以在信噪比较低的情况下得到较好的效果,同时也是数字调制信号识别长久以来研究的主要方向。因此本文对数字调制信号通过循环谱与参数统计进行识别的研究进行浅析。     

基于对称相关谱和累积量切片的QAM信号识别

随着正交振幅调制信号(Quadrature Amplitude Modulation,QAM)的广泛应用,对其调制方式的识别显得至关重要。针对M-QAM的四次方谱谱线特征在低信噪比时对调制信号识别率低,且计算复杂度高等问题,提出了一种基于信号四次方的对称相关谱和四阶累积量的一维切片值的M-QAM调制方式的盲识别算法。该算法无需波特率和采样率等先验信息,没有复杂的迭代过程,可以应用于实际信号的调制识别,对NI USRP-2930无线接收平台接收的信号进行识别。实验结果表明,提出的算法在低信噪比下具有较高的识别率、实用性和有效性。     

Rice's Representation for Cyclostationary Processes

The well-known formulas relating power spectral densities of a bandpass process and its low-pass representors in Rice's representation are generalized to accommodate cyclostationary processes, which can be more appropriate than stationary processes as models for modulated signals. It is explained that improper phase-randomization methods for stationarizing cyclostationary processes can yield misleading formulas for power spectral densities in Rice's representation. In addition, the relationships among the spectral correlation functions for a cyclostationary bandpass process and its cyclostationary low-pass representors are derived.     

基于椭圆球面波信号的连续相位调制及其频谱性能分析

椭圆球面波信号（PSWF）是时频域最佳能量聚集性信号,探索性地将PSWF信号应用于连续相位调制（CPM）技术,对基于椭圆球面波信号的CPM调制信号产生过程进行了描述;给出了采用自相关函数法对基于PSWF的CPM调制信号功率谱密度计算方法;结合数值计算、分析对比了最小频移键控、正弦频移键控、高斯最小频移键控以及基于高斯函数的CPM调制信号功率谱密度及其信号占用带宽等特性。数值计算结果表明椭圆球面波信号相比矩形脉冲、升余弦脉冲、高斯脉冲及（类）高斯脉冲作为CPM的基带调频信号,可以获得频谱性能、能量聚集性较好的CPM信号。     

Application of classical cosine series window functions to fullresponse signaling offset quadrature binary modulation systems

It is proposed that classical cosine series window functions be applied to finite symbol duration quadrature binary modulation systems as pulse shapes in the interest of increasing the spectral confinement of the resulting modulated signals. The respective modulation systems are analyzed in terms of modulated signal envelope uniformity, spectral confinement, and bit error rate in the presence of white Gaussian noise with varying degrees of modulated signal amplitude compression. Even in the presence of moderate modulated signal compression, classical cosine series window function pulse shapes offer spectral confinement for quadrature binary modulation systems that is equal to or better than that provided by the conventional pulse shapes corresponding to QPSK, MSK, SFSK, and DSFSK. It is shown that moderate levels of modulated signal compression have negligible effects on the system bit error rate. A system implementation using the sum of FSK modulated signal components to achieve spectral confinement based on the classical cosine series window functions is presented     

Corrélation spectrale théorique des modulations CPM partie I : résultat analytique pour les modulations cpfsk à 2 états (1-rec)

All modulated signals used in digital communication activities are cyclostationary, because their autocorrela tion function contains some temporal periodicities known as cyclic periods. In the frequency domain, their spectral components are correlated every time they are spaced apart by the inverse of a cyclic period, called a cyclic frequency. A number of signal processing algorithms applied to telecommunications can take advantage of the spectral correlation of the signals to be considered (detection, equalisation, source separation,). Furthermore, the knowledge of the theoretical expression of the spectral correlation of the signal to be received is often necessary in order to derive the optimal processing. The theoretical spectral correlation of modulated signals has already been the subject of some publications. However, until now there was no comprehensive reference concerning the large family of continuous phase modulations (cpm). The aim of this two-part paper is to give the theoretical expression of the spectral correlation for binary cpm signals. The first part is devoted to the study of rectangular, full-response binary cpm modulations, also called continuous phase frequency shift keying modulations (cpfsk). It gives the exact analytical expression for their spectral correlation. The second part presents a simple computation method valid for the whole class of binary cpm modulations. This method requires only the computation of some double numerical integrations over finite intervals.     

Complex Spectrum Analysis of Modulated Optical Signals Using Stimulated Brillouin Scattering

We introduce a novel method for the measurement of the spectral phase of modulated optical signals based on double-pump stimulated Brillouin scattering interaction and direct homodyne detection. This technique is combined with Brillouin high-resolution analysis to achieve a complete characterization of the complex optical spectrum. A 10-Gb/s nonreturn-to-zero signal with a 32-bit pattern modulated with an electroabsorption modulator is measured using this technique and its time-domain information (power and chirp) is obtained by inverse Fourier transform.      

分数阶傅里叶变换（FRFT）用于调制方式识别

提出了一种基于分数阶傅里叶变换（FRFT）的通信信号调制方式自动识别方法。给出了FRFT归一化修正零均值幅度这一特征,用于模拟调制信号与数字调制信号联合自动识别。新方法分析了多种常用调制信号的FRFT谱特性,利用分数阶傅里叶变换可以展示出信号从时域逐步变化到频域的所有变化特征的特性,可以很好地分类识别多种调制信号。实测数据和仿真结果表明,在调制信号的信噪比大于10dB时,其总体识别率达90%以上。该方法具有识别效率高,识别效果好等优点,并具有实用性和可行性。     

Transmission of 10 Gbit/s directly modulated DFB signals over 200km standard fibre using mid-span spectral inversion

Dispersion compensation by mid-span spectral inversion has been used to achieve the transmission of 10 Gbit/s signals from a directly modulated DFB over 200 km dispersive fibre. The spectral inversion is provided by nondegenerate four-wave mixing in a semiconductor optical amplifier     

Coherent estimates of correlation characteristics of interconnected periodically correlated random processes

Brief analysis of correlation and spectral characteristics that describe interconnections between two periodically correlated random processes is conducted. Properties of coherent estimates of signals?? mutual correlation function and estimates of mutual correlation components are studied. Expressions for biases and dispersions of estimates are specified for amplitude modulated signals.     

基于稀疏迭代的非连续谱高频雷达信号处理方法

在拥挤频谱环境中,雷达系统在多个非连续频段发射信号并在接收端进行综合相干处理,是一种获得等效大带宽的方式. 本文专门就非连续谱调频连续波(discontinuous spectrum-frequency modulated continuous wave, DS-FMCW)及其在高频雷达中的应用展开研究. 首先提出采样点平移方法,建立DS-FMCW快时间维的非均匀采样序列谱估计模型;随后,进一步建立DS-FMCW距离-多普勒二维谱估计模型,提出解决距离徙动的方案;最后为解决距离高旁瓣问题,基于一种适用于单次快拍的迭代式稀疏重构算法提出DS-FMCW的距离与距离-多普勒谱估计方法,并提出相应的快速谱求解方法. 仿真试验表明:所提DS-FMCW距离-多普勒处理方案能有效补偿距离徙动;当频带利用率大于20%时,所提谱估计方法能够稳定地分辨距离维间隔为雷达固有距离分辨率的1/3的两个目标,且距离估计精度优于经典最小二乘算法以及正交匹配追踪算法;所提快速算法单次迭代运算量低,适用于实时系统.