Parameter estimation method for a linear frequency modulation signal with a Duffing oscillator based on frequency periodicity

In view of the complexity of existing linear frequency modulation(LFM)signal parameter estimation methods and the poor antinoise performance and estimation accuracy under a low signal-to-noise ratio(SNR),a parameter estimation method for LFM signals with a Duffing oscillator based on frequency periodicity is proposed in this paper.This method utilizes the characteristic that the output signal of the Duffing oscillator excited by the LFM signal changes periodically with frequency,and the modulation period of the LFM signal is estimated by autocorrelation processing of the output signal of the Duffing oscillator.On this basis,the corresponding relationship between the reference frequency of the frequencyaligned Duffing oscillator and the frequency range of the LFM signal is analyzed by the periodic power spectrum method,and the frequency information of the LFM signal is determined.Simulation results show that this method can achieve high-accuracy parameter estimation for LFM signals at an SNR of-25 dB.

Gauss linear frequency modulation wavelet transforms and its application to seismic phases identification

Based on the characteristics of gradual change style seismic signal onset which has more high frequency signal components but less magnitude, this paper selects Gauss linear frequency modulation wavelet as base function to study the change characteristics of Gauss linear frequency modulation wavelet transform with difference wavelet and signal parameters, analyzes the error origin of seismic phases identification on the basis of Gauss linear frequency modulation wavelet transform, puts forward a kind of new method identifying gradual change style seismic phases with background noise which is called fixed scale wavelet transform ratio, and presents application examples about simulation digital signal and actual seismic phases recording onsets identification.

A TECHNIQUE FOR MEASURING THE LINEARITY OF A LINEARLY FREQUENCY-MODULATED SIGNAL

A technique for measuring the linearity of a linearly frequency-modulated continuous wave (LFM-CW) signal is presented. It uses a delay-line and a mixer to sense the slope of the output of a sweep oscillator, so that the original form of frequency function deviated from idealized linear slope is retrieved by means of spectrum analysis. Consequently,the linearity of the LFM signal is determined. The formulation is performed based on the principle that an angle-modulated signal can be approximated by an amplitude-modulated signal if the modulation coefficient is sufficiently small. To examine the validity of the procedure and to study the effect of each parameter on the accuracy of measurement, a number of computer simulations has been made. The results of simulation show that the error of the measurement is less than 2%.

NEW TYPE OF WINDOWS FOR DIGITAL PULSE COMPRESSION IN FREQUENCY DOMAIN FOR BROADBAND LFM SIGNALS

A new type of window called combined window is designed to get higher Ratio of Mainlobe to Sidelobes (RMS) and lower Mainlobe Widening Factor (MWF). Simulation results prove that the new window can solve the contradiction between RMS and MWF better than classic windows.

Frequency Sweep Linearization for Semiconductor Laser Using a Feedback Loop Based on Amplitude-Frequency Response

A scheme of frequency sweep linearization of semiconductor lasers using a feed-back loop based on amplitude-frequency response is demonstrated in this paper. The beat frequency signal is obtained by self-heterodyne detection. The frequency changes are converted to the envelope of beat frequency signal after amplitude-frequency response. The active frequency sweep linearization is realized by feeding envelope deviations back to the drive currents of the lasers by a feedback loop. A simulation model is built to verify this scheme by Simulink. This scheme does not need high-performance, expensive lasers, complex linearization or tedious post-processing processes, which are of great significance for related applications.

多分量LFM干扰及动态环境下DBOC信号的捕获

针对双二进制偏移载波(double binary offset carrier,DBOC)信号在多分量线性调频(linear frequency modulation,LFM)干扰和动态环境下捕获算法缺乏的问题,提出布莱克曼窗预处理结合分数阶傅里叶变换(fractionalFouriertransform,FRFT)的干扰抑制算法以及离散多项式相位变换(discretepolynomial-phase transform,DPT)、部分匹配滤波器(partial matching filter,PMF)结合快速傅里叶变换(fast Fourier transform,FFT)加频谱校正的捕获算法。该算法首先对接收信号做布莱克曼窗预处理,抑制干扰谱线输出旁瓣,提高干扰汇聚的精确性;然后利用分级算法快速搜索信号的最优阶数,在保证精度量级的条件下减少了运算量;最后,在对应的最优阶数下做FRFT,结合干扰判断和轮询算法,逐个完成多分量LFM干扰的抑制。对干扰抑制后的信号首先利用定阶算法确定其动态阶数;然后通过DPT对动态信号做降阶处理,去除动态项;最后利用PMF-FFT加频谱校正完成DBOC信号的捕获。仿真结果表明,所提算法能够准确并完全地对三个LFM干扰分量进行逐个抑制,且抑制后的干扰分量不会对其他干扰分量的抑制造成影响。在检测概率达到1时,经过布莱克曼窗预处理后算法的信噪比比经过汉明窗预处理和经过凯塞窗预处理后算法的信噪比分别提升了4 dB和6.4 dB,比未经过预处理算法的信噪比提升了7.2 dB。此外,使用频谱校正可以突出捕获峰值,提高捕获精度。与未经频谱校正算法相比,经过频谱校正后算法的信噪比提升了1.6 dB,且能减少一定的捕获时间。

基于FRFT的低信噪比LFM信号参数快速估计算法

基于分数阶傅里叶变换(Fractional Fourier Transform,FRFT)对线性调频(Linear Frequency Modulated,LFM)信号参数进行估计,问题关键是确定FRFT最佳阶数,根据误差迭代思想提出新的参数估计算法,该算法利用归一化带宽和旋转角的转化关系,由估计误差推算角度差值,有效降低了运算量,不需要调频斜率正负的先验信息,改进的对数搜索算法可以进一步提高参数估计结果的稳定性和可靠性。仿真结果表明,信噪比在-8 dB以上时该方法在高效率的前提下仍具有良好的参数估计性能,平均估计误差在1%以内,估计结果接近Cramer-Rao下限,满足工程实时处理需求。

ANALYSIS AND PROCESSING OF LMF-COSTAS STEPPED FREQUENCY RADAR SIGNAL

Linear Modulated Frequency (LMF) Costas Stepped Frequency Radar (SFR) signal is introduced. Its ambiguity function is derived and analyzed in detail and its feasibility is validated in theory. The scheme of the proposed signal processing is also presented. The results of theoretic analysis and simulation show that, by using the proposed signal and increasing the bandwidth of the total stepped frequency, the ambiguity sidelobe is well suppressed and the range-velocity coupling in the stepped frequency radar is also greatly weakened.

基于Born-Jordan分布的LFM-BPSK信号的参数估计

对线性调频-二进制相位编码(LFM-BPSK)复合调制信号的参数估计,现有方法在低信噪比(SNR)环境下估计精度较低。针对该问题,文中推导出了LFM-BPSK信号的Born-Jordan(BJD)时频分布,分析了LFM-BPSK信号的时频特性,并提出了基于BJD时频分布的参数估计算法。首先,该算法在BJD时频变换的基础上使用霍夫变换对LFM-BPSK信号初始频率和调频斜率进行估计,并将二维的BJD时频变换转化为一维时域信号,通过降维后的时域信号求解码元宽度;然后,为了探寻参数估计的理论下界,文中还推导了LFM-BPSK信号参数估计的修正克拉美罗下界(MCRLB);最后,利用蒙特卡洛实验对三个参数进行了测试。仿真实验表明,文中算法在低SNR下的参数估计精度明显优于现有的算法,其性能接近MCRLB。

脉冲噪声下基于DCNN的LFM信号去噪方法

由于脉冲噪声具有明显的尖峰脉冲特性,使得基于高斯假设的传统去噪方法无法有效滤除脉冲噪声。针对这个问题,文中提出了一种脉冲噪声下基于深度卷积神经网络(DCNN)的线性调频(LFM)信号去噪方法。首先,生成LFM信号和随机脉冲噪声,构建不同广义信噪比下的数据集,输入DCNN进行训练和测试。进而,从时域波形图、分数谱、时频分布三个方面验证模型的去噪能力。最后,对去噪LFM信号进行分数阶傅里叶变换,通过搜寻分数谱中的峰值点来估计LFM信号的参数。仿真实验结果表明,文中方法不仅能够有效去除含噪信号中的随机脉冲噪声,而且还可以保持LFM信号的时域特征、分数谱特征和时频特征基本不变,进而提高了参数估计的噪声鲁棒性。与传统的基于非线性变换的方法相比,本文方法在低信噪比下仍能有效保持信号的分数谱特征和时频特征,具有更好的去噪性能和泛化能力。

A High-Linear Radio Frequency Quadrature Modulator with Improved Sideband Suppression and Carrier Leakage Performance

The quadrature modulator is a crucial block in transmitters that upconverts baseband signals to theRadio Frequency(RF)band of interest using local oscillator frequencies.In this paper,non-ideal factors thatinfluence the performance of the quadrature modulator are considered,and solutions are accordingly taken inthe quadrature modulator design.A high-linear RF quadrature modulator with improved sideband suppressionand carrier leakage performance is presented in this work.The quadrature modulator implemented in the0.18-μm SiGe process uses the current bleeding technique to improve the general performance of the doublebalanced active Gilbert mixers.An on-chip prescaler followed by two cascaded limiting amplifiers is designed toprovide accurate quadrature local oscillator signals.Predrivers at quadrature baseband signal input ports areproposed to eliminate DC offsets.The measured sideband suppression achieves a performance of better than−43 dBc and carrier leakage is less than−38 dBm over the output RF frequency range of 30 MHz to 2.15 GHz.The output 1 dB compression point equals 11.4 dBm at 800 MHz.

转发式干扰条件下LFM雷达的检测概率估计

对多雷达干扰效果评估的难点在于分析干扰机对非目标雷达的影响。针对干扰机转发一部线性调频(linear frequency modulation,LFM)雷达的信号无意干扰另一部LFM雷达时的检测概率估计问题,首先计算了被干扰雷达对干扰信号在空域、频域和时域上的接收增益;其次将干扰信号在判决单元中的统计特性建模为非中心卡方分布,并给出了检测概率和虚警概率的计算方法。仿真结果表明,在一部干扰机的无意干扰下,估计的检测概率可作为被干扰雷达检测概率的上限;为达到同样的干扰效果,转发非目标雷达的信号实施干扰,比转发目标雷达的信号进行干扰所需的干扰功率更少。

基于Z变换和改进FRI的LFM信号参数估计方法研究

为了完成线性调频(linear frequency modulation,LFM)信号的稀疏采样,并利用稀疏数据对原始信号参数进行估计,本文提出了一种基于Z变换和改进有限新息率(finite rate of innovation,FRI)的LFM信号参数估计方法。以Z变换理论为基础,设计了一种数学模型,一旦信号能够表达成该数学模型的结构形式,就能通过Z变换和零化滤波器的方法估计信号参数。然后,利用了自相关延迟的FRI结构对LFM信号采样,该结构不仅完成了LFM信号的稀疏采样,而且稀疏采样结果能够与数学模型结构相符。在理论上通过数学论证的方式证明了所提方法能够用于获取LFM信号参数信息,并通过仿真和实测数据验证了所提方法的有效性,理论和实验结果表明该方法只需要4个采样点就能实现对LFM信号的参数估计,并且实验中的参数估计误差均在3%以内,极大的提高有限新息率采样的参数估计效率。

MSK-LFM一体化信号波形相关性分析与参数选择

以二进制数字信号最佳接收准则为依据,分析了MSK-LFM雷达通信一体化信号不同码元对应波形的相关性。利用fresnel函数得到了波形相关系数的表达式;通过数值仿真得到了相关系数随LFM信号的初始频率、调频率、脉冲内码元调制速率等参数的变化关系;构建MSK-LFM信号正交调制、解调仿真模型进行验证。研究表明,通过增大初始频率、调频率或减少调制速率可降低信号误码率;当初始频率远大于脉冲内码元调制速率时可认为不同码元波形正交。

基于MSK-LFM的光载太赫兹通感信号产生与传输方案

针对第六代移动通信(The 6th Generation Mobile Communication,6G)中一体化需求感知精度较低的问题,提出一种基于最小频移键控(Minimum Shift Keying,MSK)与线性调频(Linear Frequency Modulation,LFM)的光载太赫兹通感一体化信号的产生与传输方案。利用MSK-LFM信号在太赫兹频段同时实现通信和感知测距,在通信接收端使用数字信号处理降低无线通信误码率,在感知接收端利用去啁啾技术进行测距。仿真结果表明,MSK-LFM信号在235 GHz太赫兹频段能够实现8 Gbit·s^(-1)的无线通信速率,在距离感知精度上达到了3.75 cm的距离分辨率,且通信误码率低于软判决阈值。当直流偏置为0.8时,8 GHz带宽的MSK-LFM信号通感性能达到平衡,该方案与M相移键控与线性调频(Phase Shift Keying,Linear Frequency Modulation,MPSK-LFM)信号方案相比,在相同通信速率下具有更高的感知测距精度。

基于高斯短时分数阶傅里叶变换的多分量LFM信号检测与参数估计

该文针对线性调频信号,提出一种基于分数阶波包变换分析方法——高斯短时分数阶傅里叶变换。通过旋转角度的搜索及高斯窗口宽度的调整,能够在低信噪比条件下对多分量LFM信号进行检测,避免交叉项的出现,并能得到参数的估计值。通过推导分析给出变换结果的解析表达式,计算机仿真结果也表明了该变换的有效性。

强干扰下LFM信号的检测与参数估计

强干扰下ＬＦＭ信号的检测与参数估计在雷达信号处理领域具有重要的应用。本文首先分析了强干扰下ＬＦＭ信号的特点，对ＬＦＭ信号的检测与参数估计方法进行了总结和比较；随后研究了ＷＨＴ法的检测性能，绘制了ＷＨＴ检测器的接收机性能曲线（ＲＯＣ）。最后，提出一种基于ＷＨＴ的抑制强干扰和进行ＬＦＭ信号检测与参数估计的方法，并给出了实验结果。

LFM脉冲压缩雷达的随机移频多假目标干扰技术研究

针对LFM脉冲压缩雷达回波信号距离和多普勒频率存在耦合这一特点,在阶梯波移频干扰的基础上,提出了随机移频的干扰方法。在雷达回波信号中心频率正负二分之一带宽内附加随机的多普勒移频分量,可产生分布在真实目标周围的随机假目标。随机移频干扰技术产生的假目标随机性强,使得雷达无法通过假目标频率补偿得出真实目标的位置,并且随机移频产生的假目标相对于雷达是部分适配的,因此假目标在一定程度上会展宽,当附加随机移频点接近于雷达中心频率时,产生的假目标甚至可以覆盖真实目标,从而达到较好的干扰效果。

高精度LFM信号参数估计的谱校正方法

为精确估计噪声背景下线性调频(linear frequency modulation,LFM)信号的调频率和中心频率等参数,提出了一种基于Radon模糊变换(Radon-ambiguity transform,RAT)和分数阶傅里叶变换(fractional Fourier transform,FRFT)的离散谱校正估计方法。该方法将LFM信号的调频率和中心频率估计问题转化为模糊域和分数阶Fourier域上的两次一维谱峰搜索过程,并且通过对矩形窗截断的LFM信号的RAT和FRFT谱线结构分析,将能量重心谱校正法引入两次谱峰搜索过程,在不增加运算量的基础上实现了谱峰位置的超分辨率估计,较大幅度地提高了信号的参数估计精度。仿真结果显示,对于单分量LFM信号情况,该方法对信号的RAT和FRFT谱峰位置的校正误差分别降到了扫描步长的1.55%和4.94%,证明了该方法的有效性。