计算机应用 ›› 2016, Vol. 36 ›› Issue (10): 2927-2932.DOI: 10.11772/j.issn.1001-9081.2016.10.2927

• 行业与领域应用 • 上一篇    下一篇

基于部分相关的LFM脉冲全参数估计

王思秀1, 徐舟2, 汪晓洁1, 汪江桦1   

  1. 1. 新疆财经大学 计算机科学与工程学院, 乌鲁木齐 830012;
    2. 电子工程学院 雷抗系, 合肥 230000
  • 收稿日期:2016-03-31 修回日期:2016-06-26 发布日期:2016-10-10
  • 通讯作者: 王思秀,E-mail:mypapershow@sohu.com
  • 作者简介:王思秀(1981—),男,江苏徐州人,讲师,硕士,主要研究方向:信号分析、数据处理;徐舟(1990—),男,辽宁沈阳人,助教,硕士,主要研究方向:SAR信号处理、SAR图像分析;汪晓洁(1980—),女,安徽黄山人,讲师,硕士,主要研究方向:计算机网络;汪江桦(1982—),女,湖北黄冈人,副教授,博士,主要研究方向:信息处理、数据挖掘。
  • 基金资助:
    国家自然科学基金资助项目(61562079);新疆维吾尔自治区高校科研计划青年教师科研培育基金资助项目(XJUEDU2014S042,XJUEDU2014S046)。

Whole parameters estimation for linear frequency modulation pulse based on partial correlationtle

WANG Sixiu1, XU Zhou2, WANG Xiaojie1, WANG Jianghua1   

  1. 1. College of Computer Science and Engineering, Xinjiang University of Finance and Economics, Urumqi Xinjiang 830012, China;
    2. Department of Radar Countermeasures, Electronic Engineering Institute, Hefei Anhui 230000, China
  • Received:2016-03-31 Revised:2016-06-26 Published:2016-10-10
  • Supported by:
    BackgroundThis work is partially supported by the National Natural Science Foundation of China (61562079), the Foundation for Young Teachers Scientific Research and Cultivation of Xinjiang Uygur Autonomous Region (XJUEDU2014S042,XJUEDU2014S046).

摘要: 针对线性调频(LFM)脉冲信号的侦察问题,提出了调频率、中心频率、信号到达时间、脉宽全套参数集的估计方法。首先,使用分数阶傅里叶变换(FrFT)对信号的调频率与时频关系进行估计;紧接着选取部分相关脉冲对信号进行积累,利用自相关处理完成中心频率、信号到达时间、脉宽参数的估计;然后推导了估计参数的克拉美劳下界(CRLB),分析了信噪比对估计误差的影响;最后分析了部分积累脉宽对估计误差的影响,给出了积累脉宽的选择范围。仿真分析表明,调频率估计误差几乎达到CRLB,在信噪比0 dB、基带与调制参数均未知的条件下,中心频率估计均方根误差约为10-1MHz数量级,信号到达时间和脉宽估计均方根误差约在10-1 μs数量级。参数估计误差受到相关脉冲宽度的影响,随着相关脉冲宽度的增加,估计误差呈现先减小后增大的趋势。所提方法特别适用于脉压、合成孔径等新体制雷达的侦察。

关键词: 线性调频脉冲信号, 参数估计, 分数阶傅里叶变换, 部分相关, 克拉美劳下界

Abstract: Focusing on the reconnoitering problem of Linear Frequency Modulation (LFM) pulse signals, a method to estimate the whole parameters containing frequency modulate rate, center frequency, time of arrival and pulse width, was proposed. Firstly, frequency modulate rate as well as time-frequency relation was estimated based on Fractional Fourier Transform (FrFT), then partial correlation pulses was used for signal accumulation, at last the autocorrelation technology was used to estimate the center frequency, time of arrival and pulse width. The Cramer-Rao Low Bounds (CRLB) for the parameters were derived and the effect on estimation error caused by signal to noise ratio was analyzed. Finally, the effect on estimation error caused by the width of partial accumulation pulse was analyzed, and some advice was given on choosing the width of accumulation pulse. Simulation results show that the estimation error of frequency modulate rate is close to CRLB. When signal to noise ratio is 0 dB without any knowledge of baseband and modulation parameters, the Root Mean Square Error (RMSE) of center frequency is about 10-1MHz orders of magnitude, and the RMSE of time of arrival as well as pulse width is about 10-1 μs orders of magnitude. The estimate error, which is affected by the correlation pulse width, decreases with the increase of correlation pulse width, and then increases. The proposed method is especially applicable to the reconnoitering of new system radar such as chirp radar, and Synthetic Aperture Radar (SAR).

Key words: Linear Frequency Modulation (LFM) pulse signal, parameter estimation, fractional Fourier Transform (FrFT), Partial Correlation, Cramer-Rao Low Bound (CRLB)

中图分类号: