| 摘要: |
| 全球定位系统(GPS)窄带干扰抑制常用的方法是用频域变换估计干扰频带，然后用频率内插或外推滤波削弱干扰信号的影响。为缓解在干扰频带探测阶段较高的信号采样率和频域变换所带来的高计算量的压力，提出了一种新的利用多层级原子库进行窄带干扰频带估计的方法。分别研究了多种GPS窄带干扰在傅里叶基、原子库稀疏表示基下用压缩感知重构进行干扰频率估计的方法。用Monte Carlo仿真分析了不同的信噪比、搜索步长和测量值数目条件对压缩感知构造过完备原子库方法进行窄带信号频带估计效果的影响，验证了文中提出方法的有效性。理论分析表明：所提方法与传统频域变换方法相比，具有能够极大减少测量值与计算量的优点。 |
| 关键词: 卫星导航 窄带干扰 压缩感知 频率估计 原子字典 |
| DOI： |
|
| 基金项目:国家自然科学基金资助项目（61302073，61471021） |
|
| Multiple category estimation for GPS narrow-band interference based on compressed sensing |
| YAO Yanxin |
| () |
| Abstract: |
| The frequency domain transformation for detecting interference frequency and the frequency interpolation or extended extrapolation algorithm for suppressing the interference are used in conventional Global Positioning System（GPS）narrow-band interference suppression methods. In order to relieve the pressure of high sampling rate and computation load caused by frequency domain transformation in the narrow band interference frequency detection,a novel narrow-band inference frequency detection method is proposed utilizing multiple category atom dictionary construction compressed sensing. The methods for frequency detection based on compressed sensing are studied in the spare representation styles of Fourier transformation basis and atom dictionary,respectively. For the atom dictionary compressed sensing method,the effect of narrow-band interference frequency detection versus various signal-to-noise ratios(SNRs),various search resolutions and various numbers of measurements are studied using the Monte Carlo simulation methods. The correctness of the approach proposed has been validated by simulations,and the conclusion is derived from theoretical analysis that the proposed approach reduces sharply the measurement and computation cost,offering theoretical basis for the realistic interference application. |
| Key words: global navigation satellite system narrow-band interference compressed sensing frequency detection atom dictionary |
《电讯技术》编辑部 