| 引用本文: | 肖垂欣，石长安，郑春弟，王爱国. 基于变步长双曲正切LMS的汽车雷达干扰抑制方法[J]. 雷达科学与技术, 2025, 23(2): 142-149.[点击复制] |
| XIAO Chuixin, SHI Changan, ZHENG Chundi, WANG Aiguo. Automotive Radar Interference Mitigation Based on Variable Step Size Hyperbolic Tangent LMS[J]. Radar Science and Technology, 2025, 23(2): 142-149.[点击复制] |
|
| 摘要: |
| 针对车载毫米波调频连续波（Frequency Modulated Continuous Wave, FMCW）雷达大量应用导致的雷达间严重的相互干扰问题，利用干扰信号类似脉冲特性，提出了一种变步长双曲正切最小均方自适应算法（Variable Step Hyperbolic Tangent Least Mean Square, VSHTLMS），有效缓解了雷达相互干扰所引发的噪底抬升问题。首先通过引入归一化误差，新设计了一种基于双曲正切函数框架的代价函数，提升了算法对干扰信号的抑制能力，然后运用归一化误差的Logistic函数来实现步长的自动更新。仿真与实测数据表明，与几种现有雷达干扰抑制方法相比，本文提出的算法能够有效提高信干比（Signal?to?Interference Ratio, SIR），在FMCW雷达抗干扰方面有着优越的性能。 |
| 关键词: 汽车雷达 雷达干扰 变步长 干扰抑制 自适应降噪 |
| DOI：DOI:10.3969/j.issn.1672-2337.2025.02.004 |
| 分类号:TN911；TN972 |
| 基金项目:国家自然科学基金（No.61972092， 62176082） |
|
| Automotive Radar Interference Mitigation Based on Variable Step Size Hyperbolic Tangent LMS |
|
XIAO Chuixin, SHI Changan, ZHENG Chundi, WANG Aiguo
|
|
1. School of Electronic and Information Engineering, Foshan University, Foshan 528225, China;2. Unit 63891 of PLA, Luoyang 471000, China;3. Institute of Science & Technology Innovation, Dongguan University of Technology, Dongguan 523808, China;4. School of Computer Science and Artificial Intelligence, Foshan University, Foshan 528225, China
|
| Abstract: |
| The large number of millimeter?wave frequency modulated continuous wave radars on the automotive system beget severe mutual interference among these radars. An interference mitigation method based on a novel hyperbolic tangent least mean square with variable step size is proposed to handle the mutual interference, which exploits the impulse?like characteristics of the interfering signal. As a result, the increment of noise level caused by interference is effectively alleviated. The novel cost function that embeds the normalization error employs the hyperbolic tangent function, which improves the interference mitigation ability and the robustness. Moreover, the logistic function of the normali?zation error is used to realize the automatic updating of the step size. The results from the numerical simulation and the field data demonstrate that, compared with several existing methods, this proposed method can offer competitive signal?to?interference ratio, which shows the superiority of the proposed method in terms of anti?interference performance. |
| Key words: automotive radar radar interference variable step size interference mitigation adaptive noise reduction |
