Journal of Systems Engineering and Electronics ›› 2026, Vol. 37 ›› Issue (4): 1120-1137.doi: 10.23919/JSEE.2026.000123

• RADAR PERFORMANCE IMPROVEMENT USING BEYOND LINEAR SIGNAL PROCESSING • Previous Articles    

Mutual interference mitigation for automotive FMCW radar based on TRPCA

Ken Chen1,2(), Xinyu Guan1(), Hui Zhang1(), Ruizhe Zhang1(), Yan Huang1,3,*(), Xiaowei Zhu1(), Wei Hong1,3()   

  1. 1State Key Laboratory of Millimeter Waves, Southeast University, Nanjing 210096, China
    2Sichuan Digital Transportation Technology Co., Ltd., Chengdu 610095, China
    3Purple Mountain Laboratories, Nanjing 210096, China
  • Received:2025-01-25 Accepted:2025-06-30 Online:2026-08-18 Published:2026-09-03
  • Contact: Yan Huang E-mail:chenken@foxmail.com;xinyu_guan@seu.edu.cn;huizhang@seu.edu.cn;220200965@seu.edu.cn;yellowstone0636@hotmail.com;xwzhu@seu.edu.cn;weihong@seu.edu.cn
  • Supported by:
    This work was supported by the National Natural Science Foundation of China (62271142; U2341206), Nanjing U35 Strong Foundation Engineering (4204002501), Key Research and Development Program of Jiangsu Province (BE2023021; BE2023011-2), Young Scholar Funding of Southeast University, and Fundamental Research Funds for the Central Universities (2242022k60008).

Abstract:

The dramatic rise in vehicle numbers and the development of autonomous driving technology have underscored the importance of accurate environment perception to prevent traffic accidents. Due to its distinctive advantages, millimeter-wave radar has emerged as the dominant type of vehicle-mounted sensor. However, frequency band for automotive radar is restricted to 76−81 GHz, resulting in insufficient spectrum resources because of the high-resolution requirement. The rapid increase of radars in the actual traffic environment will inevitably lead to mutual interference between radars, posing a potential threat to public transportation. To address this problem, this paper investigates the mutual interference of automotive frequency modulated continuous wave (FMCW) radars, and mitigates the mutual interference utilizing a low-rank tensor model. By stacking the short-time Fourier transform (STFT) matrix of all chirps in a single channel as tensors, a tensor robust principal component analysis (TRPCA) algorithm is employed to separate the target signal and the interference. Mutual interference of multiple input multiple output (MIMO) radar could be suppressed by applying TRPCA in all channels. Both simulations and numerical experiments demonstrate that the proposed method achieves higher signal to interference plus noise ratio (SINR) and lower root mean square error (RMSE) compared with other methods, thereby proving its superior interference suppression capability in different practical scenarios.

Key words: automotive frequency modulated continuous wave (FMCW) radar, mutual interference, interference mitigation, tensor robust principal component analysis (TRPCA)