GNSS Multiple Interference Mitigation With TF-Unet Method Based on High-Resolution Interference Sensing

作者：Jie Song, Zukun Lu, Xiaoyu Zhao, Wei Xiao, Xiaomei Tang, Guangfu Sun · 发表于：IEEE Sensors Journal · 年份：2025 · DOI：10.1109/jsen.2025.3623113 · 被引用次数：4 · 研究领域：GNSS positioning and interference、Soil Moisture and Remote Sensing、Indoor and Outdoor Localization Technologies

Interference mitigation is crucial for ensuring the reliability of Global Navigation Satellite System (GNSS) in complex electromagnetic environments. However, GNSS receivers frequently encounter various types of interference from radar, communication systems, airfields, and unmanned aerial systems. While existing interference mitigation methods for single-antenna GNSS receivers are only suppress specific types of interference. This paper proposes a multi-interference mitigation method based on the time-frequency Universal Network(TF-Unet) network. The TF-Unet method enables fine-grained segmentation based on high-resolution TF images, sensing and mitigating multidimensional interference features within TF domain. This method employs a signal fusion mechanism that optimizes interference mitigation by minimizing the TF Equivalent Mean Square Error (TFEMSE), thereby enhancing interference edge detection probability and reducing spectral impacts on GNSS signals. Leveraging high-resolution TF analysis, this approach constructs the first GNSS TF signal dataset encompassing multiple interference types and enables phase-aligned GNSS signal reconstruction. Simulations on both conventional and power-enhanced GNSS signals demonstrate that the proposed method effectively suppresses narrow-band interference (NBI), linear frequency sweeping interference (LFSI), and pulse interference (PI) by maximum 10dB·Hz carrier-to-noise ratio(C/N0) gain compared to conventional methods. Practical exper...