Kernel Design for Combined Time–Frequency Distribution in GNSS Interference Detection

This paper proposes a novel interference detection method for Global Navigation Satellite System (GNSS) receivers based on a kernel-designed reduced interference distribution (RID) and a combined time–frequency (TF) representation. The proposed TF distribution is built via the Hadamard product of the Wigner–Ville distribution (WVD) and the RID (HP-WVD-RID), enabling effective suppression of cross-terms while preserving strong TF energy concentration. Compared with existed transformed-domain analysis techniques, the proposed method provides improved detection performance under severe interference conditions. Experimental results using interfered GPS-L1 C/A signals demonstrate a 47.19%improvement in interference frequency estimation accuracy over the smoothed pseudo Wigner–Ville distribution (SPWVD), and a 91.94% increase in zero transformed-domain energy percentage (ZTEP) compared with the fractional Fourier transform (FrFT), confirming its effectiveness for GNSS interference detection under challenging conditions.