Low‐Dimensional Robust Mismatched Filter Design for Passive Bistatic Radar Using 5G NTN Downlink Signals

ABSTRACT

Ambiguity sidelobes caused by deterministic frame structures in 5G non‐terrestrial network (NTN) downlink signals may generate spurious peaks and mask weak targets in passive bistatic radar (PBR) sensing. To address this issue, this letter proposes a low‐dimensional robust mismatched filtering (LD‐RMMF) method. By exploiting the OFDM symbol organisation, a symbol level filter model reduces the optimisation variables from the sample domain to the symbol domain. To improve robustness under low reference signal to noise ratio (SNR) conditions, a bounded reference signal uncertainty model is introduced into the mainlobe preservation constraint using a worst case formulation. The resulting problem is reformulated as a convex optimisation problem. Simulation results show that the proposed method suppresses the structured ambiguity sidelobes of 5G NTN signals while producing a much lower mainlobe SNR loss (SNRL) than conventional MMF.