Enhancing UAV Surveillance in VLRC Systems: A Unitary CP Factorization Approach for OFDM-Based Passive Radar

The integration of unmanned aerial vehicles (UAVs) and Vehicle Low-altitude Road Cloud (VLRC) systems has greatly enhanced traffic efficiency and urban safety. However, their widespread use has raised airspace safety issues, especially for "low-slow-small" (LSS) targets, which are difficult to distinguish from birds due to their similar radar signals. In this paper, we study a joint estimation of delay-Doppler-Direction of arrival (DOA) parameters for passive radar using orthogonal frequency division multiplexing (OFDM) signals. A tensor decomposition method enhances super-resolution and parameter estimation through a unitary transform, the forward-backward technique, CANDECOMP/PARAFAC (CP) decomposition, and the EOMP algorithm. Firstly, the unitary transformation and forward-backward process are exploited to construct the real-valued signal model. Then, the CP decomposition method is used to derive the delay, Doppler, and Direction of Arrival (DOA) factor matrices. Finally, the EOMP algorithm is used to extract the parameters for target estimation from these factor matrices. The simulation results have demonstrated its effectiveness in LSS target parameter estimation, supporting safer and more efficient VLRC collaboration.