Compressed-Sensing-Based Virtual Array Reconstruction for Multi-Node OFDM Radar

This paper presents a technique for generating a wide-aperture virtual uniform linear array (ULA) using cooperative orthogonal frequency division multiplexing (OFDM) multiple input multiple output (MIMO) radar nodes. A generalized phase-reconstruction approach enabled consistent alignment of bistatic and monostatic measurements, while compressed sensing completed the sparse array to enable high-resolution direction of arrival (DoA) estimation. Simulation studies confirmed notable performance gains, including reduced sidelobe levels and a 50% lower root-mean-square localization error compared with individual node processing. The proposed architecture shows promise for scalable and computationally tractable implementation, suggesting a viable path toward real-time, high-accuracy sensing in distributed automotive radar networks.