Low angle estimation in MIMO radar based on unitary ESPRIT under spatial smoothing

Abstract

Aiming at the problem that the overlapping of multipath signals seriously affects the radar's performance of the elevation angle estimation for low-altitude targets, the authors combine the unitary ESPRIT (UESPRIT) and the multi-input and multi-output (MIMO) radar system and propose an algorithm based on UESPRIT to estimate the direction of arrival for low-altitude targets. Firstly, the virtual matrix after generalised matched filtering of MIMO radar multipath received signals is vectorised. Secondly, for the coherence of direct and reflected wave signals, which cannot be directly processed by the UESPRIT algorithm, the signal preprocessing is performed by spatial smoothing of the sampled data matrices. Finally, the low-altitude target estimation is carried out by using the UESPRIT algorithm. The Cramer–Rao bound (CRB) for arbitrary unbiased estimation of angle estimation is derived. The relationship between the estimation performance of the algorithm and the signal-to-noise ratio, the number of snapshots and the number of elements is analysed by simulation and compared with CRB. The simulation results show that the algorithm can still effectively estimate the elevation angle of low-altitude targets under the mutual weakening of direct and multipath reflection signals, and has better performance for low-altitude targets than generalised MUSIC.