Direction‐of‐arrival estimation using estimator banks in low‐angle tracking for S‐band radar

Abstract

Tracking low‐angle targets over an uneven surface are challenging because of the highly correlated, complicated, and volatile multipath signals encountered in radar. Especially, in the context of irregular reflector modulated by rough sea, which results in performance depravation of the existing direction‐of‐arrival (DOA) estimation methods. An effective DOA estimation approach is based on maximum likelihood (ML) estimator which is referred to as optimal synthetic vector maximum likelihood (OSVML) method. The approach is essentially different, multipath signals are present that without known prior information of ideal model. The optimal projection space in estimator banks is spanned with direct subspace and indirect subspace to match the signals received by S‐band radar. Then, the performance of the proposed method is evaluated by simulations in terms of the signal‐to‐noise ratio (SNR) and snapshot. Finally, field data sets acquired from S‐band radar are carried out to verify the practicability of the proposed method in a naval environment.

Elevation angle measurement for low‐angle targets using synthesized pattern null method

“…As a subcategory of array super‐resolution, maximum likelihood (ML) method, compared with other array super‐resolution methods, has the most significant effect in reducing the elevation angle error of low‐angle targets and has become a research focus in related fields. The standard ML method is too computationally complex to be used in practical applications, so several deformation methods aiming at reducing the computational complexity were proposed 11–13 . Among them, the synthesized vector maximum likelihood (SVML), which is optimized in terms of computational complexity, shows excellent angle measurement performance and has been used in some 3D search radars with digital array system 12,13 .…”

“…The standard ML method is too computationally complex to be used in practical applications, so several deformation methods aiming at reducing the computational complexity were proposed. [11][12][13] Among them, the synthesized vector maximum likelihood (SVML), which is optimized in terms of computational complexity, shows excellent angle measurement performance and has been used in some 3D search radars with digital array system. 12,13 Although the computational complexity of SVML has been reduced, computational overhead still reaches F s × J × (N 3 + 2N 2 + N) × M real multiplications per second and F s × J × (N 3 + 2N) × M real additions per second (N refers to the number of array elements, F s is the baseband sampling rate, M is the number of targets processed simultaneously, and J is the number of searches, that is, the ratio of the estimated range of elevation angle and the estimated interval of elevation angle, usually the range is taken as 0°-4°, and the interval is taken as 0.1°, then J = 40).…”