Covariance Estimation of Generalized Monopulse for Radar Tracking

Generalized Monopulse is an asymptotically efficient estimator of the Direction of Arrival (DoA), offering manifold array processing such as anti-jamming through adaptive beamforming, or clutter cancellation through space-time adaptive processing. In addition, this estimator is unbiased even in the general case of arbitrary antenna arrays, making it preferable over classical monopulse techniques. Its performance is well predicted in analytical formulae which are unfortunately inconvenient for real-time applications. Indeed, the unknown covariance matrix of the DoA estimation is usually required to tune the tracking stage, such as Kalman filters, for instance. This paper proposes a simple estimator for the covariance matrix of Generalized Monopulse intended for real time processing.