Study On Moving Targets Detection And Estimation For Multi-channel Airborne Radar

For its flexibility, mobility and broad vision, the airborne radar plays an important role in the contention of air supremacy in modern war, and moving target detection and localization is an important mission of airborne radar. The airborne radar usually faces widely distributed ground clutter and active jamming since it usually works in down-looking mode. The space time adaptive processing (STAP) is an effective method for the suppression of the ground clutter and active jamming. Due to the large number of IID samples and the high computational complexity required by the optimal STAP method, the optimal STAP method can not be satisfied in practical radar environment, the rank reduced STAP method and the non-homogenous STAP method are the focus and emphasis of the researchers in recent years, and many achievements have been obtained. At present, the STAP method urgently needs to combine with the new radar system and technology to achieve new development. After detection, it also needs to localize, image, and identify the target. In addition, angle estimation is an important part of localization. In practical radar environment, coherent or highly correlated sources are often encountered (such as multi-path propagation), its angle estimation has important meaning for determining the real targets and decreasing the false alarm probability. In the conventional angle estimation methods for coherent sources, the dimension-reduced algorithms are simple in implementation, but its aperture loss may lead to the decrease of the estimation accuracy. Though the non dimension-reduced methods has no aperture loss, they are complicated in realization and are often aimed at specific environments, so the practical value of non dimension-reduced methods are inferior to that of the dimension-reduced algorithms. Currently, combining with the new radar system and technology to find more effective angle estimation method of coherent sources is also the research emphasis. As a new radar system, MIMO radar has the advantage of larger degree of freedom (DOF) in spatial domain, which has important meaning in the performance improvement of clutter suppression and angle estimation.This thesis closely focuses on the practical problems in moving target detection and localization of multi-channel airborne radar, and investigates the rank reduced STAP method in phased array radar, the rank reduced and non-homogenous STAP method in MIMO radar and the angle estimation method for coherent sources in MIMO radar. The main contents of this thesis can be summarized as follows:1. The basic principle and the performance evaluation standard of STAP method are given, and the power spectrum distribution of the clutter of collocated MIMO radar is also discussed, which provides theoretical foundation for the subsequent research.2. In available post-Doppler processing STAP method, the interested and the secondary Doppler channels are orthogonal without weighting, even if the weighting window is added, their correlation is still not high. According to this fact, a Doppler channel distance reduced post-Doppler STAP method is proposed, in which the correlation of adjacent Doppler channels is improved, thus the clutter in the interested channel can be cancelled by not only the sidelobe clutter but also some of the mainlobe clutter of the auxiliary channels, and consequently better clutter suppression performance can be obtained. Meanwhile, because the steering vector of the target is strictly constrained, the reduction of the Doppler channel distance will not result in target cancellation.3. Firstly, considering that the spatial domain DOF of the collocated MIMO radar is large, and the typical rank reduced STAP method-mDT method can not be used directly, a subarray domain mDT method is presented. By using the heterogeneous subarray partition to the spatial aperture of the MIMO radar, this method can effectively reduce the spatial domain DOF of MIMO radar without grating lobe. Secondly, regarding to the severely power nonhomogenous environment in which the isolated interference occupies little proportion, a time-domain smoothing based two stage cascaded method is presented, to suppress the clutter and the isolated interference successively. Compared with the DDD method, this method has advantages of making good use of the statistical information, low spatial-- temporal aperture loss and robustness to array errors.4. The existing angle estimation methods in collocated MIMO radar are mainly for the independent sources, so in this section a data matrix reconstruction based angle estimation method for coherent sources is proposed. This method can realize completely decorrelation, fully utilize the large spatial DOF of MIMO radar and can also be used to colored spatial noise circumstance.5. In view of the problem that the existing angle estimation method for coherent sources in bistatic MIMO radar only uses two transmit elements and can not estimate the multi-path direction of departure (DOD), a data matrix reconstruction-based joint direction of arrival (DOA) and DOD estimation and pairing method is proposed. Itthe proposed method can use many transmit elements, and can estimate both the multipath DOD and the multi-path DOA. Meanwhile, it has good estimation performance and low computational complexity for pairing.