Adaptive Detection For Multichannel Radar Signals

Signal Detection is one of the primary problems in the field of signal processing. Early theories about the signal detection are mainly for single channel data. With the development of science and technology, the data to be processed are usually multichannel, such as the data received by the radar when the phased array is utilized. Moreover, the multichannel data are often complex-valued. In addition, the noise environments encountered in radar systems are colored due to clutter and jamming. For the multichannel radar signal detection in unknown colored noise, in this dissertation we aim to design effective adaptive detection approaches. The main contributions of this paper are listed below:In the second chapter, we propose the Rao and Wald criteria for the multichannel complex-valued signal detection, as well as the Cramér-Rao Bound(CRB) for the complex-valued parameter estimation. Compared with the existing method, the proposed one does not need to cascade the real and imaginary parts of the complex-valued data. Instead, it directly processes the data in the complex number field. Consequently, it reduces the design complexity and computational burden.In the third chapter, we study the point-like target detection in the presence of signal mismatch. According to Wald and generalized likelihood ratio test(GLRT) criteria, we propose selective detectors by adding fictitious random or deterministic interference. By comparing the similarity of the existing detectors, we propose single-parameter tunable detectors and two-parameter tunable detectors. By adjusting the tunable parameters, the tunable detectors can flexibly control the mismatched signal. Moreover, we analyze the statistical properties of all the proposed detectors, and derive the closed-form expressions for the probabilities of detection(PDs) and false alarm(PFAs).In the fourth chapter, we consider the point-like detection in the presence of interference, and three types of interferences are investigated.(1) For the known interference, we propose the detectors according to the Rao and Wald criteria. Moreover, we introduce a novel detection scheme, namely, interference cancellation before detection(ICBD). A remarkable property of the ICBD scheme is that it is much easier to design a detector, which can work even when the number of the secondary data is less than the dimension of the primary data.(2) For the constrained interference, we propose the detectors according to the GLRT, Rao, and Wald criteria, and analyze their statistical properties.(3) For the completely unknown interference, we propose two detectors according to the GLRT and Wald criteria. The two proposed detectors have improved detection performance, compared with the existing ones.In the fifth chapter, we consider the distributed target detection in the presence of signal mismatch. By adding fictitious deterministic interference, we propose two selective detectors. Moreover, we introduce a novel tunable detector, which can achieve mismatched signal detection or rejection for distributed targets.In the sixth chapter, we consider the detection problem for the double-subspace signal(DOSS), for which the rows and columns of the matrix-valued signal both lie in certain subspaces. The DOSS model generalizes the existing signal model for the distributed target. According to the GLRT, Rao, and Wald criteria, as well as the spectral norm test(SNT), we propose many effective detectors and give the interesting special cases of proposed detectors. Finally, we show the difference between the detection performance of the detectors under different parameter settings.In the seventh chapter, we investigate the detection problem for three kinds of direction uncertainty.(1) For the distributed target of completely unknown direction, we propose two effective detectors according to the Rao and Wald criteria.(2) When the echoes all come from the same direction, and the corresponding signal steering vector belongs to a known subspace, the detection problem for the distributed target is referred to as the problem of direction detection. According to the Wald criterion, we propose an effective direction detector.(3) When the spatial and temporal steering vectors both lie in certain known subspaces, we propose four generalized direction detectors(GDDs) according to the GLRT and Wald criteria, as well as their two-step variations. The special cases of each GDD are also given.In the eighth chapter, we consider the space-time adaptive detection(STAD) for airborne radar in low sample support. According to the diagonal loading, principal component analysis, and Krylov subspace technique, we propose many kinds of effective detectors when the number of the training data is insufficient. We also analyze the asymptotical statistical distributions and derive the corresponding PDs and PFAs.In the ninth chapter, we consider the detection problem for the multiple-input multiple-output(MIMO) radar with colocated antennas. According to the Rao and Wald criteria, we propose two adaptive detectors, and analyze the statistical performance. Moreover, we compare them with the phased-array counterpart.