Angle Estimation Algorithms For MIMO Radar In Complex Environments

Recently,a novel radar system named Multiple-Input Multiple-Output(MIMO)radar is presented,which has been drawn widely considerable attention since its unique advantages by comparison with the phased radar.MIMO radar employs multiple antennas to simultaneously transmit diverse waveforms and uses multiple antennas to receive the reflected signals.In general,according to the configuration of the transmit array and receive array,MIMO radar can be classified into two types.One is statistical radar and another is collocated radar,including bistatic MIMO radar and monostatic MIMO radar.The angle estimation is a significant problem in MIMO radar.In this dissertation,on the one hand,we focus on the problems of estimation the direction of departure(DOD)and direction of arrival(DOA)in impulsive noise environments and the complexity of angle estimation algorithm in bistatic MIMO radar,respectively.On the other hand,we research on DOA estimation in the situation of lower signal-to noise ratio(SNR)/limited snapshots,closely spaced targets and coherent source in monostatic radar.The main contributions of this dissertation are summarized as follows:1)Angle estimation approach based on fractional lower order moment in bistatic MIMOradarIn this section,we address the problem of the existing MIMO radar angle estimation algorithms based on second-order statistics performance of angle estimation declining considerably in impulsive noise environments.A fractional lower order moment Unitary ESPRIT(FLOM-Unitary ESPRIT)algorithm has been devised for joint estimation the DODs and DOAs of the targets.Firstly,according to the definition of fractional lower order statistic,we constructed the fractional lower order moment matrix.Moreover,we can effectively estimate the DODs and DOAs by exploiting Unitary ESPRIT algorithm in impulsive environments.Finally,Then extensive computer simulation results demonstrate that the proposed algorithm can efficiently suppress impulsive noise,which has high angle accuracy.2)An angle estimation approach based on infinite-norm in bistatic MIMO radarAs mentioned above,we address the angle problem of MIMO radar in impulsive noise environments.However,the angle estimation performance of the FLOM-Unitary ESRPIT algorithm degrades in the severe impulsive noise environments,which cannot efficiently estimate the DODs and DOAs of the targets.To combat the significant problem,we investigate the infinite-norm method combining it and the ESPRIT-MUSIC algorithm,and propose the infinite-norm ESPRIT-MUSIC algorithm(IESPRIT-MSUIC)in the presence of severe impulsive noise environments.The proposed algorithm overcomes the disadvantage of the FLOM-Unitary ESPRIT algorithm and has much better the performance of angle estimation compared with the FLOM-Unitary ESRPIT algorithm in the case of impulsive noise.Moreover,the presented algorithm has work well in Gaussian noise environments.Last but not least,several numerical computer simulation results verify the effectiveness and correctness of the proposed algorithm.3)Angle estimation method based on MSWF in bistatic MIMO radarIn this section,we discuss the principle of the Multi-Stage Wiener Filter(MSWF),combining it and Unitary ESPRIT algorithm,and present a computationally efficient angle estimation algorithm in bistatic MIMO radar.The presented algorithm obtains the comparable angle estimation with Unitary ESPRIT algorithm as well as much better angel estimation performance than ESPRIT algorithm.As it does not need of eigen-decomposition operation with heavy computational complexity,the presented algorithm has lower computational complexity in contrast to the Unitary ESPRIT and ESRPIT algorithms.Finally,the extensive computer simulation experiments demonstrate the effectiveness and correctness of the proposed algorithm.4)Angle estimation for monostatic MIMO radar using a real-valued subspace fittingsparse representation approachIn this section,we discuss the problems of the existing angle estimation algorithms for monostatic radar will encounter their performance degradation in the cases of few snapshots,closely spaced targets,low signal-to-noise ratio(SNR),closely spaced targets,or strongly correlated sources.To address those challenge problems,we investigated the representation method,combining it and subspace fitting algorithm,and presented a new real-valued subspace fitting representation algorithm for angle estimation in monostatic radar.The proposed algorithm has many advantages over the existing algorithms such as low sensitivity to the assumed number of targets,improved separation angular performance,enjoying on coherent source environments without requiring decorrelation,high estimation accuracy and so on.Finally,the several computer simulation results have been demonstrated to support the effectiveness and correctness of the presented algorithm.5)DOA estimation based on l_p-MUSIC in monostatic MIMO radar using the real-valuedand reduced dimensional techniquesIn this section,we consider the DOA estimation problem for monostatic MIMO radar in impulsive noise environments and introduce an unfinished work.In this work,we propose the l_p-MUSIC approach by exploiting the real-valued and reduced dimensional techniques for angle estimation.Finally,some simulation results are demonstrated.