Signal Detection And Parameter Estimation Algorithms For MIMO Radar

MIMO radar is becoming popular in radar field. It has many channels between transmitting and receiving array due to the orthogonal transmitted waveforms. MIMO radar provides a novel technique for increasing the degrees of freedom of the radar system and improving the performance of detection and parameter estimation. This dissertation mainly focuses on the signal detection and parameter estimation of MIMO radar. The research of this dissertation is summarized as follows:1. Ambiguity function of MIMO radar is established based on the signal model, and the performance is analyzed. The ambiguity function is a combined function of range, Doppler frequency and angle, and demontrates the resolution of MIMO radar in range, Doppler frequency and angle. For the waveform diversity MIMO radar, there are no couples among range, Doppler frequency and angle. The range-Doppler resolution is determined by the sum of the resolution of all the transmitted waveforms, and the angle resolution is determined by the resolutions of the transmitting array and the receiving array.2. The number of virtual sensors of MIMO radar with different linear array geomitries are analyzed and compared. A new array optimization algorithm for MIMO radar is proposed. First, the minimum number of physical sensors is used to get virtual sensors as many as possible, so the efficiency of the physical sensors is maximum, and then the virtual array of MIMO radar is subject to a minimum redundancy array, so the resolution of the MIMO radar system is improved and the high side-lobe levels are reduced. With the same number of physical sensors, the MIMO radar optimized in this dissertation can improve the ratio of mainlobe to sidelobe of the beampattern than that of the ULA MIMO radar and minimum redundancy (MR) MIMO radar.3. Adaptive pulse compression (APC) techniques of MIMO radar are researched. Based on minimum mean-square error criterion, an efficient approach for adaptive pulse compression of MIMO radar is presented, in which weight coefficients of the filter for each individual range cell are adaptively estimated from the received signal. Compared with the standard mached filter, the proposed method can suppress range sidelobe level of a transmitted waveform, and lowed the correlated level between different transmitted waveforms. To lower the computational complexity, APC of MIMO radar based on multi-stage wiener filter (MSWF) is proposed, in which, the m-th transmitted waveform is considers as the desired signal and the other M-1 transmitted waveforms are considered as inteferences, the principle of adaptive beamforming is applied to the APC of MIMO radar. The weights of forward and backward recursion of MSWF are computed to obtain the weight coefficients of the APC. The proposed method does not involve the estimate and the inverse of the sample covariance matrix, thus indicating much lower computation complexity. The proposed method is superior to the conventional matched filter, and performs similarly as the APC based on MMSE and GSC.4. Based on the receiving signal model of MIMO radar, several conventional DOA estimation techniques is applied to MIMO radar. Maximum likelihood DOA estimation of MIMO radar is derived, and the estimation consistency is proved and the estimation variance is derived theoretically. Theoretical analysis shows that the variance of MIMO radar is lower than that of the conventional phased array radar and the estimation performance of MIMO radar is better than that of phased array radar. The estimation variance with different SNRs, snapshots and receiving array elements are simulated, the results are consistent with the theoretical analysis. Last, a virtual spatial smoothing algorithm of MIMO radar is presented using the diversity of the transmitted waveform and the independence of the transmitting-receiving channel. In the technique, the received signal of the virtual sub-arrays are extracted from the output of the matched filter banks of MIMO radar, and then the auto-correlation matrix of the virtual sub-array is derived. The algorithm in this dissertation can improve the precision of DOA estimation compared with that of the conventional phased array spatial smoothing (SS) technique and MIMO virtual array SS technique.5. A novel signal subspace reconstruction method for multiple-input multiple-output (MIMO) radar is proposed. The developed method realizes signal subspace resconstruction by the eigendecomposition of two low dimension matrixes. It provides lower computational complexity and fewer training samples compared to the full DOF method. When the number of training samples is small, the DOA estimation precision is improved than the full degree of freedom algorithm.