| Multiple-input multiple-output(MIMO) radar is a new radar system, which utilizes the diversity of transmitted waveforms to improve radar performances. According to the distance between radar antennas, MIMO radar can be classified into two types. One is distributed MIMO radar, which with widely separated antennas. The other is colocated MIMO radar including monostaic and bistatic MIMO radar, which with closely spaced antennas, similar to phased-array radar. Compared to its phased-array counterpart, the colocated MIMO radar can use the waveform diversity to gain more degree-of-freedoms, then improve the performance of parameters estimation. This dissertation mainly discusses some issues in colocated MIMO radar signal processing in practical application, including array calibration and robust parameter estimation in the complicated environment, and fast transmit beampattern design. Robust beamforming is also studied for high-speed moving radar in this dissertation.The main contribution of this dissertation is as follows.1. A method of calibrating the sensor position error and gain-phase error for monostatic MIMO radar is proposed. This method utilized the received signals which are spatially and temporally disjoint. By emitting orthogonal waveforms and using a matched-filter bank in the receivers, the orthogonal waveform components can be separated. Then sub-matrix corresponding to transmitting/receiving array can be extracted to estimate sensor errors, avoiding parameters coupling. The calibration of transmitting/receiving array can be done off-line, which can be calculated efficiently and easy to implement.2. The mutual coupling effects and gain-phase errors can drastically degrade the direction estimation performance of monostatic MIMO radar array. Based on the Capon beamforming algorithm, a new auto-calibration method is proposed. The Toeplitz structure of the mutual coupling matrix for uniform linear array is exploited to reduce the number of unknown parameters. Using the property of Kronecker product, the transmitting and receiving array uncertainties can be decoupled for the error compensation at each iteration. Finally, the estimation of the target direction can be obtained.3. In most cases, the effect of different kinds of sensor errors present together can be described as direction-dependent gain-phase errors. A method based on reduced-dimension MUSIC for target direction estimation in bistatic MIMO radar is developed. Using the instrumental sensors and the orthogonality between signal subspace and noise subspace, the array errors and the target direction can be decoupled. Then the direction estimation can be obtained by double one-dimensional searches without the knowledge of sensor errors. The computational cost can be reduced efficiently. Simulation results show that the proposed method can work well, when the sensor position errors, gain-phase errors and mutual coupling effects are present together in both transmiting and receiving arrays of bistatic MIMO radar.4. When the sources are distributed in space, the mathematical constraint of point source is not satisfied. For bistatic MIMO radar with gain-phase errors, a robust angle estimation algorithm for coherent distributed targets is proposed. Firstly, the signal model of coherently distributed targets is established. Then, based on the second rotational invariance property of the steering vectors of coherently distributed targets, the central angles of targets can be estimated. Simulation results verify the validity of the proposed algorithm, and show that it is not sensitive to the form of source distribution functions.5. Transmit beampattern design for MIMO radar is obtained by synthesizing the signal covariance matrix, which can be achieved by convex optimization approaches. Due to the high computational complexity, these approaches are not easy for practical implementation. To reduce the computation load, a cyclic iterative method for MIMO radar transmit beampattern design is proposed. Based on the weighted least square criterion, signal covariance matrix can be obtained by optimizing the quadratic cost function with respect to its Hermitian square root. For the uniform linear array, especially when the sampling grid in normalized spatial frequency is uniform and the weights for grid points are the same, FFT can be used to further decrease the algorithm’s complexity.6. For the array mounted on high-speed platform, the performance of conventional adaptive beamforming algorithm can be seriously degraded, due to the presence of fast changing of interference direction and look direction error. To combat the impact, a robust beamforming method based on interference location prediction is proposed. First, we define the normalized spatial frequency and predict the range of the interference location. Then, a robust adaptive beamformer is derived by a matrix weighted approach, which can suppress the gain in the predicted region and maintain the gain in the mainlobe region. Therefore, the proposed method is more robust to interferences and look direction errors. Simulation results demonstrate the effectiveness of the proposed method.
|
PDF Full Text Request