Research On Design And Optimization Of New Radar Systems

The new scheme of radar can adapt to the increasingly complex electromagnetic environ-ment, and becomes the research hotspot in the radar field in recent years. Since the system design and optimization can improve the specific performance of radar system, the study of dif-ferent system design and optimization methods have also been widespread concern. This thesis focuses on the new radar systems including cognitive radar (CR), multiple-input and multiple-output (MIMO) radar and compressed sensing (CS) radar, and proposes a variety of optimization methods to effectively improve the estimation, detection and localization performance for the specific target. The main contents of this thesis include the following five aspects:1. In the CR system, a method based on Kalman filter (KF) to estimate the target scatter-ing coefficients (TSC) is presented for the temporally correlated extended target. Then,a direct algorithm is proposed to optimize the transmitted waveform by minimizing the mean square error in each KF iteration. Additionally, the optimization algorithm also considers the constraints including transmit power, peak-to-average power ratio, and de-tection probability with constant false alarm ratio. However, the waveform optimization problem is non-convex and cannot be solved effectively. Therefore, this thesis proposes a two-step method to convert it into several convex optimization problems. Finally, the optimized waveform for the temporally correlated extended target is obtained, and the better TSC estimation performance is also achieved.2. For multiple extended targets, the joint TSC estimation methods are proposed in the sce-narios with close and distantly separated targets,respectively. Then,this thesis proposes a joint waveform optimization algorithm to improve the joint TSC estimation performance,and achieve a trade-off among the different targets. Moreover, the joint waveform opti-mization problem is also non-convex, and can be converted into a semidefinite program-ming (SDP) problem by the relaxation operations. Then, the optimized waveform for multiple targets can be obtained by solving the SDP problem, and can significantly im-prove the joint TSC estimation performance.3. To detect the moving targets,a radar system with multiple moving platforms is proposed,where each distributed transmitter and receiver platforms are equipped with the colocated MIMO radar. Therefore, this radar system have advantages of both distributed and colo-cated MIMO radar systems, and can effectively improve the detection performance for moving targets. First,for the problem of heterogeneous clutter, a clutter sparse model is proposed to estimate clutter information by exploiting the sparse characteristics of clut-ter. Since the model needs to discretize the clutter, and may cause the clutter scattering points to deviate from the discrete grids. Therefore, an off-grid problem of clutter estima-tion is established, and a two-step method is proposed to solve the problem and estimate the heterogeneous clutter. Second, using the estimated clutter and target information, a waveform optimization algorithm is proposed to maximize the signal-to-clutter-and-noise ratio of the echo signals. Finally, the moving target detection method based on generalized likelihood ratio is realized in the fusion center for the multiple moving platforms. Simu-lation results show that the radar design scheme and the waveform optimization method can significantly improve the detection performance of moving targets.4. For the velocity and distance estimation problem with multiple extended targets, this the-sis models the parameter estimation problem as a sparse reconstruction problem, and an estimation method based on CS for the extended target is proposed. Since the echo sig-nals from the extended targets are the convolutions between the transmitted signals and the target impulse responses, this these proposes a method of constructing the dictio-nary matrix for multiple extended targets. Additionally, to further improve the estimation performance of TSC, a waveform optimization method is proposed to improve the recon-struction performance by minimizing the mutual coherence of the dictionary matrix, and to improve the corresponding the estimation performance of TSC.5. To localize multiple targets in the distributed MIMO radar, a method of optimizing the locations of radar antennas is proposed. First, a localization method based on sparse re-construction is proposed after formulating the overcomplete dictionary matrix with all the location information. Then, the coherence coefficient of the matrix is adopted to in-directly describe the localization performance. Furthermore, an algorithm for optimizing the antenna positions is proposed to minimize the mutual coherence, and improve the per-formance of both sparse reconstruction and the localization accuracy in the radar system.To describe the localization performance with different antenna positions, the probability distribution of the mutual coherence is deduced for different antenna positions, and can be used to improve the localization performance by optimizing the antenna positions in the distributed MIMO radar system.