Study On MIMO Radar Waveform Optimization With Imperfect Target Prior Knowledge

In recent years, multiple-input multiple-output (MIMO) radar is attracting the attention of researchers and practitioners alike owing to its good performances. Since MIMO radar can improve the performance of target detection and direction of arrival (DOA), and detect slowly moving targets, it has been widely used in military and civil regions. In MIMO radar system, the waveforms are usually designed according to the prior knowledge of targets and environments, but the prior knowledge is obtained by estimation, during which errors are unavoidable, so the performance of conventional waveform degrades seriously. This dissertation is mainly concerned with the research on MIMO radar waveform optimization with imperfect target prior knowledge. The author’s major contributions are outlined as follows:1. In multiple-input multiple-output (MIMO) systems, the waveform optimization is usually focused on certain one other than the whole performance. In this paper, a novel waveform design method is proposed to improve MIMO radar’s overall performance, such as:target detection and parameter estimation. In the proposed method, the waveform covariance matrix (WCM) is optimized under the three following constraints: improving target detection probability, lower parameter estimation variance and suppressing sidelobe. Firstly, the equivalent formulas of detection probability and Cramer-Rao bound (CRB) are derived. After that, with maximizing the difference between the sidelobe and mainlobe as well as weighting each constraint, the waveform optimization problem, the constraints of which can be adjusted flexibly to satisfy the practical demands, is formulated. This waveform optimization problem can be recast as a linear programming, so it can be solved efficiently. The simulation results verify the effectiveness of the proposed design.2. In MIMO radar system, the waveforms are usually designed based on the prior knowledge of targets and environments. And the prior knowledge is obtained by estimation, during which errors are unavoidable, so the performance of conventional waveform design degrades seriously. For solving this problem, this paper presents two robust waveform design methods to improve detection and parameter estimation performance of MIMO radar separately. With target location errors and channel errors bounded, construct two optimization problems with signal to noise ratio (SNR) and CRB as cost function separately. To maximize SNR and minimize CRB, iterative algorithms are provided, and each step of the algorithms is transformed into a convex problem, which can be solved efficiently. Simulation results show that the proposed methods can improve the detection performance and parameters estimation performance of MIMO radar.3. Conventional waveform design methods of MIMO radar are sensitive to transport matrix errors, so the optimal matched-waveform is hard to achieve, and the detection performance degrades dramatically. To mitigate this problem, a novel robust waveform design method is introduced for MIMO radar based on probabilistic constraint. In this method, the probability of the worst case is considered very small. Therefore, the probability of output SNR less than the acceptable level is constrained no more than the outage probability, the waveforms are designed to maximum the output SNR. Using the characters of statistical distribution of transport matrix errors, the probabilistic constraint is transformed to a deterministic convex constraint. So the statistical optimization problem is converted to a convex optimization problem. This method maximizes the performance with high probability under transport matrix error. The simulation results show that the method increases the output SNR and detection performance.4. The combination of conventional uniform circular array (UCA) and MIMO radar leads to a flexible technique which enjoys the advantages of MIMO radar without sacrificing the main advantage of UCA radar. However, the UCA MIMO radar is limited to range-independent directivity; this limits the radar performance to mitigate non-desirable range-dependent interferences. In this paper, we propose a new UCA MIMO radar with frequency diversity (FD) for range-dependent beamforming. The essence of the proposed method is each element of the UCA transmits a distinct waveform with a small frequency increment, so the summation of signals in far-field has relationship with distance, while the traditional UCA radar does’t have this relationship. Thus the beamforming is range-dependent. The effectiveness of the proposed technique is verified by numerical simulation results.