Research On Radar Adaptive Waveform Optimization

The radar transmitted waveform influences the performance of radar systems. Optimizing adaptive waveforms according to the operating environment is favorable for extracting more information from the echoes, which could improve the performance of radar systems. In additional, signal processing technology in radar receiver have been mature, however, the technology of the transmit adaptivity has not mature, it has a huge space for research and is one of the hot research topics.The main work of this thesis focuses on the transmitted adaptive waveform optimization, including design and selection. The main contributions are as follows:In order to improve the performance of target detection, an algorithm of waveform design based on relative entropy under energy and constant modulus constraints is proposed in clutter environment. Since the relative entropy is inversely proportional to the error rate, then maximizing the relative entropy is as the criterion to optimize the waveform. Finally, the optimization model for waveform is created considering the energy and constant modulus constraints. Then the optimization model is extended for target recognition. The proposed algorithm could improve the performance of the target detection and recognition. The optimized waveform satisfies the radar systems transmiter's requirement of the hardware.Traditional algorithms for waveform optimization focus on performance improvement of single task of radar. In contrast, an algorithm based on combining waveform design with waveform selection is proposed. The optimal waveform for target detection is firstly designed based SNR criterion under energy constraint in frequency domain. Then a set of signals is generated according to the constant envelope signal method under the given Fourier amplitude, every signal in the set has the same optimal frequency waveform, which is for target detection. According to the waveform selection based on control theoretic approach, a best signal is selected from the signals set, which could minimize the both tracking mean square error. The proposed algorithm could improve the target detection and tracking performance.For robust target tracking improvement, an algorithm based on sum-zero game is proposed. The minimax filter should be designed for robust tracking firstly, then, the relationship of tracking mean-square error and transmitted pulse parameter is created, the tracking mean-square error is as criterion for waveform selection in order to select the best waveform from waveform database. The proposed algorithm could improve the robust target tracking performance.The most research on waveform optimization uses the current target impulse response (TIR) estimated. Optimizing waveform aims to optimize the next waveform that could mach the next TIR. However, the TIRS of different time in a short interval are different, thus, used current TIR estimated for next waveform optimized is improper. For what's mentioned above, a method based on TIR prediction is proposed. The concept and properties of the TIR are discussed. Then, the dynamical model of the TIR is created. The next TIR prediction is derived by Kalman filter. The the waveform is designed according to the TIR prediction with the SNR criterion and mutual information criterion, respectively. The performance of the propose method approximates the real performance.