Research On Direction-of-Arrival And Related Parameters Estimation Of Radar Signals

Because of the adoption of two independent, analog processing systems to the estimation of frequency and direction-of-arrival (DOA), the traditional electronic reconnaissance equipments are very complex with simplified algorithm of signal processing. With the development of digital receiver and array signal processing, it is possible to make the frequency estimation system integrate with the DOA estimation system. With the reconnaissance equipments intergration, the equipments will be simple. At the same time, many processing algorithms can be used flexible, and then the signal information can be used efficacious.The aim of the dissertation is to research the algorithms of spatial time-frequency reconnaissance. The dissertation can be concluded as follows:Research on joint estimation of signal parameters based on phase unwrapping is addressed. When the signal-to-noise ratio is high enough, the additive complex white Gaussian noise can be equivalent to the real phase noise. Because the frequency information are contained in the phase, and the DOA is also reflected on the phase difference of the signals which received by two sensors, the joint estimation of frequency and DOA can be realized by phase and phase difference unwrapping. The simplicity and real-time of the algorithm make it will realization in practice.The enhancement of time-frequency distribution (TFD) is discussed. Wigner-Ville distribution is a good tool in time-frequency analysis. For the concentration of the signal energy along the instantaneous frequency and the average distribution of the noise energy in the whole band of the noise, WVD can withdraw the time-frequency feature from the noise. It is necessary to enhance the time-frequency feature when signal-to-noise ratio is low. Using the array, the sensors receive multiple duplicates of the signals. The TFD can be enhanced by using the cross Wigner-Ville distribution (XWVD) of the outputs of the sensors because the noises between the sensors are irrelevant. When symmetric array is used, the arithmetic average of the XWVDs of the multiple group symmetric sensors will achieve extra enhancement of the TFD.The algorithm of joint estimation of time-frequency and DOA based on time-frequency analysis is presented. Using the outputs of the sensors, the XWVD ofoutputs can be obtained. The phase of the points along the time-frequency ridge of the signal is the product of the frequency of the point and the time delay. With the conclusion, time-frequency interferometer is used to joint estimation of time-frequency and DOA. The time-frequency interferometer can be easily realized because the sensors needed are few and the computations are little relatively. The time-frequency interferometer has the inconsistency of the precision of estimation and the range of the DOA which will be estimated. In wideband array signal processing, the wideband signals can be transformed into narrowband using focusing matrices. And then the traditional algorithms based on subspace can be adapted to the narrowband signals. Unfortunately, the frequency and DOA of the signal which are the parameters be estimated must be known in advance. Using the symmetric array and bilinear TFD, the signal can be focused onto the instantaneous frequency of the signal. Thus, the enhanced TFD can be obtained by array average, and then the frequency and phase of the point along the time-frequency ridge can be withdrawn. Finally, we can construct the correlation matrix and use the subspace algorithm in order to estimate the DOA. Symmetric array is needed in spatial focusing. The spatial time-frequency distribution (STFD) matrices are presented for the arbitrary geometry of the array. The STFD matrices are adapted to narrowband signals. The modified STFD matrices are addressed for the wideband signals.The effect of the channel uncertainty is analyzed. The channel uncertainty model based on channel gain and group delay is given. The TFD enhancing based on symmetric array average and DOA estimation is robust with the channel gain uncertainty. But the channe...