Research On Anti-interference Waveform Design For Polarimetric Cognitive Radar

Cognitive radar(CR) is an intelligent radar system, which has recently attracted considerable attention. CR has a feedback channel from receiver to transmitter, it means that CR can intelligently update the transmitting waveform or the receiver filter, based on the extracted signal of targets and environment from radar returns. This characteristic of CR makes it possible that the transmitter can adaptively transmit electromagnetic wave, according to the specific radar tasks and circumstances, and maximize the radar performance. As a kind of additional information, Polarization strengthens the radar performance of anti-interference and parameter estimation. Thus the point of study is focused on the design of optimal polarized waveform for CR. The article is organized as follows.In the first part, the basic theory and knowledge of polarimetric cognitive radar(PCR) is investigated. First of all, the system model of polarimetric cognitive radar is built, PCR is the cognitive radar where polarization is added, so the transmitter can adaptively design and transmit the polarized waveform. Then the polarized signal is formulated with the polarization ellipse, which builds a solid foundation of polarization parameter estimation below. Next, the concept and importance of polarization scattering matrix are introduced, finally, the paper addresses the clutter models and figures out that the real clutter data can be very well approximated by the compound-Gaussian model with inverse gamma distributed texture.In the second part, the optimal polarized waveform design for parameter estimation is considerd. Firstly, the conditional probability density function is formulated, then the parameter-expanded expectation-maximization algorithm is introduced based on maximization likelihood estimation. It is used to estimate the target and clutter parameters accurately. Next, the optimal polarized waveform design algorithm is obtained by minimizing the Cramer Rao Bound. Finally, the sub-optimal algorithm is derived to reduce the computation and complexity, while its perfomance is almost comparable as former. Simulation results demonstrate that the algorithms in this part can acquire the precise target and clutter parameters, which provides the exact prior information and environment analysis.In the third part, the design of orthogonal waveforms for instantaneous polarization radar in the cognitive framework is studied. To decrease estimation error caused by high correlation level of the two transmitted waveforms, a method named Phase Only Spectral Approximation Algorithm(POSAA) is proposed to design a couple of waveforms with low correlation level. First of all, the object function is established under the criterion of integrated sidelobe level(ISL). Next, the object function is deduced in frequency domain according to the relationship between correlation function and the power spectral density of the waveforms. Finally, the optimization problem is tackled by trust region algorithm using its gradient and Hessian matrix. Besides, POSAA takes advantages of sequencial design, that is to say, it designs waveforms one by one, which avoids the high complexity arised from simultaneously optimization. The numerical simulations prove that the designed waveforms have a good correlation level, and the estimation error of target polarization information is very low.