Research On Cognitive Radar Adaptive Waveform For Clutter Suppression

Cognitive radar is a fully adaptive radar with capacity of environmental perception, reasoning study and decision-making, which is considered to be the future direction of radar development. Compared with the traditional radar, the most notable feature of cognitive radar is the feedback from the receiver to the transmitter, so that the radar could transmit adaptive waveform matching the characteristic of the target and envirmoment and eventually achieving to improve the performance of radar. The perception of the environment and transmission of adaptive waveform are the two key technologies to the implement of cognitive radar. This thesis will focus on these two issues.The first section briefly reviews the notion of cognitive radar and its development status of key technology, and summarizes the existing methods of adaptive waveform design and time domain signal synthesis. Aiming at meeting the need of optimal detection waveform design, the clutter doppler spectrum perception is introduced.The second part studies the model of sea clutter doppler spectrum using simulation data for the need of detection problem. Firstly the backscattering coefficient of two-dimensional sea surface is generated after which the doppler spectrum estimation is operated under various wind speed, wind direction, emission signal bandwidth. Secondly a curve fitting is used to research the relationship between the doppler spectrum and these factors, which provides a priori information for clutter suppression and detection performance analysis.In the third part, an adaptive waveform selection method is discussed with the target detection problem as the application background. The selection problem for LFM signal of different bandwidth and envelope is discussed using the shift coefficient as the cost function. Then aiming at the condition of the possible existent target whose doppler frequency shift is unknown but assumed a reasonable distribution, the waveform selection problem is studied under different prior distribution of target doppler frequency shift. We use different cost function for the application of different requirements and then the detection performance using different LFM signal parameters are compared for each condition.The fourth part studies the optimal detection waveform design and its realization in the time domain. Under the constraints of signal bandwidth and energy, an optimal NP detector and waveform are deduced in the frequency domain for extended target. Then the optimal signal synthesis in the time domain is researched for the implement of the optimal signal. In order to make full use of the radar transmitter power, we impose the maximum waveform modulus constraint upon the phase-modulated waveform and then propose a gradient descent method to find the optimal phase sequence. With the proposed method, the synthetic time domain signal energy spectrum density and the optimal signal energy spectrum density have the minimum mean square error, which means that we achieve the best approximation with the optimal signal spectrum.