Research On Parameter Estimation Of Radar Signals Based On Time-Frequency Distribution

Time-frequency analysis was introduced as a effective means of representing signals whose frequency is varying with time. For the typical non-stationary signal of linear frequency modulation (LFM), both time domain representation and frequency representation are inadequate to describe the signal appropriately. Time-frequency analysis and radar signal processing are combined in this dissertation, and mainly discusses estimation of direction of arrival, detection of moving target and Doppler parameter imaging of SAR.1. A group of vectors is obtained by correlating each receiving array element with output signal of reference array element. The output Toeplitz matrix of array is constructed with the using Toeplitz matrix structure algorithms. Simulation results show that this method can effectively enable the estimation of coherent signal number compared to traditional algorithm.2. The dissertation is focused on the DOA estimation of LFM signal based onTime-frequency. Spatial average processing technology in FRF domain is proposed to construct SATFD matrix to eliminate the coherence between the signals. Finally, simulation results validate the method can improve the estimation accuracy compared to the spatial smoothing technique.3. The signal model is superposition of multiple LFM signals after SAR/DPCA clutter suppression technique. FRFT and STFT-Radon will make the moving target signal present regular distribution by using the property that LFM signal can be de-chirped in time-frequency domain. Simulation results testify the algorithm improves the performance of moving target detection and parameter estimation.4. Moving targets of the different slant-range velocity which perhaps locate in the same resolution cell can not be imaged to correct positions. To solve the difficult point, a two-channel detecting method replacing the FFT with high-resolution MUSIC spectrum estimation technology is proposed. The simulation results validated the method can improve the capability of detecting multiple moving targets which locate in the same resolution cell.