Moment Estimation Of Amplitude Distribution Parameters Of Sea Clutter And CFAR Detection

Recently, with an increasing importance of sea clutter research in the military and civilian fields, moving target detection in sea clutter has become a hot issue in radar community. The accurate analysis and estimation of the statistics of sea clutter is the indispensable precondition to effectively detect targets in sea clutter background. Nowadays, for its effectiveness and examination by a mass of real datasets, the K-distributed clutter model is increasingly used in analysis, modeling and simulation of sea clutter. Moreover, moving target detection in the K-distributed clutter has been widely investigated and applied. As a result, the parameter estimation of the K-distributed clutter and the constant false alarm rate(CFAR) detection scheme become an important issue in practical applications.In the thesis, the statistics of sea clutter is firstly reviewed, including the characteristics of the amplitude distribution and the correlation of clutter time series. On the amplitude distributions of sea clutter, several commonly-used models are introduced and their related properties and their ranges for application are analyzed. On the correlation of clutter time series, the correlation along time and space are taken into consideration.Second, several classical target detection schemes in sea clutter background are reviewed, including the cell average CFAR(CA-CFAR) detector, the generalized likelihood ratio detector, the adaptive matched filter detector(AMF) and the adaptive normalized matched filter(ANMF) detector. In view of the flaws of CA-CFAR detector, the method of the ratio median detection which uses the two-dimensional reference window and has the ability of resistance to anomalous scattering cell is proposed. Moreover, the detection performance comparison between the CA-CFAR detector and the ratio median detection with the real sea clutter datasets are made. Compared with the CA-CFAR detector, the proposed method obtains better performance in the case of the environment with the abnormal scattering cells.Finally, the moment estimation method of the K distribution parameters of sea clutter is investigated, which form the basis to set up the constant false alarm rate thresholds of the matched filter(MF) detector and the normalized matched filter(NMF) detector in the K distributed clutter. It is proved that the CFAR thresholds of the MF and NMF detectors are solely determined by the integrated pulse number and the shape parameter of the K distributed clutter. Next, the CFAR thresholds of the MF and NMF detectors are derived and a method using the Monte-Carlo integral is proposed to calculate the CFAR thresholds under the K distributed clutter. Finally, the simulated K distributed clutter data with spatially-varying shape parameter is used to examine the CFAR property of the MF and NMF detectors. Moreover, the real sea clutter datasets are used to verify the effect of the CFAR detection using the AMF detector.