Research On Clutter Modeling And Characteristic Test Algorithm

Simulation of radar's working environment is inevitable in research and development of modern radar system. Radar clutter is an important part of radar's work environment. Modelling and simulation of radar clutter can reconstruct radar's working environment, which plays a vital role in the development of modern radar system. To test whether the clutter simulated by computer is consistent with expected requirements, the test methods of clutter characteristics will be directly used to assess the statistical properties of simulated clutter. So it is important to have clutter characteristics test methods with good performance.A new type of probability density function estimation algorithm based on windowed Fourier transform of characteristic function was proposed. Frequency rectangle and Blackman 4 terms window were chosen and their corresponding expressions in time domain were derived. The mainlobe/sidelobe characteristics of those two time domain expressions were discussed. Secondly, probability density function estimators corresponding to frequency rectangle and Blackman 4 terms window were concerned. Furthermore, the steps of how to choose the frequency window length, which is the key factor determining probability density function estimation performance, was illustrated based on Parseval theorem. Simulation comparisons were given to verify the validity and effectiveness of the proposed algorithm.We also proposed a ZMNL scheme to generate coherent lognormal clutter. The relationship between real part autocorrelation function of input complex Gaussian sequence and that of output lognormal sequence was derived. Analogously, the relationship of real/imaginary parts correlation function was also been derived. In addition, those two nonlinear equations have been illustrated by function curves concerning with different parameters. A method of how to solve those two nonlinear equations was given after that. Furthermore, steps of how to generate coherent lognormal clutter was given in detail. Finally, the proposed method has been tested with two examples with different power spectrum density, and the experimental results indicate that the scheme works very well.