Radar Passive Interference Modeling And Noise Suppression Method

In order to get better performance in a passive jamming background, some signal processing must be done to enhance the signal-to-clutter ratio. Passive jamming modeling and clutter suppression signal processing methods based on clutter modeling are studied in this thesis, and some useful conclusions are obtained after computer simulation. The main jobs in this thesis are summarized as follows: Clutter modeling methods of single-channel radar are studied, including distributive model, relational model, chaos model, ocean clutter time domain model and α-stable noise model. Simulation process of Rayleigh, Log-normal, Weibull and K-distribution clutter with special power spectrum is illuminated. In addition, the modeling approach for chaff jamming echoed signal is studied from the point of view of interference characteristic and generation mechanism. The AMTI methods based on the maximum average improvement factor criterion and the minimum output clutter power criterion are studied for the rejection of clutter with a power spectrum density function of the Gaussian or like Gaussian type. Another adaptive clutter suppression method based on the minimum residue noise power is also discussed. The clutter in one range-azimuth cell is used to filter the clutter in neighbor range-azimuth cell where a target signal exists, and the residue noise power is equivalent to or less than the required system thermal noise level. Simulation results show the effectiveness of these methods. The ocean clutter received by radar can be modeled by two narrowband signals, with time-varying frequencies centered around the two Bragg frequencies. A method of tracking the time-varying frequency by singular value decomposition (SVD) and reduction of a Hankel matrix of the time series data is studied. The frequencies of clutter signals can be removed and another reduced rank Hankel matrix is constructed by a similar but reverse signal processing, hence a new time series data with clutter signals suppressed can be obtained. The α-stable processes can better model the impulsive random signals and clutter in physical radar systems. The adaptive linear prediction methods based on the fractional lower order statistics are studied for suppressing the α-stable noise, and simulation results show the effectiveness of them.