Netrebenko K. Development of radar objects recognition devices in presence of noise using higher-order statistics

Object - the recognition process of non-Gaussian signals by their parameters of the generalized autoregression models and higher-order statistics. The purpose - enhancement of radar-echo objects recognition devices which operate in presence of noises. Methods - methods of recognition theory, stochastic modeling, elements of linear prediction theory, probability theory, recursive filters theory, applied analysis of non-Gaussian stochastic processes. Results - new non-Gaussian signals recognition features: samples and the standard deviation of moment function of prediction error, cross-moment function between the moment functions of the pattern and test signal. It is introduced - into a budgetary scientific research program in Kharkiv national university of radioelectronics. Sphere of application - pulse non-coherent radars with narrow-band probing signal.