THE METHOD OF ESTIMATING OF THE PARAMETERS OF THE GAUSSIAN NOISE DISTRIBUTION FOR DETECTION OF PULSE SIGNALS

 

I. S. Osadchiy

Scientific and Production Association “Lepton”

 

The paper is received on March 24, 2015

 

Abstract. In this article the problem of estimation of mean and variance parameters of additive Gaussian noise for detection of non-coherent pulse signal is described. Analysis of this problem revealed that it is similar to the robust statistics task of estimate of data with outliers. Some of the popular methods that can be implemented for digital real-time data processing were described. New method of parameters estimate was offered. It consists of two steps: removal of outliers and robust estimation. Research and comparative analysis of effectiveness and stability of estimation for different methods is done.

Key words: problem of signal detection; pulse signal; robust estimation.