Estimation of Loss When Detecting Signals by a Receiver with Adaptive Threshold on the Basis of the Method of Ordered Statistics

Abstract

We consider the method of formation of adaptive threshold of signal detection against the background of receiver inherent noise using the nonparametric algorithms on the basis of ordered statistics. The simulation results and the efficiency of using this method on the basis of estimating the quantiles of the statistical distribution of the process compared with the classical methods of “moving average” in the case of complicated signal-interference environment (weak-signal masking by intense interference, mutual masking of several signals simultaneously staying in the sliding data window, and the useful-signal presence in the region of the jump-like variation of interference) are shown. A mathematical model of estimating the loss introduced when detecting a useful signal by the threshold device based on the method of ordered statistics is developed.