Detection of a Chirp Ionosonde Signal under Conditions of a Priori Non-Parametric Uncertainty

Abstract

We consider the model of a mixture of probability distributions for background and target samples in the problem of extracting a signal from radiophysical sounding under conditions of a priori nonparametric uncertainty. Based on the study of this model, approaches relying on the risk function and selection of anomalous readouts in the sample are proposed for detection and extraction of a continuous ionosonde signal. Semi-empirical modeling and testing by experimental data are performed.