Source Number Estimator In Pulse Train Data
J.E. Perkins · 2005
Data that consists of a number of superimposed pulse trains (each train having a simple pulse arrival pattern) is considered. The number of independent trains present is of interest in a number of applications. By considering the variance (over window position) of the number of pulses lying in a time window the number of periodic sources can be determined. This method of Cox and Smith is generalized from strictly periodic emitters to those with noise or periodic patterns. 1. INTRODUCTION There are a number of applications that concern the analysis of pulsed signals, such as radar, bat echo-location, neuronal signals, packet telecommunication networks, and the study of pulsars in astronomy. A subset of these deal with a number of sources each emitting a simple pattern of pulses. In these cases the number of sources and their statistics is of interest. In many cases this must be determined with knowledge only of the time of arrival of the pulses. Traditional methods [2] for pulse train ...