The Signal Track Search algorithm

摘要

In order to detect gravitational waves from partially modelled/unmodelled sources we will need to have in place a group of robust data analysis algorithms. We discuss here, a particular time-frequency strategy, namely the Signal Track Search (STS) strategy, which is both robust and sensitive for a certain class of signals buried in Gaussian noise. We assume that the signal is a a broadband burst containing several cycles. Our basic strategy involves the construction of time-frequency maps of the detector output and subsequently a search for curvilinear structures in the maps. Though this strategy has been found to work extremely well when the noise is Gaussian, it remains to be seen whether the strategy will be effective when the detector output is further contaminated by non Gaussian noise sources. In this article we will discuss various unresolved issues pertaining to the STS algorithm. [References: 4]