GPU acceleration of Threat Map computation and application to selection of sonar field controls

Abstract

The Threat Probability Density Map displays the outcomes of the search effort prior to detection and is a digital representation of the probability density function of location of the existing undetected threat. The Threat Map readily provides such diagnostics as the probabilities of the threat being present in different areas of interest and this information can be utilised in selection of sensor field controls. In this work we consider a GPU acceleration of the Monte-Carlo technique for Threat Map computation and discuss application to sonobuoys.