Source localization using widely separated special microphone arrays: study of the applicability of higher order statistics

This paper presents an application implying higher order statistics in source localization. Such techniques have been first introduced in the early 1990’s. Some interesting properties of higher order statistics are exploited in zone survey applications. A large global array constituted by a set of moving sub-arrays (soldiers wearing a Head Equipment (HE) equipped with one or more microphones, a GPS receiver and a magnetometer) allows the localization of the menaces near a survey zone. Studying a large global array, with second order statistics high resolution localization techniques, the local informations are not significant enough in order to have an influence on the estimation of the position of the acoustic sources (for instance, the orientation of the HE has no influence on the estimation of the position of the sources). We expect that the use of higher order statistics and their properties concerning especially the virtual equivalent network (VEN) will improve the significance of these local information and thus, allow to have a better estimation of the source position. A first study has been based on simulations, and then, some results with real data, collected in october 2003, are presented.