Estimating distributed objects inside buildings by moving sensors Marija M. Nikoli 1 , Arye Nehorai 1 , and Antonije R. Djordjevi 2 1 Department of Electrical and Systems Engineering, Washington University, One Brookings Drive, St. Louis, MO 63130, USA nikolic2@ese.wustl.edu, nehorai@ese.wustl.edu 2 School of Electrical Engineering University of Belgrade, P.O. Box 35-54, 11120 Belgrade, Serbia edjordja@etf.bg.ac.yu Abstract: We develop a radar system with a moving sensor scheme for localizing distributed objects hidden inside buildings. We first estimate the unknown building wall parameters, using geometrical optics (ray tracing) model and maximum likelihood approach. Time-of-arrival information is evaluated at different locations and utilized to estimate the position and shape of the object. We consider the influence of the wall on the signal propagation. Simulations results in the 2D case verify the accuracy of the proposed method. Keywords: through-the-wall, radar imaging, time-of-arrival, range-gate, wideband signals. 1. Introduction In this paper, we address an important problem in urban warfare - that is peering inside buildings using electromagnetic sensing. This is a difficult task due to complex and usually unknown environment. Recent publications on this topic focus on simplified scenarios, such as estimating locations of point scatterers behind a single wall with unknown parameters [1-2]. In realistic situations, the point-scatterer approximation may not be valid since objects are close to the sensing system. We propose a radar technique for detecting distributed objects hidden behind walls and other opaque obstacles. We use moving sensors to form a virtual array and increase spatial resolution. We tackle some problems that may be encountered in practical realizations, such as unknown wall parameters. In contrast to [1-2], we use parametric (physical) models to determine the permittivity and thickness of the wall. The proposed technique is simulated using a 2D electromagnetic model, which is described in Section 2. One possibility to distinguish among various reflections, to remove the clutter, and to determine the distance between the antenna and the target is to employ range-gating [3], as presented in Section 3. Finally, in Section 4, we present some results of the simulations. 2. Electromagnetic modeling In our modeling, we use a 2D method of moments (MoM) code, since electromagnetic modeling of electrically large 3D structures (such as buildings with occupants) is still inadmissibly time-consuming. However, the approach we propose for estimating positions of hidden objects is general and can be easily applied to 3D problems.