Publications [#63850] of Lawrence Carin

Papers Published

1. Carin, L. and Dogaru, T., Optimal detection of deterministic targets buried under a rough (random) air-ground interface, International Symposium on Electromagnetic Theory. Proceedings, vol. vol.1 (1998), pp. 81 vol.1 -, Thessaloniki, Greece
   (last updated on 2007/04/14)

   Abstract:
   Summary form only given. We consider the detection of a deterministic target buried under a rough air-ground interface. The rough surface is parameterized as a random process (i.e., while the details of a particular realization of the surface are not known exactly, each surface is assumed to represent one realization of an ensemble, characterized by known statistics). Since the rough surface is treated statistically, the fields scattered and transmitted (for penetrable surfaces) at such a surface must be parameterized statistically as well. In the frequency domain, the fields scattered at a given frequency are parameterized as a random variable, usually by the mean and variance (under the assumption that the random variable has a Gaussian distribution). For time-domain scattering, of interest here, the fields must be parameterized as a random process. We consider implementation of an optimal detector which properly accounts for the random nature of the target's scattered signature. An optimal scheme for detection of a random signature requires integration over the signature's density function. Unfortunately, such a density function is difficult to obtain in general (due to the finite support of the transient target response, simple. e.g., Gaussian, models are inappropriate). Therefore, we implement the optimal detector approximately via Monte Carlo integration, through consideration of multiple realizations of the target signature (each for a particular rough surface, from an ensemble of such). While fast algorithms are well known to be requisite elements in inverse-scattering schemes (generally for deterministic scattering data), here we introduce the use of such in optimal detectors (for stochastic scattering data)

   Keywords:
   buried object detection;deterministic algorithms;electromagnetic wave scattering;electromagnetic wave transmission;Monte Carlo methods;optimisation;random processes;signal detection;statistical analysis;stochastic processes;time-domain analysis;