Model-Based Inversion Algorithms for Improved Tunnel Detection via Borehole Ground Penetrating Radar

Abstract

In recent decades, there has been an increasing emphasis on the border protection between America and its neighboring nations. The U.S. Customs and Border Protection, and Department of Homeland Security make ceaseless efforts to prevent illegal activities and threats from traversing the borders. As a result to the tightened security above ground, the utilization of underground tunnels has become more prevalent for illegal activities such as weapons smuggling and drug trafficking. To help further the security of the United StatesÍ borders, there is a strong interest in developing technologies to detect and localize these tunnels. æææDetection and imaging of tunnels in any given region of ground are made possible because the air that ?lls them is materially different from anything else underground, but challenging because of the large noise to signal ratio introduced by unknown variation of the ground surface and volumetric inhomogeneities. A Borehole Ground Penetrating Radar (BH-GPR) system has recently been explored for such purposes. The data collected by this system reflect the soil properties underground that in turn could suggest the presence of potential tunnels. æThe primary objective of this research is to develop new inversion algorithms for the BH-GPR system to detect tunnels utilizing a waveguide-based model to represent the layered profile of soil. These algorithms will more accurately represent the wave propagation properties that arenÍt taken into account in current techniques. They are also likely to enhance the accuracy of the detection and localization of tunnels existing in the complex media of the soil.