The detection of buried land mines is a problem of military and humanitarian importance. Electromagnetic (EM) sensors (ground-penetrating radars) use signals at radio and microwave frequencies for this purpose. In the past, EM sensors for land-mine detection have been empirically developed and optimized. This has involved experimental tests that are complicated, time consuming, and expensive. An alternative is to carry out initial development and optimization using accurate numerical simulations. One objective of this paper is to show, for the first time, that such simulations can be done using the finite-difference time-domain (FDTD) method. The separated-aperture sensor has been under investigation by the United States Army for land-mine detection for many years. It consists of two parallel dipole antennas housed in corner reflectors that are separated by a metallic septum. It is a continuous-wave sensor tuned to a particular frequency (typically 790 MHz). When the sensor is over empty ground, the coupling between the antennas is very small. As the sensor is moved over a buried mine, the coupling between the antennas increases indicating the presence of the mine. In this paper, the complete EM system composed of the separated-aperture sensor, air and soil, and buried land mine is modeled using the FDTD method. The finite computational volume is truncated with an absorbing boundary condition: the generalized perfectly matched layer. Detailed studies made with the simulation increase the understanding of this sensor. Results computed from the simulation are in good agreement with experimental measurements.
展开▼
