 |
| |
|
This case study follows the work of Montoya and Smith[1]. In this paper Montoya investigates a resistively loaded vee dipole antenna for use in a landmine detection GPR system. The objective of this case study is to calculate the signal at the antenna feed resulting from the soil interface and buried mine. The geometry is shown left, illustrating the vee dipole antenna, soil interface and simplified mine target. Validation is carried out against results published by Montoya.
This application is demanding as the signal to noise ratios are extremely low, which means that clutter must be reduced to an absolute minimum, ideally without losing any signal information. This is achieved in impulse systems by using a loaded antenna to prevent ringing. The downside of this approach is the reduction in the efficiency of the antenna and the optimisation of such a system relies on the trade-off between damping and signal level. In the case studied here the antenna is constructed from material with constant resistivity. The loading profile of the antenna with length is achieved by changing the profile of the dipole arms as a function of distance from the feed. The dipole is excited with a Gaussian impulse and radiates its derivative with a peak frequency of 4GHz. The dipole arms have a length approximately 3 times the characteristic width of the Gaussian impulse, sufficient to dissipate the energy in the antenna within a single transit. In the present case the antenna is situated 4cm above the soil interface. The soil is a dispersive medium with suitable parameters for a representative soil of 10% moisture content. The simplified mine target is a dielectric cylinder of relative permittivity 3.0, diameter 8cm, thickness 3.2cm, with its top surface 2cm below the soil interface. This scenario is typical for this application.
The Celia model is constructed in an identical fashion to the model of Montoya. The antenna feed is a balanced transmission line type and a 10 cell PML was used to terminate the computational domain. The cell size used was 0.75mm, providing at least 10 cells per wavelength within the soil at the highest frequency of interest. This was classified as a frequency at which the signal is 40dB down on its peak amplitude, or around 14GHz in this case. The Celia model is shown below, the soil being in cross section to illustrate the buried mine target.
The results show the 'difference voltage' measured at the feed of the antenna. This is the difference between the feed voltage with antenna in-situ above soil and target and the antenna in free space. Montoya publishes full calculations using only a simplified non-dispersive soil model. Calculations using Celia illustrate results for the simple non-dispersive soil and the full dispersive soil model in comparison to the Montoya result.
1 Montoya T P, Smith G S
"Land Mine Detection Using a Ground-Penetrating radar Based on Resistively
Loaded Vee Dipoles"
IEEE Transactions on Antennas and Propagation
Vol 47, No 12, Dec1999, pp 1795-1806