Intrinsic modeling of antenna array in near-field conditions

Antenna arrays have been increasingly used in many civil engineering and geoscience applications as they allow collecting multi-offset measurements simultaneously, thereby providing additional information for subsurface imaging and characterization. We extended a new near-field intrinsic antenna modeling approach to antenna arrays. The array was considered as a combination of couples of transmitting-receiving antennas with different offsets. Each couple of antennas was characterized using an equivalent set of infinitesimal source/field points and reflection/transmission transfer functions. We proposed an iterative approach to calibrate the model through which the antenna model was progressively completed. To reduce the number of simultaneous unknown parameters, linear and nonlinear optimization algorithms were combined together. We also applied the time gain for both modeled and measured array data to compensate for the wave attenuation, which is expected to improve the accuracy of the calibration. We validated the proposed calibration approach to an antenna array with two-Vivaldi antenna elements operating in the frequency range 0.8-3 GHz. The offsets between the two antennas were 20 and 40 cm, respectively. Calibration data consisted of 100 measurements corresponding to the antenna array at 100 different distances from a copper plane. The calibrated and measured antenna array data closely agree, with correlation coefficients larger than 0.9979 and root mean square error less than 2.3×10-5. These results open a new development avenue to apply the antenna array for digital soil mapping and non-destructive testing of materials using full-wave inverse modeling.