Comparison of Data Processing Techniques for Calibrating Low Frequency Antennas with Long Ring Down Time in An Extrapolation Range

Extrapolation method is regarded as one of the most accurate methods for obtaining the absolute far field gain of an antenna. This paper will compare the efficacy of several data processing techniques for calibrating low frequency antennas with long ring down time. Traditionally, measurement data are pre-processed to remove ripples from multipath reflections before a curving fitting is applied. We will first investigate two traditional data processing techniques. The first technique is to apply time domain gating to the vector response vs. frequency data at each separation distance. Then the gated data as a function of distance is fitted to the polynomial equation. The second technique is spectrum domain filtering. The vector response as a function of distance is transformed to k x domain at each frequency. A band pass filter is applied in k x domain to keep only the direct antenna response. In this study, we propose a new approach - the magnitudes of the antenna response as a function of distance including the ripples are fitted to a more complete generalized antenna response equation with the antenna-to-antenna multiple reflection terms included. This paper will compare the three techniques using a set of measurement data on double-ridged waveguide horn antennas in a fully anechoic extrapolation range.