Experimental-calculated method for searching far-field of electrically and physically large reflector antennas

Reflector antennas of large electrical and physical size are widely used in spaceborne radars. Far-field measurements of such antennas are very organizationally and technologically hard. The far-zone antenna polygon for measuring such antennas must have propagation path for about hundreds meters or some kilometers. In case of near-field measurements the size of the measuring bench must be con-siderably larger than the reflector antenna size. The measuring process of the big reflector antennas can be greatly simplified by using the experiment-calculated methods. The main feature of these methods is using of measured electrical characteristics of the feed-antenna when computing far-field of reflector antenna. As the measured data can be used measured 3-D far-field or near-field of the feed-antenna.

In the article the experiment-calculated method was illustrated through the example of the parabolic reflector antenna with dimensions of 12 meters on 4 meters with linear vibrator S-band antenna array as feed-antenna. Computational modeling was made in program FEKO. There was made computational model of reflector in accordance with design documentation and imported measured near-field of feed-antenna was used as the source. There was not any possibility to compare measured far-field of real reflector antenna with far-field of reflector antenna obtained with experiment-calculated method. As known from technical literature modern computational programs such as FEKO allow to create computational antenna models with electrical characteristics identical to real antennas, and the electrical charac-teristics of these models can be used as reference. To validate the offered experiment-calculated method far-field of the reflector anten-na, obtained with this method, was compared with far-field of computational model of the reflector antenna.

There are several main negative factors which can cause errors in far-field obtained with experiment-calculated method when validating: 1) the manufacturing errors in feed-antenna; 2) the near-field measuring errors because of the limited sizes scan region; 3) the errors when converting measured near-field data to near-field format using in FEKO; 4) the errors in computational modeling with imported measured near-field using as excitation source. As a result of made research it was established that measured and computed far-fields of the feed-antenna were practically identical in wide beam sector, greater than reflector beam sector. This fact shows that feed-antenna and its computational model are identical in their far-fields and so that the factors 1) and 2) have very small negative influence.

The comparison of far-fields, obtained with the experiment-calculated and computational methods, showed its good correlation. This fact shows that the negative influence of factors 3) and 4) is very small, and the offered experiment-calculated method is very good suitable to determine far-field of large electrical and physical size reflector antennas.