Dense Array Feed for Parabolic Reflector Antennas

Authors: J.D.Bregman, ASTRON, Dwingeloo, the Netherlands

M.J.Montero, Eindhoven University of Technology, the Netherlands

Abstract

Multi beam functionality of arrays can also be realized with a reflector antenna with an array feed to:

» improve observational throughput

» increase the field-of-view of a parabolic reflector with overlapping beams

» perform electronic scanning by interpolation of adjacent feed signals

» allow interference mitigation by combining signals of different beams

Overview

A brief description is given of the points that are treated in this presentation. First the configuration used in the model is illustrated in figure (1). Next the beam interpolation principle is explained in figures (2) and (3). This is done at two different frequencies to show the dependency of the beam overlap for a given separation of the feeds in the focal plane. Subsequently attention is paid in figures (4) and (5) to complex weighting based on the Conjugate Field Matching principle for the same two frequencies. In the figures (6) and (7) nulling according to the Spatial Projection technique is also shown. Finally conclusions and recommendations are given.

Configuration

Figure 1. Reflector Antenna with a linear 9-element array feed, F/D= 0.35, d= 0.125 m. By the use of multiple feeds the field-of-view of the parabolic reflector is increased. This principle of lateral displacement of the feeds can be interpreted as a sort of mechanical scanning. Feed displacement causes phase aberrations. Then symmetric patterns are no longer obtained. By combining electronic weighting and mechanical scanning beam interpolation is done.

Beam interpolation for 1.4 GHz

Figure 2. Beam interpolation for 1.4 GHz. The green pattern is the gain pattern of the third feed and the blue pattern is the gain pattern of the fourth feed. By using two neighbour feeds an interpolated beam is formed. This is done by weighting the electric far field of the two neighbour feeds and then apply superposition to obtain the total electric field. Then the gain is computed with the normalisation factors included. Trying to obtain a beam exact between the two adjacent beams can be thought as a “worst case”. Both feeds must have the same weight to achieve this. Here the electric far field of the two feeds are weighted by 0.5. The red pattern is the obtained interpolated beam.

Beam interpolation for 2.0 GHz

Figure 3. Beam interpolation for 2.0 GHz. Here the same procedure described for the 1.4 GHz is followed. The red pattern is the interpolated beam. It can be seen that there is a dip now in the beam. This can be easily explained by the fact that the element spacing between the feeds is too large.

Conjugate Field Matching for 1.4 GHz

Figure 4. Scan samples according to the Conjugate Field Matching [1],[2] weighting coefficients at 1.4 GHz. In the CFM method complex weighting coefficients are used to maximize the gain in a given direction. This also means that the phase aberrations introduced by lateral feed displacement are minimized. The scan angle here is the direction where the maximum gain is desired. This value does not always coincide with the top of the overall antenna pattern. In the figure one can see that from approximately 3° there is not much difference among the scan samples. This is because the maximum of the scan range has been reached. The scan range is limited by the edge feed element.

Conjugate Field Matching for 2.0 GHz

Figure 5. Scan samples according to the Conjugate Field Matching weighting coefficients for 2.0 GHz. This is the same principle as in the previous figure but for a higher frequency. One can see that the main beam is highly distorted in the scan range. Out of the scan range (beyond 3°) a constant pattern is visible. This is again the pattern of the edge feed. So here even the CFM optimum coefficients can not minimize the phase aberrations to obtain a nice beam. The distance between the elements is still too large.

Beam steering and nulling for 1.4 GHz

Figure 6. Steered beam at q=1° with a null at q=-2° according to the Spatial Projection (SP) [1],[3] principle for 1.4 GHz. The SP principle is very simple. A beam is steered to the angle of an undesired side lobe and then the two patterns are subtracted. Very deep nulls are obtained with this method.

Nulling in a range of angles for 1.4 GHz

Figure 7. The scanned beam is at q=1°. Beginning from –8° the interference angle (null angle) is each time incremented by 0.1°. In the figure one can observe the behaviour of the null in and out of the scan range. The “moving null” is given in blue. When the null is too close to the main beam this will be affected.

Conclusions and recommendations

It can be concluded that smooth scanning (continuous beam steering in a given range) is possible only in a limited frequency range for our specific feed element separation and F/D ratio of the telescope. For frequencies higher than 1.4 GHz smooth scanning is no longer possible. The element spacing is then too large for effective beam interpolation.

For frequencies below 1.4 GHz where smooth scanning is possible, multiple beams can be formed as well. Nulling can be obtained as well, within and even out of the scan range which seems surprising at first sight. Beam steering, multibeaming and nulling which are the main features of a conventional array can be realized with a hybrid antenna system as well.

These conclusions are based on a 1-D linear array feed and are very promising. This is a good reason to continue with a full analysis of a system with a 2-D array feed.

Till now only a scalar approach has been treated. It is important to know what effects cross polarization components could have in this hybrid antenna. So one of the next steps should be making a vectorial analysis. Also measured patterns of the array elements should be used in the next analysis. The final goal is to analyse the whole THEA tile, so a 2-D array instead of a 1-D. 2-D scanning can then be done as well.

References

[1] Montero, M.J., A Feasibility Study on Reflector Antennas with phased array feed, ASTRON Dwingeloo/Eindhoven University of Technology, Master Thesis, August 2001.

[2] Mano, S., Katagi, T. and Tsutsumi, T., Pattern Synthesis of Array Fed Reflector Antennas, Electronics and Communications in Japan, Vol. 67-B, No. 7, 1984, p. 38-45. Translated from: Denshi Gakkai Ronbunshi, Vol. 67-B, No. 7, 1984, p. 202-208.

[3] Subbaram, H. and Abend, K., Interference Suppression Via Orthogonal Projections: A Performance Analysis, IEEE Transactions on Antennas and Propagation Vol. AP-41, No. 9, September 1993, p. 1187-1194.