Notch Antennas

Reflector Systems

Passive Components

Multiband Systems

Wideband Systems

Installed Performance

High-power Components

ANTENNA INSTALLED PERFORMANCE AS APPLIED TO LEOs

Introduction

This was the text of an invited presentation to the QMWC (London, UK) Symposium on Antennas in a special session devoted to Low Earth Orbit (LEO) Satellite systems.

There are a number of problem areas in designing antennas for an LEO communications system which are caused by the local structure. The local structure is defined as anything in the vicinity of the antennas which affects the antenna. This structure may be conducting, dielectric or lossy and may affect the antenna match, its bandwidth and the radiation pattern in any direction. The antenna may be affected in one or all of these parameters.

In the case of systems which include LEOs, the antenna on the ground is likely to have a low gain and a broad beamwidth which makes the antenna more vulnerable to the local structure. In the LEO itself, the antenna will have a high gain but coupling to other antenna by way of the structure is important.

Simple Example

A simple example is that of halfwave dipole parallel to a cylinder of height, H, and diameter, D. If the distance between the dipole and the axis of the cylinder, L, is varied, then the dipole performance will vary with the distance, L. In the far-field, that is, when L is large, there is no effect. When L decreases to about 1.0 lambda, the radiation pattern will change. When L is < 0.25 lambda, the cylinder is in the nearfield of the dipole and the impedance changes. At a distance L < 0.1 lambda, the centre frequency changes.

Figure 1a and 1B show the effect when the cylinder has a diameter of 1.0 wavelengths and a height of 2 wavelengths. Clearly the impedance and the centre frequency have changed with changing L. The corresponding radiation patterns are shown in Figure 2.

Every real geometry for an installation is different and analytical methods are of little help. For the effect on match and centre frequency, a modelling tool is needed which will provide the currents on the antenna and the structure. The Method of Moments (MoM) can be used as can Finite Difference Time Domain (FDTD) but there are problems with runtime and memory. However only the structure nearest to the antenna needs to be modelled. For the effect on radiation patterns, the whole structure is needed and the Geometric Theory of Diffraction can be used. Table 1 provides a Review of CEM Tools which can be used.

Table 1 Review of CEM Tools

Application to Earth Terminals

The gain will be moderate, say, a peak gain of 15 dBi. The frequency is 1.8 GHz with circular polarisation and the antenna will be mounted on a vehicle, for example, a lorry, small boat, car, aircraft. All these structures can be partially conducting and can contain dielectrics. On these vehicles, the antenna boresight must be variable in order to access the satellite at all times.

Figure 3 shows an array antenna mounted on the metal roof of a bus, towards the front. Figure 4 shows the radiation pattern of the array in free space with a beamwidth of 45.0 degrees. Figure 5 shows the Elevation radiation pattern when the array is installed and with four different Elevation angles, looking over the front of bus at Elevation 15.0 and 45.0 degrees and looking over the rear of the bus, also at 45.0 and 15.0 degrees Elevation. There is severe disruption of the pattern at low Elevation angles and the gain has decreased from the free space figure of 17 dBi. Figure 6 shows the effect of moving the array forward to a position only 0.5 GHz from the front. The pattern is still distorted at an Elevation of 15.0 degrees to the front and the peak gain has decreased to 11.7 dBi.

Figure 7 and Figure 8 show a contour plot of the radiation pattern from an antenna in free space and then mounted on an aircraft.

Application to Spacecraft Antennas

These will have higher gain than the Earth terminals discussed above and the local structure will have little effect on match and bandwidth. A well-designed satellite antenna will have low sidelobes to meet CCIR specifications on radiation and so local structure is not an important effect on the radiation pattern. The problem area for these antennas is isolation. Since very good isolation is demanded, the solution depends on the frequency and dimensions and since much of the satellite structure will have to be included, MoM and FDTD is unlikely to be of use. Ray tracing (Diffraction Theory) will have to be used. Higher order interactions, that is, multiple bounce, must be included in the calculations.

Accuracy

The accuracy of impedance depends on how well the local structure has been modelled. It is usually necessary to model the structure or the part nearest the antenna, optimise the position, the antenna design and the structure and then build part of the structure to check the modelling results. The impedance from the model should be accurate to 10%.

Radiation patterns depend on how well the structure has been modelled. typically +/- 1 dB for gains greater than 0.0 dBi and +/- 3 dB for gains greater than -20 dBi. Computations of isolation are less accurate and might be +/- 10% in dB down to -60.0 dB and +/- 30.0 dB at -100 dB, for example, -45.0 +/- 4.5 dB (Figure 9).

Conclusions

The installation site and geometry must be examined for problems. For matching, use MoM/FDTD for the antenna and as much of the structure as possible. Check on the centre frequency and whether there is any need for special re-matching when the antenna is installed. The same antenna may be installed in many different sites and it may be more useful to rematch in situ. for radiation patterns, use MoM if the structure very small, otherwise GTD/UTD for both the earth and spacecraft antennas. For isolation use MoM if the dimensions are very small, otherwise use GTD/UTD and ray tracing.