Telecommunications Research

The Group has an interest in the application of antennas in the general area of Telecommunications. Recent areas of interest were in the area of Geodesic space-frame radomes involving work to predict the effects of struts and Earth station clutter looking at the problems associated with locating many antennas within a small footprint. Until fairly recently satellite earth-station antennas have been located in open sites and so clutter associated with objects close to the antenna did not exist. However with the rapid diversification of satellite communication services have come the need for more ground stations and many of these (for example VSAT terminals and TV uplink stations) need to be located in compact urban sites. In addition to these constraints, once an urban site is chosen there are strong economic and urban planning pressures to place as many earth-station antennas on that site as possible. The most recent work in the area of Telecommunications Research has been supported by TLS and BT.

Geodesic space-frame radomes are used extensively throughout the world to provide weather protection and a stable environment for a variety of large microwave reflector antenna systems. The original metal strut space-frame radomes have been largely superseded by dielectric space frame radomes, which incorporate membrane and struts in a single moulding, see opposite. When these are bolted together on site the space-frame is formed from the struts at the edges of the panels. The scattering and loss from the dielectric struts dominate distortion of the enclosed antennas radiation pattern. As the specifications of radome enclosed antennas get ever tighter it is becoming increasingly difficult for radome manufacturers to be able to predict, or indeed demonstrate, that these specifications can be met. We have applied FDTD to the problem of dielectric strut scattering. Once the scattering characteristics of a single strut is predicted the electromagnetic performance of a complete geodesic radome can be determined. To reduce the scattering cross section of a strut several parallel wires can be placed along the strut length to provide tuning and so reduce the amount of forward-scattered power. FD-TD is highly suited to the modelling of this effect and the field inside the strut with and without tuning shows how the wires perform the tuning effect, see below.

Field With and Without Tuning

Several scenarios with a variable number of wires and variable wire positioning have been studied. We have found that the inclusion in the model of the bolts that hold the two parts of the strut together is essential for an accurate prediction of performance. An experimental near-field measurement system to measure the forward scattering of test struts has been developed and a typical result is shown overleaf.

Having established that our single strut models are accurate, we are now modelling a complete radome so the electromagnetic degradation resulting from placing a given antenna inside a radome can be determined. We intend to verify this model by building a planar network of struts and perform far-field pattern measurements of a high gain test antenna with and without this network of struts placed across the antennas aperture. This work will be undertaken at 90GHz using our mmCATR facility.

Until fairly recently satellite earth-station antennas have been located in open sites and so clutter associated with objects close to the antenna did not exist. However with the rapid diversification of satellite communication services have come the need for more ground stations and many of these (for example VSAT terminals and TV uplink stations) need to be located in compact urban sites. In addition once an urban site is chosen there is strong economic and urban planning pressure to place as many earth-station antennas on that site as possible. This inevitably means that clutter resulting from nearby buildings as well as other earth-stations can become a serious problem. Near-field clutter can, in some cases result in the on site performance of the earth-station not meeting the satellite operators specification. Clearly this situation needs to be avoided, preferably at the earth-station antennas design stage. What is required is an electromagnetic model of the earth-stations radiation characteristics in the presence of near-field scatterers. This would then enable the earth-station operator to determine the performance of a particular manufacturers earth-station at the intended site. It would permit the operator to determine the best location for the earth-station on the site as well as the effects on the performance of any other earth-station antennas on that same site.

Ideally the electromagnetic model of the near field clutter environment of an earth-station antenna would take into account every possible near-field source of scattering, including not only all the local buildings but all the varied metallic furniture located on the roof and sides of buildings, such as railings and ventilation covers etc. Clearly this would result in a massively complex electromagnetic software package requiring large amounts of data input. To date, little if any work on the subject of earth-stations operating in the presence of near-field clutter has been published in the open literature and so one of the tasks for this study has been to determine how far simple models of nearby scatters can be taken in providing acceptable performance prediction.

Experimental verification of the theory, for example, the effects of site changes such as different earth-station location, additional buildings etc., is seen as being vital to the successful outcome of the study. A scale model of a typical earth-station antenna and nearby scattering objects has been constructed and experimentally measured at 95GHz, a typical result is shown opposite. The use of 95GHz means that the model requires plane wave illumination of order a cubic metre which is easily provided by our millimetrewave CATR. On the theoretical side we are currently determining how well a commercial antenna design package (TICRA's GRASP 8) can model this basic blocking process.