hd.gif - 12Kb

These pages describe an experiment that was carried out over a period of around 15 months during 2002 and 2003 to record the effects of the atmosphere and environment upon signal transmission at 38 GHz. This work was funded by the Radiocommunications Agency of the United Kingdom. The purpose being to collect data that would be valuable in assessing and predicting atmospheric effects upon links at mm-wave frequencies. The work at St Andrews was carried out by Dr. Lesurf, Dr Pryde of the MM-Wave Group and assisted by Dr Robertson of the PIC. The line of sight chosen for this experiment is illustrated in the image below.

COlineofsight.gif - 201Kb

The path is 26 km long and crosses an estuary, an various other geographical features. The transmitter was placed on Criagowl Hill and produced a CW (i.e. unmodulated) signal at 38 GHz. The transmitted power was 100 mW using a feed horn with a gain of around 25 dBi.

The receiver system was placed upon the roof of the Physics building of St Andrews University in St Andrews, Fife. This system was a 3-port spatial interferometer whose operational schematic is shown below.

3port.gif - 17Kb

The system used was placed with its three input ports in a horizontal line approximately perpendicular to the nominal line of sight to the transmitter. On clear days the transmitter location could be seen from the roof where the receiver was placed. The spacing between the centres of the receiver antenna/ports was 2 metres (centre-side distance).

The signal processing used a set of IQ demodulators operated as PSD/multipliers to provide outputs that indicated both the magnitude of the received signals and the relative phases received between the central port and the side ports (left and right). Four output values were taken at regular intervals.

plane.gif - 15Kb

The above illustrates the situation where the arriving signal phasefront is plane. In such a case the difference in phase when we compare the center-port with each of the sides will have the same magnitude, and indicated the nominal direction of arrival of the signal. (Provided we know the port spacing and the wavelength.) If the signal passed through a uniform atmosphere, clear of any scattering objects or multipath, this will also indicate the direct line of sight to the transmitter.

curved.gif - 14Kb

In practice, the signal source is at a finite range, and the intervening environment may also affect the propagated phasefront. Hence we will find the arriving signal may have a curvature whose centre of curvature is not located at the source. By comparing the phase difference between the central port and the two sides, and differences can be used to indicate such effects, and any time variations in prism or lensing effects in the environment/atmosphere detected. Hence the system can monitor variations in link transmission in terms of both received signal magnitude and in terms of phasefront variations due to variable inhomogeniety in the air, etc.

The following pages give some examples of typical results obtained using this system.

Link to first page of results =>

Content and pages maintained by: Jim Lesurf (jcgl@st-and.ac.uk)
using HTMLEdit, TechWriter, and Compo, on a RISCOS machine.
University of St. Andrews, St Andrews, Fife KY16 9SS, Scotland.