Antenna Metrology Theory and Applications
As well as having high quality antenna measurement facilities, the group undertakes research into microwave and millimetre-wave antenna metrology. This research is focused around the major antenna measurement facilities of the Antenna Measurement Laboratory, consisting of:
- two compact antenna test range operating from 5 GHz to 300 GHz;
- fully-screened anechoic chamber for mobile communications antenna applications (750 MHz to 5 GHz);
- NSI Planar near-field range operating to 100 GHz;
- a 9 m x 3 m x 3 m anechoic chamber for feed measurements and general measurement use.
Research in this area is illustrated by the following activities.
Mathematical Absorber Reflection Suppression (MARS)
This mathematical post-processing technique can be deployed to antenna pattern data taken using a far-field or CATR facility using only a single great circle cut to efficiently correct far-field, frequency domain data. Reflections in a CATR can often be the largest source of measurement error within the error budget of a given facility. This mode orthogonalisation and filtering technique requires only a minimum amount of information about the AUT and measurement geometry, and is able to suppress reflections in a direct far-field one-dimensional antenna range measurement.
Contrary to usual antenna measurement practice, the MARS technique deliberately displaces the AUT from the centre of rotation on the turntable. This has the effect of making the differences in the illuminating field far more pronounced than would otherwise be the case, and it is exactly this greater differentiation that makes their identification and subsequent extraction viable. Once the far-field great circle pattern cut has been acquired in this displaced condition, the AUT pattern can be mathematically translated back to the origin of the measurement co-ordinate system and from this, the equivalent cylindrical mode coefficients (CMCs) can be deduced from the measured fields.
The cylindrical mode coefficients for the, now ideally centrally located, AUT are then recovered, so any mode representing fields outside the ideal conceptual minimum MRE can be filtered out, thereby removing contributions that are not associated with the AUT. As these transforms and their inverse operations can be evaluated with the fast Fourier transform (FFT) this makes the F-MARS algorithm very efficient in terms of computational effort.
To demonstrate the method, repeat measurements were taken of the far-field great circle azimuth cut of a medium gain X-band corrugated horn in the QMUL CATR (see figure 1 below), where a single parametric change was introduced into the measurement. This change consisted of introducing a 0.6 m by 0.6 m flat reflecting plate into the chamber that was located in the same horizontal plane as the AUT, and was chosen as it constituted a worst-case configuration, as the specular reflection of the main beam of the corrugated horn directly illuminated the CATR reflector. Figure 2 below shows the great-circle far-field co-polar amplitude and phase patterns of the AUT, where the reference trace was taken without the reflecting plate. Conversely, the measured trace was taken with the reflecting plate installed within the chamber and clearly shows the effects of the additional scattering as an additional large amplitude side-lobe at around 50°.
Figure 1: corrugated horn in the CATR facility.
Figure 2: Power and phase patterns for the corrugated horn.
From inspection of the F-MARS processed patterns, it can be seen that the effects of the spurious scatterer have been effectively suppressed in both the amplitude and phase plots. Details can be found at:
- Gregson, S.F.; Dupuy, J.; Parini, C.G.; Newell, A.C.; Hindman, G.E., “Application of Mathematical Absorber Reflection Suppression to far-field antenna measurements”, Loughborough Antennas and Propagation Conference (LAPC), 2011, Page(s): 1–4, DOI: 10.1109/LAPC.2011.6114016.
Poly-Planar Near-Field Measurements
Conventional single cut planar near-field measurements are restricted to high-gain antennas and to radiation pattern angles covering much less than the forward hemisphere, due to the truncation of the extent of the planar surface that the radiation fields are measured. We have developed techniques to combine several near-field planar scans to calculate the far-field radiation pattern of a given AUT, to much wider angles, and indeed full spherical coverage.
As an example, we demonstrate a poly-planar technique in which the AUT is mounted in one of six discrete orientations, these orientations being the positive and negative directions of the three orthogonal positional axes. This cubic geometry is the most demanding case, as the orthogonality between adjacent partial scans would constitute a worst case scenario whilst being relative simple to realise. Using our NSI planar scanner and the corrugated horn (shown in Figure 3) as the AUT, all six surfaces of the cube surrounding the AUT were sampled by performing the following rotations from position:
- AUT nominally aligned to axes of range;
- positive rotation of 90° about y-axis;
- negative rotation of 90° about y-axis;
- positive rotation of 180° about y-axis;
- positive rotation of 90° about z-axis, followed by a positive rotation of 90° about new x-axis;
- positive rotation of 90° about z-axis, followed by a negative rotation of 90° about new x-axis.
The requirement for the inclusion of the back plane follows from the requirement to perform the pattern integration over a closed surface. When the partial scans were completed each of the six partial scans were processed using our transformation algorithm. The y-polarised electric near-field can be found plotted below and as expected, the fields at the intersection between adjacent partial scans are continuous.
Figure 3: Corrugated horn and the NSI near-field scanner.
These cubic data sets were subsequently transformed to the far-field and resolved onto a Ludwig III polarisation basis. An example pattern is shown below, where the blue trace represents patterns obtained from the poly-planar technique, whilst the red trace denotes results obtained from our compact antenna test range (CATR). The high frequency oscillatory behaviour evident within the azimuth cut of the CATR at wide angles is a result of multiple reflections within the facility and should be ignored.
Details can be found in chapter 9 of:
Figure 4: Y-polarised electric field for the six surfaces.
Figure 5: Comparison of far-field patterns from the Poly-Planar Near-field and CATR measurements.
Highlights and Research Outcomes
Along with colleagues Stuart Gregson (Near-field Systems, Inc. [NSI]) and John McCormick (SELEX ES Edinburgh), Prof. Clive Parini published a comprehensive text covering the theory and practice of planar Near-field antenna measurement:
In summer 2014, along with colleagues Stuart Gregson and Daniël Janse van Rensburg (Near-field Systems, Inc. [NSI]) and John McCormick (SELEX ES Edinburgh), Prof. Clive Parini will be publishing the research text “Theory And Practice of Modern Antenna Range Measurements” in the IET Electromagnetics Series.
This single volume will provide a comprehensive introduction and explanation of both the theory and practice of all forms of modern antenna measurements, from their most basic postulates and assumptions, to the intricate details of their application in various demanding modern measurement scenarios.