Millimetrewave Antenna Research

Since the late 1980s the Group has built up a considerable reputation for work in millimetrewave antennas, both in measurements and theory.

The group was subcontracted to undertake the measurements for both the engineering and flight models of the Advanced Meteorological Sounding Unit B (AMSU-B), a mm-wave radiometer designed to monitor temperature and humidity in the upper atmosphere. These were completed successfully and the system is now in operation. With the increasing need to move up in frequency, there are many aspects of Microwave research that need to be explored and advanced.

The group was a major contributor to an ESTEC study entitled "Millimetre Wave Antenna Testing Techniques" concerned with the measurement of large spacecraft antennas designed to operate up to 1.5THz. The outcome of this project led to a further area of research and the development of a new millimetrewave antenna measurement system supported by EPSRC.

Moving to the small scale we believe that one of the next major advances in millimetre and sub-millimetre wave antennas will be the integration of arrays of active antennas on a single silicon, or GaAs, substrate. This technology is especially well suited for submillimetre/millimetre frequencies which are the most promising bands for future communication links, Wireless LAN, space scientific and earth observation instruments and other applications such as automotive collision avoidance radar. The group is presently active in this area aided by the support of ESTEC.

Since the late 1980's the Group has built up a considerable reputation for work in millimetrewave antennas. In 1997 the Group was a major contributor to an ESTEC study entitled Millimetre Wave Antenna Testing Techniques which sort to determine the best method of measuring spacecraft reflector antennas of up to 4 metres in diameter operating up to 1.5THz. This requirement is being lead by the next generation of space-based radiometers planned by ESA. The study concluded that the QMUL proposed Tri-reflector CATR with spherical main reflector was the most viable option. We currently have an EPSRC grant to build a CATR demonstrator with 1 metre diameter spherical main reflector and operating from 60 to 200GHz. An artist's impression of the system is shown opposite. The surface accuracy of the reflector directly determines the phase quality of the quiet zone. A displacement of the reflector surface by one hundredth of a wavelength (from the true paraboloid results in a phase change at the quiet zone of 7.2°. Since the generally accepted definition of a CATR quiet zone is one with less than ±0.5dB amplitude ripple and ±5° degrees of phase ripple, surface quality needs to be very high. The principle advantage of a spherical main reflector is that only one radius of curvature is needed for the entire reflector and so manufacturing costs can be significantly reduced over that of an offset paraboloid. The use of two shaped subreflectors serves two purposes. One to control the illumination of the main reflector by the feed so that illumination is uniform over a large proportion of the main reflector but falls rapidly as the edge is approached. This minimises edge diffraction and maximises the quiet zone size. The second function of the dual reflector feed system is to correct for the small difference in shape between an offset paraboloid and a large radius of curvature spherical reflector. The subreflector shapes have been designed using a Geometrical Optics (GO) reflector synthesis technique and the performance of the CATR predicted from Physical Optics and GTD using the TICRA GRASP 8 reflector analysis package which is used extensively within the Group.

Continuing with the measurement theme the Advanced Meteorological Sounding Unit B (AMSU-B) is a 5 channel mm-wave radiometer system developed for the TIROS-N and NOAA series of polar orbiting weather satellites. The radiometer has channels at 89GHz, 150GHz and three separate channels covering 183.3GHz. It is principally designed to monitor temperature and humidity profiles in the upper atmosphere and in addition uses the "window" at 89GHz to view the earth's surface temperature profile. The main contractor for the AMSU-B project is British Aerospace (Space Systems) Ltd. and to perform the antenna measurements on the breadboard, engineering and flight models of the instrument subcontracted QMUL. The antenna system for AMSU-B consists of a 90 degree dual offset main reflector with an elliptical aperture having a 200mm major axis. The subreflector feeds into a quasi-optical package that splits the incoming signal into the 5 channels. The beamwidth of the radiation pattern is about 0.9 degrees and it is required to measure the antenna radiation patterns with a dynamic range of better than 60dB. The measurements required ranged from measuring the radiation patterns of feeds up to frequencies of 191.3GHz (performed in our far-field anechoic chamber) to the more challenging of measuring the radiation patterns, beam efficiency and RF pointing, of the complete instrument. This being undertaken in our millimetrewave Compact Antenna Test Range details of which are described the Antenna Measurement Facilities section. The AMSU-B Engineering model and 3 flight models have now all been successfully measured and delivered to the customer. The launch of the first of 3 flight models took place in the summer of 1998. The figure above shows AMSU-B in the QMUL mmCATR undergoing tests. The UK Met. Office and NASA have now decided to upgrade the AMSU-B Engineering model to Flight model status so that it can be flown on a future mission. It is expected that the radiation characteristics of this upgraded AMSU-B will be measured at QMUL in spring 1999.

Moving to the small scale we believe that one of the next major advances in millimetre and sub-millimetre wave antennas will be the integration of arrays of active antennas on a single silicon, or GaAs, substrate. This technology is especially well suited for submillimetre/millimetre frequencies which are the most promising bands for future communication links, Wireless LAN, space scientific and earth observation instruments, and others applications such as automotive collision avoidance radar. However there is a lack of proper design tools for the theoretical analysis of these devices. Most of the electromagnetic analysis tools are not able to handle the non-linear behaviour of the solid state active elements, and those that do are not compatible with a good electromagnetic model for the radiating patch. To this end we have extended our FDTD software to model both non-linear devices and passive circuit and radiating elements. Schottky diodes and others active devices are included in the FDTD method as non-linear differential equations associated with a lumped element circuit model. Using this method, features such as the effect of feed/bias transmission lines, air bridges, and optimised matching between passive and active structures can be included in a design. The electromagnetic performance of the complete circuit, including the interaction of the active element with the passive part of the circuit, can thus be predicted. To experimentally verify our code a quasi optically fed ring mixer for operation at 90GHz is being manufactured on a silicon substrate, and is shown opposite. The type of antenna selected for the preliminary design is an annular slot antenna which produces highly symmetrical patterns at the resonant frequency of the ring ((=(d approx.), where the energy is mostly contained within the dielectric half-space rather than in the air. For these planar antennas to radiate efficiently they need to be coupled to a lens, constructed from the same material as the substrate, in order to form the beam and to prevent significant radiation loss into surface waves, see figure opposite We have developed a Physical Optics based lens model to take the FDTD predicted imbedded radiation pattern for the integrated antenna inside the substrate, and determine the far-field pattern. In collaboration with ESTEC we are also designing a 650GHz version of this mixer for use in imaging arrays in space based astronomy and radiometry. Our overall goal is to be able to take into account the effects of mutual coupling when arrays of active elements are to be designed.