Within the framework of the GOCE Earth Explorer mission, which is scheduled for launch in 2007 by the European Space Agency (ESA), a high accuracy Satellite-to-Satellite Tracking Instrument (SSTI) with two hemispherical-coverage receiving antennas will be used for real time navigation and precise orbit determination. The two GPS antennas are directly mounted on the edge of one of the metallic solar array wings of the spacecraft.

The GOCE requirements on the calibration procedure for the GPS antenna are demanding. From an elevation angle of 15° up to the boresight antenna direction, the maximum error on the phase center position should not exceed ± 1.8 mm for the L1 and ± 2.4 mm for the L2 carrier phase signal. The challenge is further increased when the near field multipath effects generated by the satellite modify the antenna phase pattern.

The characterization of the antenna phase variations has been performed by RYMSA [1] using numerical simulations and measurements on the stand-alone GPS antenna in an anechoic chamber. In parallel, ESA has first simulated the stand-alone antenna and then with spacecraft parts to include the interactions with the spacecraft body [2]. Finally, in order to have a completely independent set of measurements, the antenna has been characterised outdoor directly receiving live signals from GPS satellites. For this test campaign, the Automated Absolute Field Calibration Technique developed by the Institut für Erdmessung (IfE) and Geo++ has been used [3].

The calibration consists in mounting the antenna on a moving robot installed on the roof of a building. Then, rapidly changing the antenna orientation through the robot, absolute phase center offsets and variations are determined on submillimeter level while unwanted multipath interferences caused by reflecting surfaces of the site environment (roofs, walls, etc.) are completely eliminated. Furthermore, the influence of the GOCE solar panel on phase and amplitude is estimated by calibrating the GPS antennas not only in stand-alone mode but also together with a representative cut-out of the spacecraft’s solar wing which was mounted beneath the GPS antenna [4]. The differences of both set-ups confirm and identify the near field influences caused by the wing: these contributions must be taken into account in order to fully meet the accuracy requirements of the GOCE mission.

Measurements collected at Rymsa and IfE are now used by ESA to validate the simulation model including the GPS antenna plus a section of the GOCE wing. As no tests measurement are planned at full spacecraft level, the validated (antenna + wing) model will be embedded in a larger model of the complete GOCE spacecraft body towards determining the ultimate antenna performance at system level.

[1] M. Manzanares, A. Torre, "Technical Note of GOCE GPS Antenna Phase Center", GO-TN-RYM-0057, March 2006. [2] G. Toso, D. Maeusli, D. Polimeni, P. Nepa, "Analysis of GOCE SSTI/GPS Antenna Performance", 28th ESA Antenna Workshop 2005, Noordwijk, 2005. [3] G. Wübbena, M. Schmitz, F. Menge, V. Böder, G. Seeber, "Automated Absolute Field Calibration of GPS Antennas in Real-Time", Proceed. ION GPS-2000, pp. 2514–2522, 2000. [4] F. Dilßner. G. Seeber, M. Schmitz, G. Wübbena, G. Toso, D. Maeusli, "Characterisation of GOCE SSTI antennas", German geodesy magazin "Zeitschrift für Vermessungswesen (ZfV)", 2006.

This work presents the study results of the X/Ka High Temperature High Gain Antenna ( HTHGA) sub-system being developed by Alcatel Alenia Space and Konsberg Space and Defence, in the frame of the ESA BepiColombo mission. Bepi Colombo is an Interdisciplinary Cornerstone Mission to the planet Mercury, in collaboration between ESA and ISAS/JAXA of Japan.

The HTHGA configuration is driven by the following key specific mission driven parameters:↓ ←

„h High range operating temperature [-180°C,+400°C],
„h Dual frequency band operation ( X/Ka),
„h Wide steering range ( more than hemispherical ),
„h Low mass,
„h High phase stability,
„h Limited envelope constrained by the S/C accommodation capability,
„h High pointing stability due to ka band operation,

The extreme environmental conditions induced by Mercury's proximity to the Sun (up to 14.500 w/m2 direct solar fluxes, up to 5000 W/m2 infrared flux and up to 1200 W/m2 albedo shine form the planet surface), have dictated the need for a specific high temperature development of the key antenna subsystems:

Main and Sub-reflector ( antenna reflector assembly, ARA) Dual band feed chain (radio frequency chain, RFC) Pointing mechanism and of its integrated RF Rotary Joint. ( antenna pointing assembly, APA )

Innovative solutions have been deemed necessary, in the selection of the HTHGA configuration, whose architecture, components, material and processes must be compatible to the harsh high temperature environment. The proposed antenna optics is a shaped axially symmetric ellipse dual reflector geometry (ADE) fed by a dual band, dual depth, self diplexed feed chain operating at X and Ka. The main advantage of this configuration, is the consistent RF improvement of the antenna efficiency due to losses reduction of the sub-reflector spill over and blockage and the capability to reach even with small diameters (24 wavelengths) an efficiency higher of 70 %.

The feed architecture, that avoids any dielectrics within the horn, relies on a flight proven design flown in the CASSINI HGA ( mission to Saturn ) and permits to fulfill the high phase stability Vs temperature . Due to the high temperature, both main and sub reflectors are realized with a process that relies on a particular ceramic based technology ( Cesic® ) , that allows a single piece realization of the reflector(s) compatible with a dedicated coating for RF losses reduction. Regarding the mechanism, innovative solutions have been deemed necessary at system architecture level, in the design of the mechanisms rotary joints, particularly in the selection of high temperature compatible materials and processes.

Saab Ericsson Space has recently developed a K/Ka-band antenna family. These antennas are used for satellite applications such as telemetry, command and beacons.

This paper presents the design, bread board work and flight model performance of one of its members, a dual mode K/Ka-band biconical type antenna with toroidal coverage.

The structure basically consists of two radiating elements stacked in height with the receiving Ka-band radiator positioned on top of the transmitting K-band radiator. The radiating elements consist of a circular waveguide with 7 inclined resonant slots around its circumference. The circular polarisation is obtained with an external polariser, consisting of a parallel plate region. In the parallel plate the vertical components phase varies like a free space TEM mode, while the horizontal components phase varies as a TE10 mode. An equal amplitude distribution, obtained by proper inclination of the slots, in combination with 90° phase difference will produce the desired circular polarisation. Outside the polariser corrugations give the desired radiation pattern shape and suppress unwanted lobes.

Each element is fed with a septum polarizer with two waveguide ports. A septum polarizer gives left or right hand circular polarisation depending on which of the two ports is used. The two ports give opposite circular polarisation in the waveguide, but in the radiation pattern only the azimuth phase variation differs. Thus we get two identical power patterns from the two ports without any gain loss.

Two cables from the Ka-band polarizer needed to be routed down to waveguide interfaces at the base plate without affecting the radiation pattern performance of the K-band radiator. Also the K-band radiator had to be robust for structural reasons. A routing of cables on the outside of the K-band radiator caused very high disturbance, even with a routing between the slots. These high disturbances create high omni variation and low cross polar discrimination. We selected a design where the cables are embedded within the K-band radiator wall.

For easier bread board handling and tuning all design work was initially done at X-band, followed by a final scaling to the proper frequency band. The development started with the design of the Ka-band element. Hand calculations followed by simulations and optimization with the method of moment code AKBOR resulted in manufacturing of a bread board with replaceable centre part for radiation pattern tuning possibilities.
Based the Ka-band design work with the K-band radiator started. For the K-band waveguide a thicker radiator wall was needed. This reduces the bandwidth and affects the resonant frequency of the slots. Optimization of the slots was performed using the simulation tool Ansoft HFSS and work with the bread board.

Research on lightweight antennas is taking momentum as there are many potential uses for that kind of antennas. For these reasons exploration of potential advantages offered by inflatable antennas is attracting increasing attention. Such antennas are convenient for storage and for transportations. They can be used as spare antennas, rescue services, low-cost imaging and remote sensing. primary disadvantages are shorter durability and vulnerability to failure. Rigidization with foam is expected to improve resistance to wind and vibrations. In last two decades, studies on inflatable antennas were focused on reflector antennas [1], [2].

In our research we have undertook efforts to utilize extensive experiences gained in the area of lightweight broadband passive microwave circuits making use of printed circuits. For this reason, our studies are concentrated on inflatable antenna arrays. Two categories are considered: planar and cylindrical arrays. Cylindrical shapes are feasible to be formed with tough tolerances by inflatable membranes. Preliminary review concludes that tolerances are worse for planar structures. To develop antenna arrays made with inflatable technique, we also have analyzed spatial power combining and that kind of beamforming eliminates severe problems with printed circuits made in air suspended membranes. New problem associated with spatial power combining are integration of distributed amplifiers with patches on flexible membranes.

Prior to development of full scale laboratory models, it is indispensable to perform thorough tolerance analysis. Design of RF circuits and antenna patches on usually less than 75 mm thick membranes must address how these radio components perform on somehow disturbed surfaces. We specified that the biggest interest will raise medium gain inflatable antenna arrays, and when the technology would mature applications their design may evolve onto high gain applications. In tolerance analysis we have applied Monte Carlo method.

Fig. 1. Example of tolerance analysis: Computed return loss of the S-band circular patch which undergoes conical distortions. Magnitude of distortions is ranging 7 in height above the ground. The patch is for use in planar inflatable antennas.

As a feasibility study, following the successful design of an X/Ka band feed system for a military satellite earth terminal [1], the CSIRO, in collaboration with BAE systems Australia, has looked at the possibility of extending the capability of the original feed system to cater for potential military maritime applications. The CSIRO X/Ka-band feed system is made of an optimised coaxial horn and a compact, complex feed-system which makes it ideal for applications where space is at a premium.

Our design approach was to look at an unshaped Cassegrain dual-reflector system with the following strict constraints:

More and more communication systems operating via spaceborne platforms are accessible with compact or home equipment. A good example are gaining popularity radioamateur space services. Widespreading of digital data transmissions and introducing of digital television have led to use of low microwave bands by radioamateur services. We have designed, manufactured and tested lightweight microstrip patches for use in space, primary for telecommanding and telemetry of small spacecraft. For the needs of emerging radioamateur digital systems we designed, manufactured and tested a dual band antenna made with microstrip technique. Most remarkable features of the antenna are:

(i) low profile - overall height is 21.5 mm,
(ii) dual band operation: 1260-70 and 2400-50 MHz,
(iii) exceptionally low weight in Earth conditions - 220 g,
(iv) excellent circular polarization within each beam, in some cut-planes axial ratio is less than 1 dB for 80° elevation angles,
(v) use of proprietary magnetic absorbing materials in match loads of two directional coupler, so there is no need for use soldered junctions,
(vi) outstanding thermal, vacuum and mechanical properties.

Model of the presented antenna is shown in photos in Fig. 1. As shown with measurements, the developed antenna model features outstanding electrical parameters.

Fig. 1. Top and bottom view of the developed L/S band microstrip antenna for use onboard spacecraft. A profile of the antenna is not planar and conforms to a cylindrical shape.

Number of minispacecraft missions is expected to raise significantly over next years. Launched in clusters or as auxiliary payload, minispacecraft must complain with severe outline constraints. Use of patch antennas, little extruding from walls, make spacecraft simpler and development time shorter. Actual antenna placements must compromise with some higher prioritizated objectives (structural factors, instrument clearance). Recently demonstrated designs of patch antennas have feature excellent circular polarization within a wide angle of the beam. Microstrip elements mounted on walls have remarkable different pattern shape from helix antennas located on long booms.

It is essential for low rate telemetry and telecommand (TTC) communication, that it works extremely reliable. Therefore, extensive studies are required to determine optimum placements for patch elements on walls of spacecraft. Our research have been concentrated on this issue. Simulations have been carried with Method of Moment (Concept of the University of Hamburg) and with Ray Tracing Method. Experimental work have been run with four electromagnetic models of spacecraft. The aim of studies was to find such placement of RHCP patches that minimizes number of nulls and other adverse effects in combined radiation pattern. Usually no more than four patch elements are used onboard. Calculated and measured radiation patterns for the patch placement considered as one of most favorable for a minispacecraft is presented in Fig. 1. There is a close agreement between simulated and measured patterns. Method of Moment can effectively handle simulations of a minispacecraft on a single modern PC analysis. The Ray Tracing Method calls for much simpler computations, however, results we have got are remarkable more generic than the results we have obtained with the MoM. In the course of our analysis we investigated failure modes, such as incomplete solar panel deployment. Results of the carried out research can be generalized onto other emerging application areas (e.g. vehicles for intelligent transportation systems).

Fig. 1. Measured and calculated (Method of Moments) radiation pattern for the cluster of three patch antennas placed on a 'washing machine' size spacecraft with improperly deployed pair of solar panels.

The fourth generation of INMARSAT satellites, provides a good example of an advanced high capacity, high gain fixed beam/cell mobile communication system, [1]. The L band Mobile satellite system utilizes a 9m deployable reflector fed via a 120 element feed array. A digital signal processor (DSP) is used to produce many different beam types that cover the earth and provide different user services. The most demanding of these is the Personal Mobile Communication (PMC) service. This uses ~200 high gain narrow spot beams to cover the earth. The beams provide spatial frequency re-use, so co-coloured beams are synthesized to have a high degree of isolation between them. The earth is covered by contiguous, (nominally hexagonal, as viewed from the satellite), cells that are fixed with respect to geographic locations on the earth. The beams are changed throughout the day so that they remain aligned to these fixed geographic cells even though the satellite orbit is slightly inclined. Users are assigned to a beam that aligns with the geographic cell they occupy.

However, performance requirements for mobile satellite communication systems are steadily increasing, both in terms of higher antenna gain and frequency re-use capacity. In order to provide the edge of cell directivity and the beam roll-off rate, to achieve inter-beam isolation, the beam size will need to decrease. So, many more beam/cells will be required to cover a given geographical area. This will lead to far larger reflectors and feed arrays with many more radiating elements. For conventional fixed beam/cell systems the antenna pointing error is likely to become a dominant loss factor in the link budget. An alternative is to abandon the fixed beam/cell scheme, and form beams that are optimized to individual user locations. This paper uses simulations of a 'dynamic beam synthesis' process to highlight the potential benefits of this approach.

The INMARSAT 4 type of antenna system architecture is used as a baseline to compare the fixed beam/cell system to the alternative of dynamically synthesizing the beams making use of the individual user locations. The simulations indicate the potential benefit of 'dynamic beam synthesis'. For example in one simulation the isolation between users shows improvements of 10 to 15dB. This improvement could be used to improve quality of service, increase the frequency re-use capacity or traded-off for increased user gain. Alternative trade-off examples will be presented, to illustrate the potential benefits it offers, to future mobile satellite communication systems.

[1] R.F.E. Guy, C.B. Wyllie, J.R. Brain, 'Synthesis of the Inmarsat 4 Multibeam Mobile Antenna', 12th International Conference on Antennas and Propagation, ICAP 2003 31st March-3rd April 2003, University of Exeter, UK, p90-93.

We present results for the optimization of helical antennas with a truncated-cone reflector. Starting from a helical antenna with a circular or a square flat reflector, it is found that the truncated-cone reflector (Fig. 1) can increase the antenna gain [1]. The dimensions of the optimal antenna with the reflector are found to be slightly different than those for an infinite or finite flat plate below the antenna.

For the numerical simulations, we use three families of helical antennas with constant heights of 1, 2, and 5 wavelengths. For each antenna, eight different reflector heights are considered: 0.25, 0.50, 0.75, 1.00, 1.25, 1.50, 1.75, and 2.00 wavelengths. For each pair of antenna and reflector heights, four parameters are varied (optimized) to find the maximal gain of the antenna in the axial direction at the operating frequency. Optimized parameters are: antenna radius, number of turns (or equivalently antenna pitch), truncated-cone lower radius, and truncated-cone upper radius. A model of an optimized antenna is shown in Fig. 1. For the antenna simulation we use programs from [2, 3]. Optimizations are carried out using particle swarm optimization algorithm [4] implemented in WIPL-D Optimizer. One optimization cycle consists of 300 iterations (EM solver calls). Each optimization is carried out 10 times with random seeded initial solutions to maximize the possibility of finding the best solution in the optimization space.

Our optimization results are optimal helical antenna parameters and cone reflector radii, expressed as functions of antenna and reflector heights. It is found that the gain of the antenna increases significantly when the reflector height is increased (while the antenna height is constant). The gain increase is about 1 dBi for 0.25 wavelengths reflector height and up to around 5 dBi for 2.00 wavelengths reflector height (compared to the antenna of the same height with flat reflector).

To verify the results found by the simulations, one helical antenna with the truncated-cone reflector is assembled and measured. The measured and computed results match very well.

Fig. 1. Model of optimized helical antenna with truncated-cone reflector.

REFERENCES

[1] A.R. Djordjevic, A.G. Zajic, and M.M. Ilic "Enhancing the Gain of Helical Antennas by Shaping the Ground Conductor," accepted for publication in Antennas and Wireless Propagation Letters.

[2] B.M. Kolundzija et al., WIPL-D Pro. v5.1, WIPL-D, 2005.

[3] A.R. Djordjevic et al., AWAS for Windows Analysis of Wire Antennas and Scatterers, Software and User's Manual, Artech House, Boston, 2002.

[4] J. Robinson and Y. Rahmat-Samii, "Particle Swarm Optimization in Electromagnetics," IEEE Trans. on Antennas and Propagation, Vol. 52, No. 2. Feb. 2004, pp. 397-407.

By loading properly arranged slots in an equilateral-triangular microstrip patch, a dual-frequency and/or broadband operations of a single-feed triangular microstrip antenna are presented. For dual-frequency operation, the proposed design is achieved by loading two pair of narrow slots in the triangular patch, symmetrically embedded close to the both side edges of the patch. The obtained two operating frequencies are of same polarization planes and by varying the positions and lengths of the inserted slots, a tunable frequency ratio of the two frequencies ranging from about 1.16 to 2.06 is obtained. Furthermore, it is found that by protruding a narrow slot out of the embedded slots close to the side edges, broadband operation of the triangular microstrip antenna near its fundamental resonant mode can be achieved. Results show that the antenna bandwidth of the proposed broadband design can be greater than 2.6 times that of a conventional triangular microstrip antenna. Details of the proposed dual-frequency and broadband designs are described and simulated results of near- and far-field parameters are presented and iscussed.

Fig. 1. Surface current distribution on the slot-loaded equilateral-triangular microstrip antenna (a), near E-field distribution of ant on handset board (b) insertion loss versus frequency (c)

The proposed design in Fig. 1(a) is denoted here as a simple slot-loaded triangular microstrip antenna The CP radiation is achieved by cutting the section at tip of H1=6.9 mm. The triangular patch is assumed to be equilateral with a side length of L=71 mm and is printed on a substrate of thickness h=1.6 mm and relative permittivity εr=4.4. The pair of slots, having dimensions l1=39 mm, w1=1 mm, are placed in parallel to the side edges of the triangular patch, with a small distance w1 away from the side edges. With the presence of slot 1, the surface current distribution of the TM20 mode is modified such that the small dip in the broadside direction of the radiation pattern is removed. A single probe feed for the excitation of the TM10 and TM20 modes can be easily located at w2=66 mm from the center line of triangle and H2=6.9 mm from the bottom edge of it. To begin with, simulated results obtained using IE3D™ and FIDELITY of Zeland Software Ltd.

for the antenna design are first presented in Fig. 1(a) in which typical results of the excited patch surface currents distributions of TM10 and TM20 modes are shown. It should first be noted that the loading of slot 1 has small effects on the TM10 mode and, on the other hand, can make the distribution of patch surface current density of TM20 mode more uniformly distributed in the center portion of the triangular patch to make the radiation patterns of the TM20 mode more close to that of the TM10 mode. Fig. 1c shows the simulated results of the return loss for the antenna design of Fig. 1(a). Results show satisfactory agreement for the present dual-frequency design of a simple slot-loaded triangular microstrip antenna, at GPS bands L1=1.575 and L2=1.227 GHz, respectively. The two operating frequencies of the proposed dual-frequency design are of same polarization planes and also have similar radiation patterns.

Circularly polarized (CP) microstrip antennas are currently receiving much attention for mobile satellite communication [1]. Several techniques have been explored to design CP microstrip patches, which use either special feed networks [1], [2] or some physical deformation in the patch shape [3]. The former employs complex excitation architecture and the latter one is relatively simple but suffers from the inherent limitation of narrow impedance and axial ratio bandwidths [3].

In this paper, we propose a new single-feed wideband CP antenna design employing a suspended substrate probe-fed circular patch loaded with a couple of slots as shown in Fig. 1. An antenna operating over comparatively a very wide axial ratio bandwidth (< 3 dB) covering a GPS band is demonstrated using simulation and experimental results. Excellent overlap of matching and axial ratio bandwidths is revealed for an optimum design as shown from the results in Fig.2. As much as 13% matching bandwidth (s11<-10 dB) covering nearly 8% axial ratio bandwidth (< 3 dB) is demonstrated with good right-handed CP (RHCP) to left-handed CP (LHCP) isolation. Detailed description of the new antenna along with some representative simulation and experimental results will be provided.

References
[1] D. Sevenpiper, et al., "Low-profile cavity-backed crossed slot antenna with a single-probe feed design for 2.34-GHz satellite radio applications", IEEE Trans. Antennas Propagat., Vol. 52, No.3, pp. 873-879, 2003.
[2] H. Kim, et al., "A single-feeding circularly polarized microstrip antenna with the effect of hybrid feeding", IEEE Antennas Wireless Propagat. Lett. Vol. 2, pp.74-77, 2003.
[3] K. P. Young and K. L. Wong, "Dual-band circularly polarized square microstrip antenna", IEEE Trans. Antennas Propagat., Vol. 49, No.3, pp. 377-382, 2001.

The problem of the Polarization of the Antenna is a very critical point in the world of Telecommunication. For example the Data Link between Earth and Satellite is generally operated with Antennas working in Linear Polarization. In many cases the communication with the Satellite is difficult to establish because the Polarization of the ground Antenna is not always aligned with the Polarization of the observed Satellite. An antenna with Variable Polarization enables to use only one Antenna for Telecommunication with Satellite whatever is the position between the ground Antenna and the Satellite.

In this article we present a specific application of the Patent "High Gain Antenna for Microwave Frequencies" assigned in ELTA. This Patent concerns Circular Polarization. But it is possible to use the element of the Patent which is planed for Circular Polarization in order to obtain Variable Linear Polarization. The principle is based on the combination of Left Circular Polarization (LCP) and Right Circular Polarization (RCP) to get Linear Polarization. If we apply a delay in summation of LCP and RCP the Linear Polarization can be changed very simply. With this method a single Antenna can generate Polarization adapted to different positions between ground and Satellite.

The application of the Patent consists of sharing the aperture of the Antenna with superposition of two Printed Circuits which do not interferes too much. Each Printed Circuit generates a Circular Polarization in different senses. The complete antenna has 2 connectors, one connector for LCP and the other connector for RCP. The Linear Polarization is obtained by recombination of the LCP and RCP Antenna through a Power Divider by 2.

When the summation of the 2 Ports is in phase we obtain a Vertical Polarization. If the summation between the 2 Ports has a phase shift of 90° (Phase Shifter or Delay Line) the Polarization is Horizontal. All the intermediates Linear Polarizations by example 45° can be obtained by tuning the Phase Shifter or the length of the Delay Line. In the complete article, we will present the measured performances of the Array described in figure 1 and figure 2.

Fig 1: Drawing of the LCP Antenna on the left side and RCP Antenna on the right side Fig 2: Drawing of superposition of the 2 prints
(LCP Antenna placed under the RCP Antenna)

This paper presents the main results of a study, for which the main objectives were to revisit at system level the future missions requesting large deployable antennas, to define and justify for each mission an associated antenna architecture, with specific attention on the deployable concept.

A mobile telecommunication mission was considered, with multiple spots (> 150 spots) and high gain (> 46dBi). Two antenna architectures were considered, associated with a 25m deployable reflector : The Focal Array Fed Reflector (FAFR) has its focal array arranged at the focal plane. An imaging array also named Defocused Focal Array Fed Reflector (DFAFR) has its array significantly moved away from the focus of the reflector. These architectures were compared in terms of performances and complexity. From the definition of the reflector, the pillow effect was predicted, using a model involving linear and nonlinear analyses. Its effect on RF performances was assessed, as well as its capabilities to be compensated by processing.

For Earth Observation Missions, an X band LEO SAR instrument was selected. The typical antenna size is around 4*2 m2. Deployable structures technologies can be used to relax the accommodation constraints imposed by low cost launch systems. Two accommodation cases were selected to assess the potential interest of deployable structures : A single satellite launched by a small launcher, Three satellites launched by a medium size launcher. A number of novel deployable reflector concepts were considered for these cases. A number of criteria, that includes Surface Accuracy, Mechanical Features, Technological Maturity were considered for these various candidates. The best concepts appear to be the Hinged Panel Reflector or Reflectarray, and the Spring Back Reflector.

For Space Science Missions, the Space Very Long Baseline Interferometry (VLBI) was selected. This astronomical interferometric technique searches for interference fringes from RF signals. For Terrestrial VLBI observations, the interferometer baseline length is limited by the diameter of the Earth. These limitations are removed when one or more interferometer elements are placed in space. In long baselines, the fringes to be detected are weak. Large antennas are therefore needed to improve the signal to noise ratio. A mission was defined that places two spacecraft in orbit simultaneously. Four bands of observation were defined (8, 22, 43, 86 GHz). The radiating aperture size was set to 20m, A sensitivity study was conducted to derive the best antenna profile which optimises the aperture efficiency.

Novel deployable concepts, such as Reflectarrays, dielectric lenses and transmit arrays, were also analysed. Such antennas have interesting features for large deployable antennas. They can have either a full planar shape or a facetted parabolic shape. This may then result in a simpler deployment scheme than for parabolic reflectors. However, the reflective surface to deploy is more complex than a mesh reflective surface. The objective of this task was then to consider whether these periodic-structure antennas can be deployed and accurately tensioned for large deployable apertures, which are challenging applications for more conventional designs. Typical applications were 25m antennas in S band (telecommunication mission), and 5m X band antenna (interplanetary mission).

Recent experiments, both in space and in the laboratory, have required thorough understanding of the behavior of wire antennas in homogeneous and inhomogeneous magnetized plasmas. The interest in the subject has substantially been stimulated by experimental observations of guided electromagnetic waves on a conducting ionospheric tether [1]. The results of these experiments are found to be in good agreement with the theoretical predictions for wire-guided electromagnetic modes [2, 3], which allows one to use the approach of [2, 3] for a detailed analysis of the characteristics of wire antennas in a magnetoplasma.

This paper discusses the current distribution and the input impedance of a perfectly conducting, arbitrary-length cylindrical antenna insulated from the surrounding cold magnetoplasma by a coaxial cylindrical sheath for the case where the antenna is aligned with an external magnetic field and is excited by a given time-harmonic slice voltage generator. The sheath characteristics are chosen so as to model either an ion sheath around a cylindrical conductor in a plasma or the dielectric insulation used in some experiments. The plasma parameters are taken typical of the Earth’s ionosphere. The emphasis is placed on frequencies lying in the resonant part of the VLF (whistler) range, in which the whistler-mode refractive index surface has unbounded branches. As is known, the conventional integral equation approach cannot be applied readily to wire antennas in this case. Therefore, the current distribution on the antenna surface is treated as a boundary-value problem. Within the framework of such an approach, a rigorous solution for the current distribution of an infinitely long antenna is initially obtained. The current induced on the antenna comprises the discrete-spectrum contribution, which is represented by a wire-guided mode supported by a cylindrical conductor in a magnetoplasma in the whistler range, and the continuous-spectrum contribution [3]. It is shown that, at the chosen frequencies, the current distribution of a sufficiently thin antenna is predominantly determined by the wire-guided mode since the continuous-spectrum current is comparatively small in the presence of a sheath around the antenna wire and, hence, can be neglected. Then the truncation of the antenna to a finite length is performed and reflections of the wire-guided mode between the antenna ends are considered. Based on this concept, a generalized transmission-line theory is developed for determining the current distribution and the input impedance of a finite-length antenna. The described approach is shown to be suitable for finding the characteristics of VLF wire antennas, both short and long (such as ionospheric tethers, for example), in space plasma. Detailed numerical results will be reported for particular cases of interest.

References
[1] H.G. James and K.G. Balmain, Radio Sci., vol. 36, no. 6, pp. 1631-1644 (2001).
[2] T.M. Zaboronkova, A.V. Kudrin, and E.Yu. Petrov, Radiophys. Quantum Electron., vol. 42, no. 8, pp. 660-673 (1999).
[3] A.V. Kudrin, E.Yu. Petrov, G.A. Kyriacou, et al., Progress In Electromagnetics Research, vol. 53, pp. 135-166 (2005).

The use of vector-sensor antennas for localization of source signals has received significant attention [1,2]. The output of these sensors, proportional to the corresponding electric or magnetic field, is intended to be used for estimation of the source direction and polarization.

These sensors are densely packed within a small volume. Hence, coupling among the elements of sensors cannot be neglected. Several attempts have been made to evaluate this coupling. For example, in [2], two collocated circular loops are considered. The loops have identical radii, their centers coincide, and the loops are located in two orthogonal planes. We present several technical problems that occur with such loops.

We tried to reproduce the analysis method and results from [2,3]. We expanded the loop currents in terms of Fourier- series harmonics. For a small number of harmonics, we obtained a good agreement with the results of [2,3]. However, as we increased the number of harmonics, our results for the input and mutual impedances converged to significantly different values, showing much stronger coupling than predicted in [2,3].

The explanation is as follows. The wire axes of the two loops intersect at two locations, which is not taken into account in the analysis in [2,3]. These intersections behave as if the two loops are galvanically interconnected at these points. Hence, the results of the refined simulations naturally converged to the same results as for interconnect loops. We verified this conclusion by analyzing the same loops using the programs of Refs. [4,5].

In [2,3], the actual feeding arrangements and the matching networks for the loops were not taken into account. We analyzed some realistic feeders for two and three collocated loops to investigate the resulting coupling and symmetrization problems.

In conclusion, the coupling among the loops is extremely strong, regardless of whether the loops are galvanically interconnected or not. This imposes significant problems for practical applications of the collocated loops.

REFERENCES

[1] A. Nehorai, and E. Paldi, "Vector-Sensor Array Processing for Electromagnetic Source Localization", IEEE Trans. Signal Processing., Vol 42, pp. 376-398, February 1994.

[2] Y. Huang, A. Nehorai, and G. Friedman, "Mutual Coupling of Two Collocated Orthogonally Oriented Circular Thin-Wire Loops", IEEE Trans. Antennas and Propagat, Vol. 51, pp. 1307-1314, June 2003.

[3] S. Krishnan, L.W. Li, and M.S. Leong, "Comments on 'Mutual Coupling of Two Collocated Orthogonally Oriented Circular Thin-Wire Loops'", IEEE Trans. Antennas and Propagat, Vol. 52, pp. 1625-1626, June 2004.

[4] A.R. Djordjevic, M.B. Bazdar, V.V. Petrovic, D.I. Olcan, T.K. Sarkar, and R.F. Harrington, AWAS for Windows Version 2.0: Analysis of Wire Antennas and Scatterers, Artech House, Boston, 2002.

[5] B. Kolundzija, J. Ognjanovic, T. Sarkar, M. Tasic, D. Olcan, B. Janic, D. Šumic, WIPL-D Pro. v5.1, WIPL-D, 2004.

The increasing demand for bandwidth in wireless communication drives the transceiver systems to higher operating frequencies. The license-free band at 60 GHz receives much attention because of the almost unlimited amount of bandwidth that is available there. Moreover, the advances in silicon technology allow the design of low-cost electronics that operate at millimeter-wave frequencies. As a result, there is a need for low-cost antennas that exploit the bandwidth of about 5 GHz that is available.

We discuss a novel design of a balanced-fed aperture-coupled patch antenna. The geometry of the antenna is shown in Fig. 1. The balanced feed and the planar layout of the antenna allow for a suitable interconnection with a balanced amplifier. The aperture coupling avoids the need for vias and is used to increase the bandwidth.

Fig. 1. Antenna geometry

There are two major problems associated with completely planar aperture coupled patch antennas. The first problem is the excitation of surface waves in the dielectric [1]. The second problem is the front-to-back ratio of the antenna, i.e., due to the aperture coupling, part of the power will radiate to the backside of the antenna. In the design, the surface-wave excitation is minimized through the use of two distant coupling apertures which cancel part of the surface-wave power, as proposed in [2]. To improve the front-to-back ratio, a reflector element is introduced, following an idea presented in [3]. Both those design strategies are used together for the first time, to enhance the global efficiency of the antenna.

In conventional aperture-coupled patch designs, the aperture is made small in terms of wavelengths such that the aperture is non-resonant and the back radiation caused by the aperture is reduced. Here, the apertures can be resonant in the operation band of the antenna because the reflector element compensates for the back radiation. As a result, the bandwidth is increased significantly since the antenna now has two resonant elements, i.e., the patch and both apertures. An increase is observed from about 3% for the conventional single-element patch to more than 10% for the novel design.

The antenna is analyzed numerically by means of an in-house developed method-of-moment implementation. This implementation allows for an analysis of the radiation efficiency and bandwidth of the antenna. The effect on antenna performance of the separation between the slots, the thickness of the layers, and the size of the reflector has been analyzed in detail. For example, the impact of the reflector element is shown in Fig. 2. The front-to-back ratio and radiation efficiency improves significantly when the reflector element is added. From simulations we observe a -10dB bandwidth of 6 GHz in the frequency range from 54 to 60 GHz. The simulated radiation efficiency is more than 82% within this band. A prototype of the structure is currently being manufactured in a low-cost PCB technology. Measurements will be performed at TNO Defense and Security and will be presented at the conference.

Fig. 2. Front-to-back ratio and radiation efficiency with (solid) and without (dashed) reflector.

[references: see attachment]

For most applications, a horn antenna radiating a nearly perfect Gaussian radiation pattern is required or preferable. A perfect candidate is of course the corrugated horn.

However, for applications over around 80 GHz, it is becoming difficult and expensive to manufacture such horns as the corrugation pitch and depth tend to be very small. To eliminate the need to use corrugated horns for high frequency application, the smooth-walled horn technology [1][2][3] has recently been applied to a number of challenging applications.

At the CSIRO ICT Centre, we have developed a software package able to optimize the radiation pattern of smooth-walled horns by shaping the profile of the horn via the coefficients of a spline [1]. Several horns were designed, manufactured and tested successfully for a number of applications, ranging from 80 to 840 GHz [4][5][6].

We will showcase horns designed for radioastronomy applications at around 100 [1], 345 [5] and 840 GHz [4], for antenna test-range illumination at 200 GHz (one giving a Gaussian pattern and one giving a sectoral pattern [6]) and for wireless communications at 86 GHz. The available references [4][5][6] tend to focus on the applications. However, in this paper, the challenges, solutions and restrictions of these applications, from a horn design point of view, will be highlighted.

We have also designed spline-profile horns for future THz applications [7] [8] and a 1.9 THz horn has been manufactured and is already running in the "low-frequency" channel of the GREAT receiver [7]. No dedicated radiation pattern measurements of the horn itself have been done as yet, but they are expected in the near future.

[1] Granet C., James G.L., Bolton R. and Moorey G., "Smooth-Walled Spline-Profile Horn for Radio Astronomy", JINA 2002 International Symposium on Antennas, Nice, France, 12-14 November 2002, Volume 1, pp 375-378.
[2] Deguchi H., Tsuji M., Shigesawa H., Matsumoto S., "A compact low-cross-polarization horn antenna with serpentine-shaped taper", IEEE Antenna and Propagation Symposium, Boston, USA, 2001, pp. 320-323.
[3] Neilson J.M., "An improved multimode horn for Gaussian mode generation at millimeter and submillimeter wavelengths", IEEE Transactions on Antennas and Propagation, Vol. 50, No 8, Aug. 2002, pp 1077-1081.
[4] Rabanus D., C. Granet C., Murk A., Tils T.. "Measurement of properties of a smooth-walled spline-profile feed horn around 840 GHz", Infrared Physics & Technology, 2006.
[5] Lüthi T., Rabanus D., Graf U.U., Granet C., Murk A., "Expandable fully reflective focal-plane optics for millimeter- and submillimeter-wave array receivers", Review of Scientific Instruments, 77, 014702, January 2006.
[6] Barker S.J., Granet C., Forsyth A.R., Greene K.J., Hay S.G., Ceccato F., Smart K.W., Doherty P., "The development of an inexpensive high-precision mm-wave compact antenna test range", Proceedings of AMTA, Newport, Rhode Island, USA, Oct. 30 - Nov. 4 2005, pp. 337-340.
[7] GREAT receiver: http://www.ph1.uni-
koeln.de/workgroups/astro_instrumentation/receiver/
[8] SOFIA platform: http://www.sofia.usra.edu

The standardisation process for IEEE802.15.3c Wireless Personal Area Networks (WPAN) allocated at 60 GHz requires transmit and receive circuits capable to handle high data rates. In this context, a planar Vivaldi antenna has been developped to be easily integrated in RF-frontends. This paper presents the requirements during antenna design, its performance and the integration into the final circuit.

I. Introduction

Up to now the WLAN at allocated at 2.4 GHz and from 5-6 GHz has reached a large acceptance. Regarding new applications, a new standard at 60 GHz has to be established. In this context, a 60 GHz communication link has been developped and a dedicated antenna has been designed.

II. Antenna Design

For a first prototyp the following antenna specifications had been fixed:
- fc = 61 GHz, B = 3 GHz
- S11 < -10 dB,
- G > 10 dBi

Considering the integration the main task is to connect the antenna directly to the amplifiers by a microstrip line. These demands lead to a Vivaldi antenna [1]. The antenna layout is shown in fig. 1.

Its broadband behaviour will not only meet the requirements, but guarantee insensibility against fabrication tolerances [2]. The two antenna wings are arranged on the two sides of the RO3003 substrate. The measured gain patterns are shown in fig. 2. As can be seen, the minimum gain of 10 dBi is reached almost in the complete frequency band.

III. PCB Layout

This antenna has been integrated in two PCB Tx and Rx-Boards and connected to specially developped amplifiers [3,4] via bond wires. First tests showed a good performance of the WPAN communication setup. Furthermore, channel measurements of indoor communication scenarios are currently performend. A more detailed description will be given in the full paper.

IV. Summary

This paper presents the design constraints, the results and the integration of a 60 GHz Vivaldi antenna for WPAN communication link. The antenna has successfully been integrated and first communication tests show promising results.

References

[1] E. Gazit, "Improved Desing of the Vivaldi Antenna", IEE Proceedings, 135 (2), pp. 89 - 92, April 1988
[2] K. Schuler, Y. Venot and W. Wiesbeck, "Innovative material modulation for multilayer LTCC antenna design at 76.5 GHz in radar and communication applications", Proceedings of the 33rd European Microwave Conference EUMC, pp- 707 - 710, Oct. 2003
[3] Y. Sun, F. Herzel, L. Wang, J. Borngräber, W. Winkler and R. Krämer, "An integrated 60 GHz receiver front-end in SiGe:C BiCMOS", Digest of Papers Topical Meeting on Silicon Monolithic Integrated Circuits in RF Systems, October 2006
[4] Y. Sun, J. Borngräber, F. Herzel and W. Winkler, "A fully integrated 60 GHz LNA in SiGe:C BiCMOS technology", Proceedings of the Bipolar/BiCMOS Circuits and Technology Meeting, pp. 14 - 17, October 2005

Terahertz (THz) fields normally refer to the electromagnetic fields between 1 and 10 THz, which is between the microwave and the infrared regions in the spectrum. This region is so far the least explored because it has multidisciplinary characters and there is very limited equipment available on the market for experiment. THz attracted very little attention before 1990s. As the wireless communications have been developed at a fast pace, more and more researchers are working on THz region in recent years. This is largely due to its special features and potentials for many attractive applications.

There are many problems associated with THz technique. For example, how to design efficient THz antennas that are very different from the conventional antennas. There has been very little publication on this subject. This paper is aimed at addressing this problem.

Due to how THz signals are generated, there are basically two types of THz antennas: the conventional antennas and photoconductive antennas. In THz region, the skin depth of a conductor is very small, thus conductive loss could be very high. Photoconductive antenna is the one attracted more attention and it seems to performance better. The principle of generating the THz wave from the photoconductive antenna is very different from the traditional one. First, a laser beam is focused on a semiconductor to make it produce some pairs of electrons and holes. Then, the electrons and holes are accelerated by the bias voltage which is connected to the antenna fabricated on the semiconductor. Thus a time-varying current is produced and the electromagnetic wave can therefore be generated.

Normally a photoconductive antenna has a coplanar stripline structure. It consists of a semiconductor substrate and a coplanar stripline with a dipole antenna fabricated on the substrate. Some recent researches observed that if being added the indentations at the edge of the dipole, the antenna can get a better performance on generating the signals. In this paper, we will not only compare various THz antennas, but also explain the reasons, such as why the indentations can improve the performance of the antenna. Some simulation results will also be provided to aid our discussion.

In this paper, important aspects of a 3D-FDTD algorithm are presented for the analysis of AIAs. Mm-wave and sub-mm wave frequencies are very promising for future communication links. Millimetre-wave frequencies have a small wavelength and active integrated antennas (AIAs) have a huge potential to exploit this. A successful CAD should be able to handle the non-linearities of the active elements as well as the radiation from the antenna. Overall successful design of AIAs depends on the availability of accurate CAD. The method chosen is the FD-TD as it is a powerful technique that solves Maxwell’s equations efficiently and accurately.

The algorithm presented here includes the use of an un-split PML as the absorbing boundary condition. One of the most important aspects of modelling AIAs is the accurate representation of the packaged active elements. As the operational frequency increases, the effects of the packaging become more prominent. The need to capture the equivalent circuit becomes evident. The non-linear lumped network FD-TD (NL2N-FDTD) method presented by Emili et al [1] was specifically developed to accurately model a packaged Schottky diode in one cell. The simulation results of the code have been verified by comparing the results with either published ones (fig.1) or results from an older version of the in-house code. This algorithm will enable further simulations of AIA arrays for mm-wave imaging applications at 90 GHz. The core element of this array will be a 90 GHz receiver slot-ring.

Fig. 1-Voltage across a Schottky diode:Comparison between the FD-TD code being developed and published results

[1] G. Emili et al, "Rigorous modelling of packaged Schottky diodes by the non-linear lumped network FD-TD approach", IEEE Trans. Microwave Theory Tech., Vol.48, pp.2277-2282, 2000.

Recently, multilayer technology using Low Temperature Co-fired Ceramics (LTCC) has become popular in high frequency applications. In fact, LTCC allows high precision 3-D structuring and vertical integration of the circuits, to increase density and electric performances and to achieve low cost and size reduction. Therefore, this study focuses on millimetre wave applications for telecommunications and on LTCC technology. In order to fully take advantages of the 3-D integration of LTCC, a global design is presented integrating both a 3-D U-shape resonator and a patch antenna, functioning at Q band. Our purpose is to conceive a two-pole radiant filter obtained by associating the resonator resonance and the antenna one. Filtering targets of this filter are the following : centre frequency at 40 GHz, 8,5 % -3 dB bandwidth and 10 dB return loss. This article aims to prove the problem feasibility using the different elements that constitute the filter.

The substrate is made of eight 99-µm-µm-thick layers of Ferro A6S(εr=5,9 and tanδd=1,3.10-3). The resonator is designed with three layers of Ferro A6S. The input excitation system is composed of CPW ports printed on the bottom face of the substrate. The radiative element (a lambdag/2 antenna) is placed on the top face. The resonator and the patch are coupled through a rectangular slot located in the middle of a ground plane between layer 5 and layer 6 (Fig.1).
The structure has been realized and measured and the simulated S11-parameter ((SigmaAg=4,7.10e7 S/m and tanδ=1,3.10e-3) compared to the experimentation reflection response is shown in Fig.3.
Theoretical and experimental results are in good agreement and show satisfying electrical performances. The calculated and measured radiation (Fig.4-5) results permit to validate this study and the concept is thus demonstrated.
This original structure allows to improve the natural frequency filtering behavior of the patch by combining the patch and the resonator and to obtain both satisfying filtering and radiating functions.

One of the applications of a such structure concerns antenna array designs. By considering the radiant filter as an elementary cell, we could develop original array with simplified design methodology while maintening high performances. Such antenna array is then considered as an elementary component that will be mounted on a flip-chip carrier. The component can be associated with other ones to build greater antenna arrays. So the use of these components will facilitate the design of the antenna array feeding circuits that will be design on the substrate carrier. The feeding circuit of each elementary component will be designed on the LTCC substrate bottom that permit to simplify the flip-chip bounding. To show the concept feasibility, the first step of our work lies in the design of a structure composed of two elementary radiant filters.These two cells are linked by the input excitation system printed on the bottom face of the substrate. The first results of simulated return loss demonstrate that it is possible to define a circuit feeding on the LTCC substrate bottom. Calculated radiation performances have been defined (Fig.7).
According to the design and to the radiation pattern, the computed radiation pattern of the two cells array shows that the beam is focused in the E plane, that validates our approach.

The new interesting applications for the frequencies above 100 GHz for example in imaging, communications and spectroscopy require the design of high performance and lower cost components. Dielectric materials are an interesting alternative and there are already many dielectric components such as waveguides, directional couplers and phase shifters. This is also one of the reasons why dielectric rod waveguide antennas of rectangular cross-section are interesting.

Such antennas have been designed already for a long time from low permittivity materials for several frequency ranges by feeding the antenna with a standard metal waveguide. The benefits of the materials with high permittivity are the reduced size without losing good waveguiding properties and the better matching with the metal waveguide [1,2]. Such designs do not need launching horns that are needed in case of low permittivity materials. Nowadays the developed technology has made it possible to manufacture these high permittivity rods that often require high precision as the required dimensions are so small.

In this paper the design of the dielectric rod waveguide (DRW) antenna of rectangular cross-section made of silicon for the 110-150 GHz frequency range is presented. A standard 110-170 GHz metal waveguide was used as a feeding element. The matching from the metal waveguide to the dielectric rod can be improved significantly by tapering the rod feed end. E-plane tapering was chosen according to simulations and earlier studies [1]. The radiated beam can be formed and matching can be improved by tapering also the radiation end. To make radiation pattern smoother, parasitic radiation from the transition region was suppressed by introducing an absorber around the dielectric rod.

Figures 1 and 2 present the measured H- and E-plane radiation patterns of the silicon antenna of 42 mm total length with 4 mm E-plane tapers in both ends. One interesting feature that can be observed from the measurement results is the rather frequency independent beamwidth. While the -10 dB beamwidth of an open waveguide varies about 20˚, the measurement results of the DRW antenna show only about 10˚ variation in both planes in the frequency range of 110-150 GHz.

Fig. 1. Measured H-plane radiation pattern. Fig. 2. Measured E-plane radiation pattern.

[1] D. Lioubtchenko, S. Tretyakov, S. Dudorov, "Millimeter-Wave Waveguides", Kluwer Academic Publishers, Boston, 2003.
[2] D. V. Lioubtchenko, S. N. Dudorov, J. A. Mallat, A. V. Räisänen, "Dielectric rod waveguide antenna for W band with good input match", IEEE Microwave and wireless components letters, vol. 15, no. 1, pp. 4-6, Jan. 2005.

Design techniques of millimeter-wave planar array antennas with several radiation patterns and gains are developed for many sensor applications such as long-range and short-range automotive radars and home and car security systems. There is no perfect millimeter-wave antenna which is always optimum for any applications. So, we should possess design techniques for several types of millimeter-wave antennas. Either slotted waveguide antennas or microstrip antennas can be used for most of the major applications in the millimeter-wave band as the situation demands because they have completely different advantages from each other.

Waveguide antennas are advantageous for its high-efficiency property even in the case of large aperture and high gain. However, a simple antenna structure which allows metal injection molding is expected for mass-production of the waveguide antenna. On the other hand, it is difficult for microstrip antennas to realize high gain because feeding loss of the microstrip line is significant in this frequency range. However, etching technique is superior in points of cost reduction and wide design variety depending on the printed patterns. Consequently, waveguide antenna is adequate for high-gain applications and microstrip antenna is for relatively low gain applications such as short-range sensors and sub-arrays of DBF (Digital Beam Forming) systems. We introduce proposed structures and performances of the developed waveguide and microstrip antennas in this paper.

First, a high-gain slotted hollow waveguide array antenna is developed for long-range sensors, as is shown in Fig. 1. A guide wavelength of a hollow waveguide is generally larger than a wavelength in free space. So, the grating lobe is a serious problem since the slot spacing is approximately one guide wavelength in general broadside arrays. Therefore, we propose "a post-loaded slot" and "a single-layer alternating-phase feeding circuit" in order to suppress grating lobes. Measured grating lobe level, gain and antenna efficiency are -28.6dB, 33.2dBi and 56%, respectively. Furthermore, we have developed two-dimensional design techniques for arbitrary radiation patterns of microstrip comb-line antennas as is shown in Fig. 2. We present the design techniques for low sidelobe, beam tilting and feeding circuit in this paper.

The PLANCK satellite will observe the most ancient radiation of the Universe, known as the cosmic microwave background, with the highest accuracy ever achieved. PLANCK will carry a 1.5-metre telescope and will focus radiation from the sky onto two arrays of highly sensitive detectors, the LFI (Low Frequency Instrument) and the HFI (High Frequency Instrument). The RFQM (Radio Frequency Qualification Model) of this satellite has been tested in the Alcatel Alenia Space CATR (Compact Antenna Test Range) in February 2006. The antenna is composed by the QM (Qualification Model) telescope structure, the QM primary and secondary reflectors and a specific FPU (Focal Plane Unit) constituted with 4 horns (LFI and HFI) working at 30, 70, 100 and 320 GHz. Radiation measurements in amplitude and phase over 4p steradian have been performed with the acquisition system based on MI-Technologies equipment. Specific transmitter and receiver modules (up-converters and down-converters) have been developed by Farran Technology at 70-100-320 GHz. Pseudo time-gating correction method has been applied thanks to frequency sweep capabilities. The dynamic range obtained during this measurement campaign was around 80 dB at 320 GHz, 95 dB at 30 GHz and better than 100 dB at 70 and 100 GHz.

Advanced radio telescope antennas for space applications are realized with application of stable composite materials, which are lighter in general than various metal realizations. Reduction of reflection loss of these materials is of paramount importance for various applications, like terrestrial (ALMA sub-millimetre wave telescope for instance) and for Space (PLANCK reflector for instance).

Here the latest results are presented for the sample measurements.

The reflection loss of the samples set, representative for the Planck reflector and a set of samples representative for the baffle material has been measured in a frequency range 110 - 350 GHz. The reflectivity of the traditional high reflective metals like Silver, Copper, Gold and Aluminium also will be presented. The configuration and measuring method for 2-mirror resonator set-up with normal beam incidence will be shortly discussed. There is no need for any external reference standard, as will be explained and good results have been obtained with the techniques for the data extraction.

Further measurements have been performed in a 3-mirror resonator set up, where the sample under test forms the third mirror of the resonator. The reflectivity is deduced from the comparison of the Q-factor of the 3-mirror resonator and the corresponding 2-mirror resonator. With this method, the reflectivity can be measured for arbitrary angle of incidence and polarization.

A comparison of the calculated reflectivity with the measured reflectivity for real metal reflectors is performed with a critical discussion, including accuracy estimations.

We demonstrated that the reflection loss of reflectors treated with a coating process as for Planck or also for a reflector metallised with Vacuum Deposited Aluminium (VDA) and aluminium layer on a kapton film, may be essentially higher than it follows from DC calculation.

Obviously this depends on the coating process technology and its associated parameters. Therefore, this leads to the clear conclusion that the measurement of reflection loss is very important, in particular for the indicated frequency range. Relevant aspects will be described in the final paper with highly accurate results.

SiGe HBT and MOSFET cut-off frequencies are higher than 230 GHz and this increase allows new millimeter wave (MMW) applications on silicon such as 60 GHz WLAN and 77 GHz automotive radar.

This study focuses on standard silicon antenna topic. Integrated antenna achieved on STMicroelectronics advanced sub 120 nm BiCMOS low resistivity (15 Ohms.cm), High Resistivity Silicon On Insulator (HR SOI) (>1 kOhms.cm) and glass technologies (>1 kOhms.cm) are described. CPW patch antenna and double slot antenna types are retained to perform a benchmarking of antennas. A specific study on the impact of substrate conductivities on these antennas is proposed. Antenna characterization is led on a dedicated on-wafer test bench. Performances in term of gain and radiation pattern are given.

Feasibility of integrated antennas on silicon has been demonstrated [Montusclat et al. JINA04, EUMC05, ICMTS05] but this paper focuses on benchmarking of substrate resistivity on standard manufacturing. Two antennas are under investigations: a coplanar patch antenna and a double slot one. For these coplanar antennas, electromagnetic field can go through substrate and antenna behavior can be modified. This choice allows a comparison of performances between substrate conductivities; this is one of the main objectives of this paper. Antenna designs and constraints are described too.

Finally, matching and radiation measurements are completed with a dedicated on-wafer test bench using HP8510XF VNA and Cascade Microtech Inc. GSG probes.

CPW patch antennas

First studied antenna is a CPW patch. This antenna must respect technology design rule constraints to ensure process reliability. Patch is realized with a grid of metallization to decrease densities on CMOS technologies. This antenna type has been realized in low resistivity, HR SOI and glass technologies to evaluate their capability. Dimensions have been fixed to obtain a 40 GHz resonant frequency on each technology and so perform a right comparison (matching, pattern and gain). Measurements have been performed in the far field region on a dedicated test bench for on-wafer measurement (described in final paper). To probe antennas, specific devices are used: the pads which specifications and effects will be shown. Figure 1 gives S11 comparison for these 3 antennas and Table 1 a summary of the achieved performances.

Double slot antennas

A same benchmarking is summarized Table 1 and will be developed in final paper.

Full characterization (radiation pattern and S11) for these devices will be provided.

In proposed advanced silicon technologies allowing MMW designs, gains have been measured for presented coplanar antennas. The range begins from nearly -8 dB for the CPW patch antenna on low resistivity silicon substrate until 2.3dB for this antenna on glass substrate.

A benchmarking on substrate conductivities is fulfilled for HR SOI, low resistivity and glass technologies. In both antenna cases (CPW patch and double slot), HR SOI substrates provide 4dB improvement on antenna gain. This is a really promising result for antenna design and developments in standard silicon manufacturing.

The Array for Microwave Background Anisotropy (AMiBA) is a hexapod telescope for research in cosmology, designed to measure the Cosmic Microwave

Background anisotropies at 85-105 GHz. The AMiBA hexapod platform can accommodate up to 19

antennae for interferometry measurements. It is located on Mauna Loa, Hawaii. The project is led by the Academia Sinica Institute of Astronomy and Astrophysics Taiwan (ASIAA).

We present design, simulations, manufacturing process and performance verification for the 0.6m diameter antennae.
7 antennae are currently used in the first operation phase and the construction of the 1.2m antennae is on the way.
Both antennae sizes are Cassegrain systems with a primary paraboloid and a secondary hyperboloid. The 0.6m primary has a f/ ratio of 0.35 with a final effective f/ ratio of the assembled system of 2.03.

Primary and secondary mirrors are novel carbon fiber sandwich structures, manufactured by CoTec Inc., Taizhong, Taiwan. The primary is composed of a carbon skin and Nomex honeycomb, the secondary is a carbon skin and potting compound.
After the mold is machined by CNC, vacuum assisted resin transfer molding is used for the manufacturing of the sandwich structures. The cured dish is then lathing machined and later aluminum coated.
In this process the required 50 microns rms surface accuracy is easily achieved for the 0.6m antenna. Typical rms values are 20-30 microns and 10 microns for the primary and secondary, respectively. The better surface accuracy increases the antenna efficiency from 95% to more than 99%.

Simulated load conditions for the structure analysis of the antennae include self-weight, temperature, wind pressure and material properties. Resulting maximum rms surface errors are 40 and 5 microns for the primary and secondary mirror, respectively.

Finally, the antennae beam pattern is measured in the far field, using a 90 GHz artificial point-source. A computer-controlled tripod mount with the antenna and with a spectrum analyzer for read-out is used to do a 2-dimensional scan of the source. With this high-resolution data we can confirm an excellent beam pattern, which is essential for the later analysis of astronomical observations.

Introduction

The superstrate EBG structure is regarded as a promising solution to achieve a high antenna gain. Patches and array of patches are generally used to feed this kind of structures; however the losses can be significantly important in the millimeter range due to the feeding network. To avoid this drawback, a radiating slot antenna fed by a rectangular waveguide, which considered as a broadband radiating source, is associated to a superstrate medium in order to enhance the inherent antenna gain of the structure.
II. The proposed antenna structures

The chosen superstrate consists of a superposition of dielectric slabs of successively low and high permittivities, where the low permittivity material is here simply made of air and the high permittivity material is a TMM10 substrate (ε=9.2). The suggested antenna structure for validation has a total size of 60 mm x 60 mm x 7 mm. The achieve results by simulations and measurements are promising. However, the obtained bandwidth makes the applications of superstrate EBG structure limited to narrow band systems. To treat this problem, we propose to replace the air by the Rexolite (ε=2.53) in order to enlarge the bandwidth. On the Fig.1 we present a photograph of the realized superstrate EBG antenna using two slabs of TMM10 separated by a slab of Rexolite. In order not to disturb significantly the near field of the radiating source, a distance of lambda 0/2 is taken between the radiating source and the first dielectric slab. Both TMM10 dielectric slabs are separated of lambda 0/4 for air and lambda g2/4 for Rexolite, and the thickness of each slab of TMM10 is of lambda g1/4, lambda g1 being the guided wavelength within the TMM10 and lambda g2 for Rexolite. The central frequency for which the focusing system will work, is directly related to the thickness and permittivity of the materials, which leads to a working frequency of 38.9GHz.

III. Results

The measured return loss is presented on Fig. 2. One can notice the improvement of the bandwidth, which pass from 150 MHz to more than 1GHz after we replace the air by the Rexolite. The measured gain is shown on Fig. 3, the gain is about 19 dBi when using air and the half-power beamwidth is about 12°, after changing towards the Rexolite the gain decreases by 3 dB to reach 16 dBi but it will be stable on a broader band, the half-power beamwidth increases to 22° what explain the decreases of gain. For the both cases, the cross polar is less then 25 dB. The antenna efficiency is increased from about 20% to 47%.

Conclusion

We use an optimized waveguide-fed radiating slot to feed an EBG substrate structure using one dielectric material with high permittivity like the TMM10, the bandwidth of this structure is about 150 MHz, after replacing the air by the Rexolite, the bandwidth is improved to cover more then 1GHz of band around the central frequency of 38.9 GHz. The performances obtained by the association of two technologies, the waveguide and the EBG cavity resonators, makes it possible to consider such antennas structures as a good choice for wireless communication systems