A novel dual polarized ultra wide band (UWB) feed with a decade bandwidth will be presented for use in both single and dual reflector antennas. The feed has nearly constant beam width and 11 dBi directivity over at least a decade bandwidth. The feed gives an aperture efficiency of the reflector of 66% or better over a decade bandwidth when the subtended angle towards the sub or main reflector is about 53 deg, and an overall efficiency better than 47% including mismatch. The return loss is better than 5 dB over a decade bandwidth. The calculated results have been verified with measurements on three linearly polarized lab models: one operating between 1 and 15 GHz, another between 150 and 150 MHz, and a third between 500 and 2 GHz. The feed has no balun as it is intended to be integrated with an active 180 deg balun and receiver. The feed is referred to as the Eleven antenna because its basic configuration is two parallel dipoles 0.5 wavelengths apart and because it can be used over more than a decade (and 11 > 10 :-) bandwidth with 11 dBi directivity. We also believe that 11 dB return loss is achievable.

The feed has been developed on projects at Chalmers for California Institute of Technology for use in the US SKA [1] (Figure 1), the US proposal for a future radio telescope of one Square Kilometer Array aperture. The paper will also describe a 150-1700 MHz version of the feed that weights more than 100 kg and has been developed at Chalmers and delivered to NRAO in Green Bank West Virginia for use in their 42 m radio telescope on a project managed by MIT Lincoln lab in Boston. The latest work has been done on a 500 MHz - 3000 MHz version for the Russian RATAN radio telescope.

The paper will describe the principles of operation of the Eleven antenna, and the details of the three different designs, including measured radiation patterns and radiation efficiencies. The radiation efficiencies have been measured in a reverberation chamber.

[1] R. Olsson, P.-S. Kildal, S. Weinreb, "Compact low-profile decade bandwidth log-periodic folded dipole feed for reflector antennas", to appear in special issue on multifunction antennas in IEEE Transactions on Antennas and Propagation, February 2006 (Sander Weinreb is with California Institute of Technology).

Planar dielectric structures used for increasing the directivity of a point source, have been investigated since many years [N.G. Alexopoulos, D.R. Jackson, 1985]. The effects of a superstrate with high permittivity, placed on a substrate with very low permittivity, have been studied demonstrating the establishment of a resonance condition, which maximizes antenna gain, radiation resistance, and radiation efficiency. Particular attention has been given to the physical interpretation of this resonance effect, which can be described in terms of leaky waves (LW), of both transverse electric (TE) and magnetic (TM) type, excited in the structure. Under certain resonance conditions, a pair of weakly attenuated TE/TM leaky waves become the dominant contributions to the antenna aperture field. Both the attenuation and the propagation constants of the TE/TM leaky waves become small under the resonance conditions thus ensuring a large phased aperture field at the free-space interface and a consequent narrow radiated broadside beam. This effect has been already explored and applied to relax the distance among array elements without occurring in grating lobes, and gaining space for the feeding network [M. Albani, F. Capolino, 2002]. More recently, the effect of highly-directive pencil beam radiation, has been used in the design of feeding arrays for Multiple Beam Antennas (MBA) in multimedia telecommunication satellites, where a multi beam coverage is requested, with high gain spots, low side-lobe and cross-polar high purity [K.S. Rao, G.A. Morin, M.Q. Tang, S. Richard, K.K. Chan, 1995]. The basic structure of the MBA is constituted by a single reflector illuminated by a feed cluster, with advantages with respect to multi aperture systems in terms of coverage flexibility, costs, mass and spacecraft accommodation. Overlapped beam footprint in the reflector aperture plane can be achieved by increasing the effective area of each feed. This requires a partial overlap between the radiators themselves. This approach is expected to reduce drastically the focal array complexity thanks to an overlapping performed at the level of the electromagnetic aperture, as opposed to the use of a bulky beam-forming network applied to the individual feeds. Some recent studies [G. Tayeb, S. Enoch, P. Sabouroux, N. Guerin, P. Vincent, 2002][R. Chantalat, P. Dumon, B. Jecko, M. Thevenot, T. Monediere, 2003][ S. I. Khureim Castiglioni, G. Toso, C. Mangenot, 2004] demonstrate the applicability of this overlapped feed concept. Nevertheless, the polarization characteristic of these structures has not been investigated, yet. For multifeed applications, this is an important aspect to be considered. The aim of this paper is to rigorously demonstrate that the conventional choice of the stratification, not only ensures a large aperture field, but at the same time realizes linear polarization of the same aperture field, when excited by any linearly polarized point source. The linear polarization performance are explained by resorting to a suitable analytical derivation, involving an appropriate closed form expression for the dominant LW term, and the complex image technique for the remainder.

The Hybrid Mechanical-Electronic Pointing System (HMEPS) is a project of the European Space Agency (ESA), which considers future scenarios using the Q/V band on satellite communications. This work has been developed under contract between University of Vigo and ACORDE (Advanced Communications Research and Development), which is one of the subcontractors of the ESA project "Hybrid Mechanical-Electronic Pointing System" lead by SENER, Ingeniería y Sistemas, S.A. and with the participation also of INASMET, and University of Cantabria. The main characteristic of the antenna of the HMEPS is to provide two separate controls of the main beam: electronic control and mechanical control. The mechanical control is produced by an adjustable subreflector whose design and development is the main objective of the project. The subreflector adjustment works cooperatively with the electronic beam control which is provided by a source array as primary feed of the antenna system. In a Q/V mission based on beam scanning from the GEO orbit, a field of view of 5 ° around the boresight direction is considered to provide coverage to European mainland, insular regions and other countries of influence.
A Confocal Gregorian antenna has been considered as a baseline design due to its good scanning properties. Furthermore, the subreflector itself can become a single and adjustable reflector antenna with a reconfigurable beam for other missions at Q/V, Ka or Ku bands.
The surface design is based in the following steps: 1. For each scanning condition an Ideal Subreflector surface is obtained by Geometrical Optics shaping techniques. There is one degree of freedom in this technique: the total path length of the rays (L) can be arbitrarily modified.
2. The Ideal Subreflector surface is approximated by a tilted best fit paraboloid by a least squares scheme. A paraboloid has six parameters: focal length (f), axis orientation (u,v) and focal location (x_F, y_F, z_F).
3. After defining the mechanisms, only three degrees of freedom exist to control the surface (u,v,f). The other three parameters (x_F, y_F, z_F) are dependent. A new mechanically restricted tilted best fit paraboloid (MRTBFP) is now defined. This MRTBFP is the objective for the design of the actuators and the flexible surface.

A prototype of the subreflector has been manufactured by SENER (mechanisms and structure) and INASMET (surface). The manufactured was also measured under different scanning conditions. The RF performance of the antenna will be presented. The results proved the feasibility of kind of technology.

The helical beam antenna first developed by Kraus is very circularly polarized in a direction along the axis of the helix, however, when the cross-section becomes ellipse, the axial ratio bandwidth (AR<3dB) starts to decrease and becomes a relatively narrow band CP antenna. In this paper, elliptical helical antenna with parasitic element is proposed to enhance the reduced bandwidth caused by change of the cross-section. Broadband CP bandwidth characteristic of the axial-mode helix with circular cross-section was attributed in the literature to the traveling current distribution along the helix wire during a very large frequency band.

When the cross-section is ellipse other than circular, the current distribution rules will be changed, and reflected waves start to appear and become the dominant component in some frequency bands. That leads to CP bandwidth becoming narrow. To resolve the problem, a parasitic helix with the same cross-section has been added to utilize the couple effect between the parasitic and driven helix to make the total current a traveling wave distribution. It is obviously observed that the total current distribution is related to the feeding point and the length of the parasitic helix, as a results, the work is mainly about selecting the appropriate starting point and the ending point of the parasitic element.

The simulation in Moment Method based commercial software IE3D is carried out to verify the hypothesis. The geometry of the proposed antenna is shown in Fig.1, and the optimized result is shown in Fig.2. It can be noticed that the single helix with eccentricity 0.94 has 350MHz CP bandwidth, and the bandwidth of the other one with parasitic element increases to 660MHz. It is also observed that the gain of the antenna in the main beam can be improved by maximum 2dB in a large band. The parasitic method leads to 88.6% increment of the CP bandwidth, and that verifies the fact that the appropriate parasitic helix can enhance the circularly polarized bandwidth.

This paper compares two different compact lens antenna approaches for a global Earth observation application at 26GHz. Lens specifications were defined by Alcatel Alenia Space within the framework of ACE Network of Excellence activities. The lenses are required to match a secant square type radiation pattern template in the elevation plane with right-hand circular polarization (RHCP). Two alternative specifications are defined for the azimuth radiation pattern template: either omnidirectional or with a mechanically scanned sector beam to enhance gain.

Two alternative lens approaches are evaluated for the above cases: one based on axial-symmetric dome lens geometry (developed at "Instituto de Telecomunicacoes", Portugal) and another one based on a fully 3D lens design (developed at "IETR", France).

Despite the axial symmetrical geometry that is used in one of the approaches, it can also provide the required shaped sector beam solution by appropriate modification of the axial symmetric lens design and the feed positioning. The dome antenna is made of ECCOSTOCK K=10 with both an inner and outer quarter wavelength matching layer of ECCOSTOCK K=3. The feed is a RHCP printed antenna and the dome lens shape was optimized for the desired azimuth template. For the omnidirectional radiation pattern case, the feed was placed on-axis. Results show a good agreement with specifications. For the sector illumination case, a new axial symmetric lens profile was designed with the feed located off-axis. This is still a dome lens, however the inner surface of the dome is now spherical and centered with the feed, rather than with the outer lens surface axis. This configuration produces the required sector shaped beam, which can be scanned just by rotating the dome lens around the feed axis with fixed feed. The gain increase with respect to the on-axis dome lens is of the order of 10 dB.

The double-shell 3D lens is fed by an aperture-coupled microstrip patch antenna. The source is centered with respect to the lens axis, and the inner and outer shells are made of MACOR and Rexolite, respectively. The shapes of both shells have been designed using a hybrid Geometrical Optics / Physical Optics (GO / PO) method of analysis coupled with a binary Genetic Algorithm. The design procedure accounts for the above-mentioned electromagnetic specifications as well as the fabrication and assembling mechanical constraints. The GO / PO results are in good agreement with the specifications and with full wave FDTD computations. To reduce the amplitude of the ripples in the secant-squared part of the radiation pattern, another 3D lens antenna with a similar configuration has been designed, but in this case the lens is made from a different pair of materials with lower dielectric constants (Rexolite for the inner shell, and foam for the outer shell); this allows reducing the effects of multiple internal reflections. In both cases, the shaped beam is scanned by rotating the whole lens antenna. Compared to the axial symmetrical geometry with centered feed, the gain increase is of the order of 9dB.

The performance of the four lens antennas presented here will be compared numerically and experimentally in the final paper.

Introduction

EADS Defence and Security Systems was selected by the DGA (French DoD) to develop a new communication system called ACN (Airborne Communication System). The ACN provides a communication service characterized by very high capacity and very high coverage area on easy and quickly deployable gateways, without any infrastructures. The ground gateways communicate through an airborne node, set up in a 33000 ft ceiling aircraft. Considering the specific constraints due to the integration under an aircraft (low weight and size, 33000 ft environment,) to the large coverage area (50 km radius, that means 170 degrees width), the circular polarization, the need to get the same power in the whole area and the need to get a low dispersion (< 1 dB) of the patterns in the wide frequency band (14.47-15.16 GHz), this leaded to develop a totally new compact antenna with profiled reflectors. This antenna is presented in the article.

Description of the antenna

A photo of the complete antenna designed with SRSR software tool [1] is shown in figure 1. The sub-reflector is positioned near the aperture of the feed (a few centimeters). A glued dielectric rod is used between the sub-reflector and the feeder to maintain the assembly. The circular septum polarizer (axial ratio < 0,5 dB for LHCP and RHCP) is integrated in the main reflector of the antenna. The diameter of the antenna is 50 cm and its height is 20 cm. Figure 1 : Photo of the antenna.

Results

The antenna was measured in our anechoic chamber (spherical near fields technique). Four cuts in circular polarization were registered for each frequency. Excellent agreement was obtained between the computed values and the measured values on the whole frequency band. The far fields patterns for the RHCP mode at 14.6 GHz and 15.05 GHz are plotted in figure 2 and 3 respectively. Results for the LHCP mode are similar. The gain is always higher than 6 dBi on all the frequency band and for both circular polarization. Figure 2 : Specific cuts at 14.6 GHz. Figure 3 : Specific cuts at 15.05 GHz.

Conclusion

All the measurements were compliant with the required specifications. Mechanical tests were made and were passed with success. Two flight models based on this antenna were manufactured. All the results of these FMs are completely similar to these results presented in this paper.

Reference

[1] Berthon A., Bills R. " Integral Equations Analysis of Radiating Structures of Revolution ". IEEE Trans. Antennas and Propagation, Vol. AP-37, no. 2, pp. 159-170, Feb. 1989.

State of the art:

Reflector antenna systems have been deeply investigated for decades and can be considered a relatively mature technology, [1]. Typically the feeds of the reflector systems aiming at high gain are realized with high performance corrugated horns. Recent investigations seem to indicate that there is some potential for an increase of the performances of reflector antennas using Electromagnetic Band Gap (EBG) super-layers. In particular, in [2]-[4] it was indicated that EBG super-strates could be adopted to increase the directivity of an aperture small in terms of the wavelength. In [5] the enhanced directivity was used to ensure that most of the power radiated by the feed was intercepted by the reflector and thus not lost as spill over. However, the decrease in taper efficiency will in spite of the improve spillover efficiency limit the global aperture efficiency [1].

In this Contribution

BG dielectric super-layers are used to improve the overall aperture efficiency of a single feed reflector system. In this scope the Green's function of the EBG superstructure is investigated in detail. In particular, from its spectral representation the most significant leaky pole contributions are extracted and their role in defining the shape of the far field pattern of an underlying feed is described. Inverting this procedure, starting from a required radiation pattern, one can identify the pertinent geometrical parameters that provide the wanted leaky wave poles. Thus an angular filtering effect that maximizes the reflector efficiency can be obtained. A feed prototype that aims at realizing the ideal pattern sec⁴(theta/2) [1] has been designed, and is being manufactured. The preliminary results indicate that the inclusion of the EBG increases the aperture efficiency from 75% to values higher than 81%, over a 10% bandwidth. Moreover a peak efficiency of 85% is obtained which would be ruled out by standard simple horn design curves.

References:

[1] Balanis Antenna Theory : Analysis and Design, Wiley Interscience 2005

[2] C. Cheype, C. Serier, M. Thevenot, T. Monediere, A. reineix, B. Jecko, "An Electromagnetic Bandgap Resonator Antenna" IEEE Trans. AP Vol.50, no.9 Sept. 2002, pp. 1285-1290

[3] Y. J. Lee, J.Yeo, R. Mittra, W.S. Park "Application of Electromagnetic Bandgap Superstrates with Controllable defects for a Class of patch Antennas as Spatial Angular Filters" IEEE Trans. AP Vol.53, no.1 January 2005, pp. 224-235

[4] Guerin, N.; Enoch, S.; Tayeb, G.; Sabouroux, P.; Vincent, P.; Legay, H. "A Metallic Fabry–Perot Directive Antenna" IEEE Trans. AP, Vol. 54,no. 1, Jan. 2006 pp. 220 - 224.

[5] S.I. Khurein Castiglioni, G. Toso, C. Mangenot, "Multi-beam Antenna Based on a Single Aperture Using Overlapped Feeds", Proceedings of 13-th JINA November 2004

Singly-fed stationary reflector antennas that are scanned by translating the feed have limited field of view (FOV) due to beam distortions appearing for off-axis directions of observation. These distortions are larger for reflectors with small focal ratios (F/D<0.5) and limit the increase of the FOV to a few beams when an array of conventional feed horns is employed. To overcome the limitations of these conventional designs and improve the FOV, dense Focal Plane Arrays (FPA) have been suggested. They can accurately reproduce the field distortions by adopting the weights of the many electrically small elements. Although, the potentials of the concept have been widely disscused in recent publications, the effects of the gain (power) losses in practical dense arrays were not taken into account when estimating the aperture efficiency of the beams. The primary goal of the paper is to investigate the impact of reflection and dissipation losses on the aperture efficiency in dense arrays. The radiation efficiency of an FPA fed reflector with different element spacings will be shown for two examples: an FPA of open-ended hard waveguides and an FPA of Vivaldi elements. We look into these issues by performing a theoretical full-wave analysis of an infinite array of open-ended hard waveguides and an 8x9x2 array of wideband Vivaldi elements. The latter is theoretically as well as experimentally characterized. In the course of the study, it will be demonstrated that a tradeoff between the beam and radiation efficiencies has to be made in order to achieve the maximum overall aperture efficiency of the FPA-fed reflector system.

To offer an initial insight into the work, the figure below shows how the radiation efficiency of a Vivaldi array FPA feed, weighted according to the Airy function, varies with element spacing. It is seen that the efficiency rises with increasing element spacing, as a result of reduced mutual coupling losses. On the other hand, the beam efficiency (consisting of different contributions such as the amplitude taper, phase uniformity, spillover, polarization etc) decreases with the larger element spacing due to coarser sampling of the Airy pattern (not shown here). The resulting product of the beam and radiation efficincies will indicate the range of the element spacings which is optimum for achieving the maximum overall aperture efficiency of dense arrays.

Analysis of dielectric lens antennas are often performed by high-frequency techniques based on geometrical and physical optics. Though, it has been proven, that proper accounting of multiple reflections enables one to improve the performance of corresponding numerical algorithms and to widen the domain of their applicability [D. Pasqualini, IEEE AP-Tr, 52(3)-840, 2004, A. Pavacic, IEEE AP-Tr, 54(2)-604, 2006], the accuracy of such approximate techniques as to the analysis of small size and arbitrary shaped lenses, which are the key components of integrated lens antennas, are still questionable.

These difficulties can be overcome by using full wave methods based on integral equations or FDTD. Unfortunately, for 3-D lens antenna analysis and synthesis, the applicability of both approaches is limited due to mathematical complexity of a 3-D vector formulation of the problem for the former and high requirements to computer time and resources for the latter. In 2-D, implementation of Muller’s boundary integral equations (MBIE) enables one to develop algorithms with controllable accuracy for any set of lens parameters [S. Boriskina, JOSA-A, 21(3)-393, 2004]. 2-D versions of FDTD can easily be realized and are much more efficient in terms of computer resources than 3-D ones. This and the easiness of handling the lenses of complicated shapes and different types of primary sources make them attractive for lens antenna analysis [M. Vorst, IEEE MWCL, 12(7)-258, 2002]. In spite of popularity of FDTD approaches, they are known to be inaccurate near the natural resonances of the considered objects [G. Hower, IEEE AP-Tr, 41(7)-982, 1993]. These numerical errors are uncontrollable and thus may spoil the analysis due to the important role of internal resonances in electromagnetic behavior of small size lenses [A. Boriskin, MOTL, 43(6)-515, 2004]. Unfortunately, there are very few papers focused on the validation of FDTD as to the reliable characterization of resonance phenomena in open resonators.

The objective of the reported work is estimation of a 2-D FDTD approach accuracy in the analysis of near and far field characteristics of small size 2-D hemielliptic lens antenna. To do this, we first consider circular cylindrical resonators to verify the performance of FDTD and MBIE algorithms as to the description of high quality whispering-gallery (WG) resonances by comparing with the exact series solution (SS). Here, we observe perfect matching of numerical results between SS and MBIE both far and near the resonance frequencies. The FDTD algorithm demonstrates a shift of the WG resonance frequencies that can be reduced (however not eliminated) by choosing a denser mesh. It also looses accuracy near the resonances both in the near and far field analysis.

We consider also extended hemielliptic lenses made of rexolite, quartz and silicon, excited by a focal line current source. Here, the MBIE method is used as a reference solution. The comparisons are performed for near and far field characteristics for lenses with the bottom size of 2, 3 and 5 wavelengths in free space and different bottom extensions. The analysis shows that inaccuracies entering the solution become significant only for silicon lenses when the so-called half-bowtie resonance is excited (Fig 1).

Key words: hemielliptic lens antenna, FDTD, Muller boundary integral equations.

Symmetrical Ortho-mode transducers (OMTs) are key components in dual-polarized antenna feed systems for their inherently wide band operation (approx.50%). The drawback of the symmetrical configuration is that any manufacturing and assembly errors are directly converted into anti-parallel field components resulting in undesired spikes in the S-parameter response and degradation of the cross-polarization performance of the antenna. Additionally, for high power applications, multipactor as well as Passive Inter-Modulation (PIM) are real concerns that have to be addressed carefully. A typical solution to mitigate the risk of generating PIM would be to monolithically manufacture the OMT using Electroforming technique. This however leads to longer procurement time and higher cost. This paper presents a symmetrical ortho-mode transducer in WR62 to cover the frequency band from 11.40 to 18.40 GHz, using a wide band spike suppressor device. The OMT is made of several aluminum parts fabricated by conventional machining technique. Multipactor analyses have been performed following the ESTEC multipactor recommendations showing a power threshold above 20 kWatts. Compared to its electroformed counterpart this new approach leads to significant savings in terms of cost and schedule. A breadboard was manufactured and tested for scattering parameters, radiated cross-polarization and PIM. The unit was bench tested to verify the spike-free return loss and isolation performance. Tested return loss was better than 22 dB and the polarization isolation was better than 56 dB across the full band. Return loss could have been significantly improved over particular sub-bands with minor geometrical changes. The impact of high order modes excitation was checked by testing the radiated cross-polarized signal of the OMT assembled with a corrugated horn. The worst case cross-polarized signal level due to high order modes was less than -45 dB. Finally, the assembly was PIM tested over temperature at the 5th and 7th order with two 180 Watts carriers. The results show a PIM level below the noise floor (<-140 dBm).