I. INTRODUCTION
The purpose of this joint French–Japanese work is to design a detection system for debris or Foreign Object Debris (FOD) on airport runways [Reference Beasly, Binns, Hodges and Badley1–Reference Mazouni5].
The fold topic became of interest with the Concorde accident in 2000. Companies started to develop systems that are actually on international airports like Vancouver and Heathrow [Reference Beasly, Binns, Hodges and Badley1], Singapore [2], and Boston [3]. Except for [2], the companies have chosen to include a millimeter wave radar in this set-up. The Tarsier developed by QinetiQ operates with a 94 GHz radar, whereas FODetect (Xsight and Thales) operates with a 77 GHz radar.
The project includes as well the design and fabrication of a high gain antenna (LEAT, France) as the development of a millimeter-wave front-end (ENRI, Japan). Measurements campaigns are carried out jointly. The complete system is composed of the high-gain antenna and a radar module, working in W-band between 76 and 77 GHz. The antenna is a printed reflectarray antenna (RA), generally used for millimeter wave radar applications because of its excellent trade-off between high directivity and low loss, low profile, and low cost. Nevertheless, a primary source from center generates a masking effect called “aperture blockage”, which decreases the antenna efficiency. To avoid this effect, the reflectarray is designed with an offset feed with a main radiation beam of the RA in the specular direction, for specular radiation minimization purpose. The side lobe level is a pertinent parameter for this system as debris can be close to each other with a high radar cross section (RCS) dynamic range. Therefore, a low RCS FOD detected in the main beam could be hidden by a higher RCS FOD detected in a side lobe. A primary feed with a prolate radiation pattern modifies the signal in such a way that 99% of the power is within the main lobe. As a consequence, the secondary lobes are greatly reduced and the overall noise level accordingly. Section II describes the antenna design and Section III deals with the antenna measurements, while Section IV presents the results from our measurement campaign.
II. ANTENNA SPECIFICATIONS AND MODELIZATION
In 2009, from a preliminary study, a 35 dBi gain circularly polarized reflectarray has been developed [Reference Mazouni5]. Elementary cells were designed to achieve linear to circular polarization conversion. This configuration has the main advantage to simplify considerably the antenna implementation. Although radar tests were conclusive for small distances [Reference Mazouni5], the gain of the antenna was not sufficient to detect −20 dBsm target at 46 m as recommended by the FAA advisory circular for this application [Reference O'Donnell6]. Furthermore, the use of a circular polarization did not show any greater interest compared to linear polarization for this application. For that reason, a second antenna with 40 dBi gain and linear polarization was designed. Moreover, in the perspective of the future 76–81 GHz radar module developed by the ENRI the antenna is designed for this band. In consequence, the design frequency of the antenna is 78.5 GHz. Nevertheless, only 76–77 GHz results are presented here (for concision purpose) because the whole radar system used for measurement presented in this paper is operating in this band.
A) 27° offset reflectarray
As for every quasi-optical antenna, RA has to compensate for the phase delay of the spherical incident wave coming from the primary feed. In RAs, the phase compensation is obtained by adjusting patch dimensions. Patches are placed on the flat surface of a back-metallized substrate (Fig. 1) [Reference Pozar, Targonski and Syrigos7].
where Rij is the distance between the feed and the patch (i, j), rij is the distance between the patch (i, j) and the RA center, r 0 is the direction of the main beam, k 0 is the wave number, and ψij is the phase that has to be reflected by the patch (i, j) for delay compensation.
Fig. 1. (a) Scheme of the offset reflectarray. (b) Scheme of an elementary cell.
Equation (1) describes the phase compensation for a focused beam. It is possible to adjust main antenna parameters such as directivity, radiation pattern shape or number of beams with proper optimization. Moreover, RAs with feed from center have a high aperture blockage that decreases the overall aperture efficiency. Therefore, an offset primary feed at 27° has been chosen. Previous studies have shown that elementary cells radiate also in the specular direction [Reference Milon, Gillard and Legay8]. In our case, a high lobe should appear at θ = −27°. To solve this problem, the radiation direction of the focused beam is chosen in the specular direction, i.e. θ = −27°. Figure 1 shows the antenna geometry.
Moreover, a primary source with a prolate radiation pattern is used to increase the total aperture efficiency [Reference Soummer9, Reference Lanteri, Migliaccio, Dauvignac and Pichot10].
B) Prolate horn
The field radiated by a reflectarray is proportional to the Fourier transform of the tangential electrical field in the aperture, i.e. the Fourier transform of the radiation pattern of the primary feed. The influence of the latter on the RA radiated far field can be compared to the effect of a window in signal processing. Therefore, a prolate function is chosen for the radiation pattern of the primary feed. The prolate window modifies the signal in such a way that the main lobe contains 99% of the power. As a consequence, the secondary lobes are greatly reduced such as the overall noise level. The aperture efficiency is improved. The template of the prolate window can be approached with enough accuracy by a Kaiser window. The coefficient β enables to adjust the level of the primary feed radiation pattern at the offset angle of 27°. We have chosen to design the horn for having −20 dB at 27°, which corresponds to β = 5. Simulations were conducted using SRSRD software, developed by France Telecom Orange Labs dedicated to axisymmetric antennas [Reference Berthon and Bills11].
Numerical results in the E, H, and 45°-planes are shown in Fig. 2. The results obtained agree well with theory over ±27° in the E, H, and 45°-plane.
Fig. 2. Simulated radiation pattern and template of the Prolate primary feed at 77 GHz.
III. MEASUREMENTS
The reflectarray is composed of elementary cells with rectangular patches etched on a 254 µm Duroïd substrate (ɛ r = 2.2). The cell lattice is λ/2. Diameter and focal length to diameter ratio (f/D) are of 160 mm and 1.125, respectively. Elementary cells are simulated using Ansoft HFSS with the Floquet port method, which allows the simulation of a single cell within an infinite array. Reflection phase is obtained by variation of the width and the length of the rectangular patch. Then 340° phase excursion is achieved at 77 GHz as described in Fig. 3(a). As it is not possible to simulate large R.A. (here we have 40λ) with HFSS and home-made program based on an equivalent aperture method is developed at the LEAT for radiation pattern calculation purposes. Due to the approximations that have to be done results are expecting to be accurate within the main lobe only.
Fig. 3. (a) Phase versus width and length. (b) Phase versus frequency. (c) Phase versus frequency and incidence.
Several studies are conducted on the reflectarray cells. First the phase behavior versus patch dimension is determined (Fig. 3(a)). From these results it is possible to simulate the gain loss due to the fabrication errors (±50 µm) with our program. They have been estimated to be 0.7 dB at 77 GHz. Moreover, the reflected phase variation of the cell over frequency has been simulated. In a first step under normal incident for several patches dimensions (Fig. 3(b)), then for different incidences (Fig. 3(c)). One observes a linear variation.
The reflectarray has been fabricated and measured at LEAT. It is aligned by stand tilted at 27° in elevation to measure the main lobe in the θ = 0° direction. Measurements are shown in Fig. 4 over the bandwidth of interest.
Fig. 4. Measured radiation patterns at 77 GHz.
The 3 dB aperture is 1.7° and secondary lobes remain inferior to 20 dB. The relatively high radiation level between −60° and −15° is due to the fixture that masks the transmitting antenna during E-plane measurement. The antenna gain value is ranging between 38 dBi at 76 GHZ and is above 40 dBi after 76.5 GHz. It corresponds to an aperture efficiency varying between 30 and 50% that are state-of-the-art values in the W-band [Reference Zeitler, Lanteri, Pichot, Migliaccio, Feil and Menzel12, Reference Byun and Cho13]. As expected it rises near the central frequency and goes up to 70% at 78 GHz, which is to our best knowledge one of the highest value for the RA. Figure 5 shows the detection system, used during the measurement campaign in Japan. It is composed of the offset reflectarray, the frequency modulated continuous wave (FM-CW) radar module and another stand made at ENRI was used instead of the one used in the anechoic chamber.
Fig. 5. View of the full system.
The antenna gain has been measured again with this stand and its value was ranging between 36 and 36.5 dBi between 76.25 and 76.75 GHz. This 2.5 dB discrepancy is due to a misalignment between the primary source phase center and the reflectarray surface. A new stand has been fabricated and preliminary results have shown that the 40 dBi gain value has been again obtained. It should be sufficient for detecting −20 dBsm objects, but the extended bandwidth of the system up to 80 GHz will also improve the detection since the antenna gain is ranging between 40 and 42 dBi in the upper frequency band.
IV. RADAR MEASUREMENT
The monostatic FM-CW radar has been developed at the ENRI. It is composed of a Ku band driver circuit (12.3–13.3 GHz) and a radar module. The module is compact (33 × 35 × 20 mm3), without connections and waveguides, and lightweight (180 g) without antenna.
Tests are conducted between 76 and 77 GHz, over a 500 MHz bandwidth. To avoid parasitic detections caused by near objects taken from the side lobes, we have seen that particular attention has been paid to the side lobes levels of the antenna (<20 dB SLL). The emitting power is less than 10 dBm according to the specifications for external measurement requested by the Japanese Administration. Measurements were conducted on pavement at ENRI site in November and December 2010, respectively. Although, several FOD systems already exist, there is no detailed evaluation of expected performances, except for the FAA circular of 2009 [Reference O'Donnell6]. It describes relevant test object for FOD application. A target of −20 dBsm has to be detected up to 46 m for a system made with distributed radar modules on the runway. To evaluate the sensitivity and the detection capability of the radar, four standard targets were chosen:
• Three metallic cylinders' monostatic RCS have been calculated analytically [Reference Knott, Shaeffer and Tuley14] from 0, −10 to −20 dBsm.
• A corner reflector (28 dBsm) was employed.
Each cylinder has the same dimensions in diameter and height: cylinder C1 with 134.5 mm, C2 with 62.4 mm, and C3 with 29.0 mm. RCS values corresponding to an incidence at 0° and receivers placed 360° around the targets were simulated using Finite Element Boundary Integral (FE-BI) new module of HFSS [15]. Simulations have been carried out between 76 and 81 GHz in the perspective of the future large bandwidth front-end that is under development at the ENRI. Since there is no significant variation over the bandwidth of interest, Fig. 6 shows the simulation results at 78.5 GHz, which corresponds to the center of the frequency band. One observes that the maximum RCS value for each cylinder is not in the monostatic configuration (θ = 0°). Therefore, it would be interesting to investigate using multistatic radar in the future for detection improvement.
Fig. 6. Targets and the RCS numerical simulations.
Figure 7 shows a view of the four targets placed on pavement at ENRI and Fig. 8 represents a 60° scanning radar image obtained with a 0.12° angular step. The rotation of the whole radar (Front-end + antenna) is carried out in the xOz-plane with a motor placed under the antenna. The height of the whole system is 30 cm.
Fig. 7. Scene scanned by the radar.
Fig. 8. Radar image with objects placed at 10 m from the radar.
The four targets are detected at 10 m together with several elements of the environment. Indeed, metallic fences, long the pavement ENRI, are detected along tens of meters. We note the high reflection of the corner reflector on the right side, which masks the closest objects due to the longitudinal resolution limitation (30 cm). Moreover, the −20 dBsm cylinder is detected up to 40 m.
V. CONCLUSION
An offset RA with 40 dBi gain at 77 GHz has been designed, fabricated and measured at LEAT anechoic chamber. Unfortunately, due to the distance separating LEAT and ENRI, the foreseen stand made at ENRI for assembling the FM-CW radar module and the RA was not enough accurate and lead to a 2.5 dB discrepancy in the antenna gain. A measurement campaign was carried out jointly in December 2010 with this configuration. Although, first results are conclusive for the radar detection capabilities, the level is not sufficient for detecting a −20 dBsm cylinder at 46 m. Fortunately, with a new stand, the 40 dBi antenna gain value has been obtained. Moreover, a new FM-CW module is under construction with an operating frequency bandwidth ranging between 76 and 81 GHz. With this configuration, we can take advantage of the best antenna gain value around 78 GHz. Finally, RCS simulation results have shown that multistatic measurements should improve the detection, but this set-up faces some important difficulties due to synchronization between the different receivers in the W-band.
ACKNOWLEDGEMENT
This work was supported by the JSPS and the French Ministry of Foreign Affairs under Sakura project number 21153ZF.
Karim Mazouni was born in France, in 1980. He received the B.S. degree in electronic and electrical engineering and automatics from the University of Nice Sophia-Antipolis, France, in 2006 and the M.S. and Ph.D. degrees in electronic engineering from University of Nice-Sophia Antipolis, France in 2008 and 2011, respectively. His major research interests focus on millimeter antenna, reflector, reflectarray, radar system, FOD and power line détection.
Christian Pichot (M'83) was born in France in 1951. He received the Ph.D. and the Doctorat ès Sciences degrees from the University of Paris-XI (Orsay), France, in 1977 and 1982, respectively. In 1978, he joined the Laboratoire des Signaux et Systèmes, CNRS/Ecole Supérieure d'Electricité, Gif-sur-Yvette, France. During the 1989–1990 academic year, he was a Visiting Researcher at the Lawrence Livermore National Laboratory, Livermore, CA. From 1992–1995, he was with the Laboratoire d'Informatique, Signaux et Systèmes de Sophia Antipolis, University of Nice-Sophia Antipolis/CNRS, Valbonne, France. He is currently a Research Director CNRS and Director of the Electronic, Antennas and Telecommunications Laboratory, joint research unit supported by the University of Nice-Sophia Antipolis and CNRS, France. He was Chairman of the Platform “Design” of the Microelectronics Integrated Center of the Provence-Alps-Riviera Region (CIM PACA), France, from 2005 to 2007. His research activities are concerned with scattering and propagation of electromagnetic waves, radiation of antennas, inverse scattering (microwave imaging and tomography, complex permittivity reconstruction, object detection and recognition), theoretical analysis, numerical and experimental aspects for applications in radar, civil engineering, non-destructive testing, engineering geophysics, security and military applications, antennas, telecommunications, and medical domain, VLF/LF frequencies, microwaves and millimeter waves.
Jérôme Lanteri received the M.S. and Ph.D degrees in electrical engineering from the University of Nice-Sophia Antipolis, France, in 2004 and 2007, respectively. He was a post-doctoral researcher at the University of Nice-Sophia Antipolis, France, from 2007 to 2008, and at CEA-LETI, Grenoble, France, from 2008 to 2010. Since September 2010, he is Associate Professor at the University of Nice-Sophia Antipolis, France. His research interests include integrated antennas for gigabit wireless communications, reflectarrays and transmitarrays at millimetre-wave frequencies.
Jean-Yves Dauvignac he is a Professor at the University of Nice-Sophia Antipolis, France. During his Ph.D. thesis (1993), his research activities are concerned the modelization of microwave antennas using finite-surface-element-method to solve integral equations of the electromagnetic field. Since 1996 he is involved in microwave imaging for the detection of buried objects and in the study of dielectric resonator antennas. He developed a new class of UWB antennas (ETS Antennas) in planar printed technology for GPR and radar synthetic impulse microwave imaging system (SIMIS) for surface penetrating radar applications. During this period, he has also contributed in several campaigns of measurements road survey, pipes and mines detection in collaboration with LCPC and ONERA. In 2002, he was involved in the design of UWB antennas for telecommunications applications and the measurements of UWB antennas in time domain. More recently he has worked on the design of high gain antennas for radar systems in millimeter waves. Actually, he continues his research and high gain millimeter-waves-antenna. Since 2000, he is the head of the team “High gain antennas and UWB detection and imaging microwave and millimeter-wave systems” of LEAT.
Claire Migliaccio received the “Diplôme d'Ingénieurs ENSERG” in 1993 and the Doctoral degree in 1996 from the “Institut National Polytechnique de Grenoble”, Grenoble, France. From 1996 to 2007, she worked as Associate Professor in the University of Nice Sophia Antipolis, France, and since October 2007 she is a Professor at the University of Nice Sophia Antipolis. From 1996 to 2001, her research interest was focused on uniplanar large band antennas. Since 2001, she has moved to mm-wave antennas and systems. It includes the design of large antennas for radar applications, such as related measurements techniques. She has conducted or is involved in several mm-wave projects in collaboration with academic or industrial partners including collision avoidance system for rescue helicopters, large antennas for satellite communications or FOD detection system.
Akiko Kohmura joined the Electronic Navigation Research Institute, Japan as a researcher after receiving her PhD degree in 2007. From 2011 to 2012, she is with the Laboratoire d'Electronique, Antennes et Télécommunications, France as a guest researcher. Her main interests are millimeter wave radar, antennas for the radar, and electromagnetic compatibility on aircrafts.
Shunichi Futatsumori received the B.E. and M.E. and Ph.D. degrees in electronics and information engineering from Hokkaido University, Sapporo, Japan, in 2004, 2006 and 2009, respectively. From 2008 to 2009, he was a Research Fellow of the Japan Society for the Promotion of Science. In 2009, he joined the Electronic Navigation Research Institute (ENRI). Japan, where he has been engaged in the research of millimeter radar systems and electromagnetic compatibility issues. Dr. Futatsumori received the Young Researcher$B!G(Bs Award of IEICE and APMC prize, both in 2009. Dr. Futatsumori is a member of the IEEE and IEICE.
Naruto Yonemoto received the B.E. and M.E. and Ph.D. degrees in electronic engineering from Saga University, Saga, Japan, in 1995, 1997 and 2000, respectively. He joined the Electronic Navigation Research Institute (ENRI), Japan in 2000. He was a visiting researcher of Laboratory of Electronic, Antennas and Telecommunication in France from 2005 to 2006. He is currently a chief researcher of ENRI. His research interest includes millimeter wave radar system and electromagnetic compatibility. He is a member of IEEE, IEICE, EuMA.
