2496 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 56, NO.

8, AUGUST 2008

Bandwidth Improvement in Large Reflectarrays

by Using True-Time Delay
Eduardo Carrasco, Student Member, IEEE, José A. Encinar, Member, IEEE, and Mariano Barba

Abstract—A significant improvement in the bandwidth of large antennas and small F/D ratios [1], [2]. Usually, the required
reflectarrays is demonstrated using elements which allow true-time phase value to compensate for the different paths of the wave
delay. Two identical, large reflectarrays have been designed using from the feed horn to each array element is only achieved at
different phase distributions to generate a collimated beam. In the
former, the phase distribution is truncated to 360 as is usual in the central frequency within a range of 360 . When frequency
reflectarray antennas, while in the second, the true phase delay varies, a phase error proportional to the differences in path
is maintained (three cycles of 360 ). The chosen phase-shifter el- length is produced. This effect produces a small reduction
ements are based on previously measured and validated patches in gain for pencil beam reflectarrays as shown in [2], and
aperture-coupled to delay lines. The radiation patterns for both re- produces a significant distortion of the beam shaping in con-
flectarrays have been computed at several frequencies and the gain
is represented as a function of frequency for both cases. Bandwidth toured beam reflectarrays [4] when frequency varies out of
curves are presented as a function of the reflectarray size. the central frequency. This band limitation is the same as that
Index Terms—Broadband antennas, reflectarrays, time-delay
which occurs in zoned reflectors [12], in which the depth of
arrays. a parabolic reflector is reduced in steps of an integer number
of half wavelengths, in a similar manner as in zoned dielectric
lenses [13]. Although the surface design of an unzoned lens
I. INTRODUCTION is independent of frequency, zoning the lens reduces its mass
and losses in the dielectric material, but limits the bandwidth
T HE most severe drawback in reflectarray antennas is
their inherent narrow-band performance, and much effort
has been made in recent years to overcoming this limitation
through the introduction of frequency dependence [14].
Parabolic reflector antennas use the physical geometry to
[1]–[11]. Reflectarray bandwidth is limited mainly by two equalize spherical waves arriving from the feed horn to form
different factors as described in [1]. The first is the narrow band a planar wave front. This phase equalization is independent of
of the radiating elements; the second is the differential spatial the frequency, and it is equivalent to the true-time delay (TTD)
phase delay resulting from the different paths from the feed to technique, as it is called in phased arrays. In contrast, printed
each point on the wave front of the radiated beam. elements on reflectarrays are normally used for phase-shift
The bandwidth limitation produced by the radiating element compensation in a limited 360 phase range.
is the more stringent for moderate size reflectarrays, as demon- A phase-delay range with more than one cycle of 360 can be
strated in [2]. Different types of reflectarray elements have been used to improve the bandwidth of reflectarrays. Stubs of vari-
proposed to improve the element bandwidth in printed reflectar- able length attached to rectangular patches [15] can introduce a
rays, including stacked rectangular [3], [4] and ring [5] patches, TTD, but with a very restricted room for long lines and impor-
double crossed loops [6], and patches with other shapes [7]. tant spurious radiation levels. The use of varying-length lines
In a recent work, a reflectarray based on artificial impedance aperture-coupled to patches [10], [11] can reduce the effect of
surfaces has been proposed to increase the bandwidth of reflec- the differential spatial phase delay through the introduction of a
tarrays using a single layer of printed elements [8] arranged in physical path that produces TTD in a wide phase range. How-
a regular lattice with period of less than half a wavelength. This ever, in this type of element, the phase of the total reflected field,
configuration, based on the reduction of the period, allowed in- including the field reradiated by the patch and the one reflected
by the grounded dielectric, exhibits a nonlinear response pro-
creasing the bandwidth up to a 20% for a reflectarray of around
ten wavelengths in diameter. For this antenna diameter, the er- duced by the resonances of the structure, as shown in [16]. In
rors introduced by the differential spatial phase delay do not pro- order to achieve TTD, the elements must be designed to provide
duce a significant reduction in the bandwidth. a linear phase response, proportional to the length of the line in
a large range of phase delay. For example, the resonance effects
On the other hand, the bandwidth limitation produced by the
differential spatial phase delay is critical for electrically large were compensated in [17] by properly adjusting the length of
the aperture, and a reflectarray element was designed to provide
a phase directly proportional to the line length.
Manuscript received October 29, 2007; revised March 10, 2008. Published
August 6, 2008 (projected). This work was supported in part Spanish Ministry
The improvement of bandwidth using true-time delay phase
of Science and Technology under the project MEC TEC 2004-02155 and by the compensation in reflectarrays is demonstrated in this paper.
Mexican National Council of Science and Technology (CONACYT). For this purpose, two pencil-beam reflectarrays have been
The authors are with the Electromagnetism and Circuit Theory Department, designed at 9.65 GHz using two different phase distributions
Universidad Politecnica de Madrid, 28040 Madrid, Spain (e-mail:carrasco@etc.
upm.es). implemented by patches aperture-coupled to delay lines. The
Digital Object Identifier 10.1109/TAP.2008.927559 first distribution is limited to one cycle of 360 and the second
0018-926X/$25.00 © 2008 IEEE
CARRASCO et al.: BANDWIDTH IMPROVEMENT IN LARGE REFLECTARRAYS BY USING TRUE-TIME DELAY 2497

TABLE I
DATA FOR THE ELEMENT SIMULATED IN WGS

Fig. 1. Reflectarray element based on patches aperture-coupled to delay lines.

(a) Expanded view. (b) Top view.

distribution is expanded up to three times 360 . A bandwidth

improvement is demonstrated in the second case, as a result of
the TTD.
Additionally, other moderate-sized reflectarrays, which re-
quire fewer than three cycles of phase-delay have been designed
using limited phase distributions and expanded TTD distribu-
tions. The bandwidth curves as a function of the reflectarray size
are presented, demonstrating that by using TTD the bandwidth
is not reduced in large reflectarrays.

II. ANALYSIS, VALIDATION, AND DESIGN

OF REFLECTARRAY ELEMENT Fig. 2. Simulated and measured phases for reflectarray element in WGS.

Reflectarray elements based on patches aperture-coupled to TABLE II

delay lines allow a wide range of phase delay if the delay line is DESIGNED RADIATING ELEMENT DATA
bent forming a U-shaped line like that shown in Fig. 1. For the
analysis of this type of reflectarray element, a full-wave modular
technique described in [18] is used. The element is considered in
an infinite array environment and it is analyzed by the method of
moments (MoM) in spectral domain by considering an arbitrary
incident field on the periodic structure, with its corresponding
angle of incidence. The modular approach is based on the in-
dividual characterization of each array layer (array of patches,
array of apertures, and array of delay lines). First, a generalized
scattering matrix (GSM) which characterizes each array layer
is computed separately, and then, the total GSM is obtained by Fig. 2 shows the comparison between computed and mea-
the iterative cascading of the individual GSMs of each interface sured phase of the reflection coefficient of the designed element
using simple matrix computations. The overall GSM of the cell in a frequency band from 7.5 to 8.5 GHz. The simulations has
provides the phase-delay, ohmic losses and cross-polarization been performed using roof-top basis functions to characterize
produced by the reflectarray element. the U-shaped delay line. Straight microstrip lines larger than the
For the experimental validation, the well known waveguide waveguide borderline has been measured to produce the phase
simulator (WGS) technique has demonstrated accurate results delay performance. Good agreement between simulations and
[19]. A reflectarray element has been designed, manufactured measurements can be seen with maximum differences in the
and measured in a WGS for different line lengths. Simulation order of 20 that is into an acceptable range, mainly produced
of an infinite array can be verified by the insertion of some ra- by photo etching errors, alignment errors, and by the assump-
diating elements inside a waveguide section and measuring the tion of an infinite array of the delay lines in the simulations.
reflection in the excitation port. The imaging action of the wave- Once the proposed phase-delay element has been validated,
guide walls simulates the infinite array for one polarization. A a reflectarray element is designed for a wide range of phase
WR112 waveguide has been chosen, imposing a unit cell size delay. The phase response can be very linear if a proper adjust-
of 14.25 mm 12.6 mm (two radiating elements has been in- ment of the geometry is performed alleviating the effects pro-
serted). The geometrical data of the manufactured element are duced by the resonances [17]. Depending on the phase values
summarized in Table I. The thickness values correspond to that required at the reflectarray elements, the phase-shifter is chosen
measured in the materials, because sometimes there are discrep- with a simple line segment or with a U-shaped TTD line.
ancies with respect to nominal values, particularly for Rohacell The phase and amplitude of the reflection coefficient provided
(layer 3). by a reflectarray element with the characteristics of Table II are
2498 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 56, NO. 8, AUGUST 2008

Fig. 3. Reflection coefficient for reflectarray element as a function of line

length for normal incidence, =0 ,' =0 . (a) Phase. (b) Amplitude. Fig. 4. Reflection coefficient for reflectarray element as a function of line
= 30
length for oblique incidence,  ,' = 40 . (a) Phase. (b) Amplitude.

shown in Fig. 3 as a function of the line-length for different fre-

are the coordinates of the element . The phase distribution is
quencies when a plane-wave impinges at normal incidence. The
usually truncated to a 360 phase-shift cycle to simplify the im-
phase variation covers the 3 360 phase range required for the
plementation of the reflectarray elements. However, the effect
phase-delay distribution and it is compared with an ideal delay
of spatial phase delay can be overcome if the phase is expanded
line, in which the phase should be , where is the propa-
to several 360 cycles.
gation constant at the corresponding frequency and L the length
Fig. 5 shows the phase distribution at 9.65 GHz to produce
of the delay line. The only drawback of delay lines is that a
a pencil beam in the direction for an el-
longer line produces a higher ohmic loss, as shown in Fig. 3(b).
liptic reflectarray antenna made up of 1888 elements arranged
In the present case, the losses have maximum values around 1
in a 50 48 grid, assuming the feed in the position
dB at the line resonances. The overall dissipative losses in a re-
mm, mm, mm, with respect to the
flectarray having patches of all dimensions are expected to be
reflectarray center in two cases: (a) the phase truncated to 360
less than 0.5 dB. These losses can be reduced if low-loss ma-
and (b) the phase expanded to three cycles of 360 . The reflec-
terials (as RT-Duroid) are used for layers d(1) and d(4). As the
tarray element shown in Fig. 1 with the geometrical data sum-
reflection response depends on the incidence angle, Fig. 4 shows
marized in Table II is used to implement the phase distributions.
the reflection behavior when the wave impinges obliquely with
The design of the reflectarray, for each phase distribution,
an angle , . As expected, slight variations in
consists of determining the length of the delay line in each ele-
phase and amplitude are produced when the angle of incidence
ment to achieve the phase distributions represented in Fig. 5.
changes. The reflected wave on each element includes the field
at central frequency. The feed horn is modeled by a
reradiated directly by the patch, and the field reflected by the
function to determine the level of illumination at each radi-
structure.
ating element. Once the reflectarray’s geometrical parameters
are defined which include aperture, patch, line width and di-
III. REFLECTARRAY DESIGN electric layers, feed position, and incidence angles, the required
Considering a reflectarray on the - plane illuminated by phase distribution is computed independently for both cases:
a feed horn, the required phase distribution, , at each phase-shift limited to one 360 cycle and phase-delay without
element of the array to collimate a beam in the ( , ) direction truncation, see Fig. 5. The reflection coefficient of the impinging
is determined as wave is calculated at each element localization for an initial
line length value. If the phase obtained for this line length is
(1) greater than the required phase defined by the phase distribu-
tion, the length of the line is incremented with and an it-
where is the propagation constant in vacuum, is the dis- erative loop begins until the phase value is less than the re-
tance from the feed horn to the element of the array and quired phase. If the phase achieved with the line length is less
CARRASCO et al.: BANDWIDTH IMPROVEMENT IN LARGE REFLECTARRAYS BY USING TRUE-TIME DELAY 2499

Fig. 5. Phase distribution required at the reflectarray surface. (a) Using a 360
2
phase-shift range. (b) Using phase delay (3 360 ). Fig. 6. One quarter of the masks layout for the layer containing the delay lines.
(a) Truncated phase distribution. (b) True-time delay distribution.

than the required phase defined by the phase distribution, an

error function is defined and a numerical method (false posi- jump while in the second case the line length increases without
tion, Newton–Raphson) is used to adjust the length of the line restriction, forming a U-shaped line.
until the error is below a threshold. In each iteration, the phase is The radiation patterns for the designed reflectarrays were
computed by the analysis technique described previously, which computed from the phase and amplitude of the reflected field at
takes into account the incidence angle of the impinging wave, each radiating element. The radiation patterns are computed in
the losses produced in the dielectric layers and the coupling ef- gain, dividing by the total power radiated by the horn, and taking
fects between the elements of the array by assuming local peri- into account ohmic losses, losses produced by phase errors and
odicity. Using the line lengths obtained in the design process, an spillover. The 2-D gain patterns for the TTD at 9.65 GHz are
Autocad command file is generated to produce the mask layout. shown respectively for the copolar and cross-polar components
The design technique has been applied to one linear polar- in Fig. 7.
ization but it is easy to achieve two linear polarizations if the In order to compare the reflectarray behavior using the two
position of the slots and the lines are redefined to allow enough phase distributions, Figs. 8 and 9 show, respectively, the radi-
room for a second orthogonal slot and line. ation patterns in u and v coordinates
, for both cases, truncated and unlimited phase distri-
butions at central frequency. In both figures, copolar and cross-
IV. RESULTS
polar components are shown. As can be seen, the pattern is prac-
Two reflectarrays with the same dimensions have been de- tically the same for the two phase distributions at the central
signed using the previously described technique to implement frequency. A pencil beam with a maximum gain of 36.3 dB
the phase distributions shown in Fig. 5. One quarter of the masks has been obtained with a cross-polar radiation 30 dB below the
for the layer containing the delay lines are shown in Fig. 6 for maximum gain, which is produced by the square patches. For
both cases, truncated phase distribution and TTD distribution. frequencies above and below 9.65 GHz, the beam distortion is
In the first case, the printed lines are reduced with every phase larger for the reflectarray with phase-shift limited to 360 . This
2500 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 56, NO. 8, AUGUST 2008

Fig. 9. Radiation pattern comparison in the plane u = sin(18 ) at central

frequency (9.65 GHz).

Fig. 7. Radiation patterns for the TTD reflectarray at 9.65 GHz. (a) Copolar
(b) Cross-polar.

Fig. 10. Radiation pattern comparison in the XZ plane (v = 0) at 8.65 GHz.

Fig. 8. Radiation pattern comparison in the XZ plane (v = 0) at central fre-

quency (9.65 GHz).

can be seen in Figs. 10 and 11 at 8.65 GHz, and in Figs. 12 Fig. 11. Radiation pattern comparison in the plane u = sin(18 ) at 8.65 GHz.
and 13 at 10.65 GHz. In the 8.65-GHz case, a sidelobe around
19 dB is overlapping with the main beam for the truncated phase
pattern, while the TTD pattern basically conserves its form and motivated because in the TTD case, the differential spatial phase
only a small sidelobe appears for the u plane 20 dB below the delay has been compensated and the only source of phase errors
maximum. In the 10.65-GHz case, a beam squint appears in the is the nonlinear behavior of the radiating element at extreme
u plane for the reflectarray with phase limitation. This is mainly frequencies.
CARRASCO et al.: BANDWIDTH IMPROVEMENT IN LARGE REFLECTARRAYS BY USING TRUE-TIME DELAY 2501

Fig. 12. Radiation pattern comparison in the XZ plane (v = 0) at 10.65 GHz. Fig. 15. Efficiency factor for the two designed reflectarrays compared with
ideal phases case.

phase distributions. The gain curves corresponding to two the-

oretical phase distributions have been included, as references.
The first ideal curve, represents the gain in the 8.5–11.90 GHz
band obtained from the ideal truncated phase distribution at
9.65 GHz, shown in Fig. 5(a). This case is equivalent to the de-
sign of a reflectarray with 360 phase truncation, using ideal re-
flectarray elements that produce the same phase-shift at every
frequency. In this case, the phase distribution on the reflec-
tarray remains fixed when frequency varies; however, the re-
quired phase to produce a focused beam should vary with fre-
quency proportionally to the true-time delay, as in the case of
parabolic reflectors. The second ideal curve represents the gain
at each frequency corresponding to the ideal phase distribu-
tion, compensated using TTD. Therefore, both ideal curves co-
Fig. 13. Radiation pattern comparison in the plane u = sin(18 ) at
incide at the central frequency and their difference represents the
diminution of gain produced by the effect of differential phase
10.65 GHz.
delay in reflectarrays. Note that the behavior of the gain for the
reflectarray with truncated phases is similar as the one of the
ideal phase at 9.65 GHz, being the reduction in gain for the re-
flectarray mainly produced by the losses in the dielectric mate-
rials (0.5 dB) and by some distortion of the element phasing at
extreme frequencies, as can be seen in Figs. 3(a) and 4(a). On
the other hand, the gain curve of the TTD reflectarray should
be similar as the one obtained for the ideal compensated phase
distribution, which is truth from 9.2 to 10.8 GHz. In this fre-
quency range, the TTD curve is similar to the ideal case with
a reduction in gain less than 1 dB, as a result of ohmic losses
and small phase errors out of central frequency. For extreme
frequencies, the gain drops because the phasing produced by
the elements are much different that the ideal case, as shown
in Figs. 3(a) and 4(a). These curves show clearly that the intro-
duction of TTD lines significantly increases the bandwidth of
the reflectarray. For example, the bandwidth for a 0.3-dB gain
Fig. 14. Gain comparison for the two designed reflectarrays and the gain using variation is 10.1% for the reflectarray with truncated phase and
ideal phases. 20.0% for the TTD reflectarray. If the gain variation is limited
to 1.5 dB below the maximum, the bandwidth increases from
17.8% to 26.7% for the TTD reflectarray.
Fig. 14 shows the antenna gain as a function of frequency for Fig. 15 shows the antenna efficiency expressed as the ratio
the two designed reflectarrays, based on truncated and unlimited of the gain and the theoretical directivity of an aperture. As
2502 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 56, NO. 8, AUGUST 2008

Fig. 16. Gain comparison for different reflectarrays size. Fig. 17. Bandwidth comparison, at 0.3 dB, between a phase-shift limited to
360 and TTD for different reflectarray sizes.

TABLE III
REFLECTARRAY FEATURES

can be seen, the designed reflectarrays allow 60% efficiency

for the working band. When the frequency is varied, the effi-
ciency for the TTD reflectarray remains at least 10% over the
efficiency achieved by a phase distribution limited to one 360
cycle. The antenna efficiency shown in Fig. 15 for the designed
reflectarrays takes into account ohmic losses, losses produced
by phase errors, spillover and antenna illumination. The antenna
efficiency for the ideal phase distribution, which is above 70%,
only includes spillover and antenna illumination.
The effect of different path lengths is only significant for very Fig. 18. Bandwidth comparison, at 1.5 dB, between a phase-shift limited to
large electrical dimensions, as in the case of antennas for space 360 and TTD for different reflectarray sizes.
applications. With the purpose of obtaining a bandwidth curve
as a function of the reflectarray size, different reflectarrays have
been designed to generate a pencil beam in the same direction. rays and can be compensated through the introduction of TTD
The reflectarrays have been designed from the smallest case lines. Figs. 17 and 18 show that, in conventional reflectarrays,
when only one 360 cycle is required to the 3 360 previously the bandwidth decreases as the electrical size of the antenna in-
designed case. The bandwidth decreases as the reflectarray size creases. On the other hand, the relative bandwidth in the case of
increases, as expected. This decrease is more important when TTD converges asymptotically for large reflectarrays, because
the phase distribution is limited to 360 . Fig. 16 shows the gain bandwidth is only limited by the reflectarray element and not
for the three reflectarrays described in Table III. As can be seen, by its size. These results show that a 20% bandwidth can be
the gain provided by the RA 1 provides a highest bandwidth achieved for reflectarrays with aperture dimensions larger than
because only one 360 cycle is required at the phase distri- 25 wavelengths.
bution and the differential spatial phase delay does not exist.
Fig. 17 shows the relative bandwidth in percentage for a vari- V. CONCLUSION
ation in gain of 0.3 dB, as a function of the reflectarray elec- Several reflectarray antennas have been designed in the
trical size for both cases: phase distribution is limited to 360 X-band, considering two cases for the phase distribution. In the
and TTD. The bandwidth of a reflectarray with phase distribu- first case, the phase distribution has been truncated to 360 and
tion limited to 360 behaves in a similar way to that of a zoned in the second case the antennas have been designed without
lens antenna. Fig. 18 shows the bandwidth for a reduction in phase limitation using TTD. A significant improvement in
gain of 1.5 dB. As can be seen in both figures, the errors pro- bandwidth has been demonstrated when TTD is used where the
duced by the nonconstant path from the feed horn to each el- only limitation is imposed by the bandwidth of the radiating
ement of the reflectarray are more important in large reflectar- element. The effect of the differential spatial phase delay is
CARRASCO et al.: BANDWIDTH IMPROVEMENT IN LARGE REFLECTARRAYS BY USING TRUE-TIME DELAY 2503

more important in large reflectarrays as has been demonstrated. Eduardo Carrasco (S’96) was born in Mexico
This type of antenna can be used in beam scanning applications, City, Mexico. He received the Telecommunication
Engineer degree from the Universidad Nacional Au-
if some switches are included in the delay lines. tonoma de Mexico (UNAM), Mexico City, in 2000.
He is currently working toward the Ph.D. degree at
the Universidad Politécnica de Madrid, Spain.
REFERENCES From 1999 to 2001 he was with Directv Latin
[1] J. Huang, “Bandwidth study of microstrip reflectarray and a novel America as a Broadcast Operation System Spe-
cialist. In 2002 he received a grant from Fundacion
phased reflectarray concept,” in Proc. IEEE Int. Symp. Antennas
Vodafone, Spain, to obtain a Telecommunication
Propag., Newport Beach, CA, Jun. 1995, pp. 582–585.
Management Degree at the Escuela de Organizacion
[2] D. M. Pozar, “Bandwidth of reflectarrays,” Electron. Lett., vol. 39, no.
Industrial (EOI), Spain. Since 2003 he has been with the Electromagnetism
21, pp. 1490–1491, Oct. 2003. and Circuits Theory Department, Universidad Politécnica de Madrid, Spain.
[3] J. A. Encinar, “Design of two-layer printed reflectarrays using patches From January to April 2008, he visited the Microwave Engineering Laboratory,
of variable size,” IEEE Trans. Antennas Propag., vol. 49, no. 10, pp. Università degli Studi di Perugia, Italy. His research interests include printed
1403–1410, Oct. 2001. reflectarrays, multilayer periodic structures, and space antennas.
[4] J. A. Encinar and J. A. Zornoza, “Three-Layer printed reflectarrays for
contoured beam space applications,” IEEE Trans. Antennas Propag.,
vol. 52, no. 5, pp. 1138–1148, May 2004.
[5] N. Misran, R. Cahill, and V. Fusco, “Design optimisation of ring el- José A. Encinar (S’81–M’86) was born in Madrid,
ements for broadband reflectarray antennas,” Proc. Inst. Elect. Eng., Spain. He received the Electrical Engineer and Ph.D.
Microw., Antennas, Propag., vol. 150, no. 6, pp. 440–444, Dec. 2003. degrees from the Universidad Politécnica de Madrid
[6] M. R. Chaharmir, J. Shaker, M. Cuhaci, and A. Ittipiboon, “Broadband (UPM), in 1979 and 1985, respectively.
reflectarray antenna with double cross loops,” Electron. Lett., vol. 42, Since January 1980 has been with the Applied
no. 2, pp. 65–66, Jan. 2006. Electromagnetism and Microwaves Group at UPM,
[7] M. Bozzi, S. Germani, and L. Perregrini, “Performance comparison of as a Teaching and Research Assistant from 1980 to
different element shapes used in printed reflectarrays,” Ante. Wireless 1982, as an Assistant Professor from 1983 to 1986,
Propag. Lett., vol. 2, no. 1, pp. 219–222, 2003. and as Associate Professor from 1986 to 1991. From
[8] D. M. Pozar, “Wideband reflectarrays using artificial impedance sur- February to October of 1987 he stayed at Polytechnic
faces,” Electron. Lett., vol. 43, no. 3, pp. 148–149, Feb. 2007. University, Brooklyn, NY, as a Postdoctoral Fellow
[9] J. A. Encinar and J. A. Zornoza, “Broadband design of three-layer of the NATO Science Program. Since 1991 he has been a Professor of the
printed reflectarrays,” IEEE Trans. Antennas Propag., vol. 51, no. 7, Electromagnetism and Circuit Theory Department at UPM. In 1996 he was with
pp. 1662–1664, Jul. 2003. the Laboratory of Electromagnetics and Acoustics at teh Ecole Polytechnique
[10] A. Bhattacharyya, “Slot-coupled patch reflect array element for en- Fédérale de Lausanne, Switzerland, as Visiting Professor. His research interests
hanced gain-bandwitdh performance,” U.S. 6 388 620, 2000. include numerical techniques for the analysis of multilayer periodic structures,
[11] M. E. Bialkowski and H. J. Song, “Dual linearly polarized reflectarray design of frequency-selective surfaces, printed arrays, and reflectarrays. He has
using aperture coupled microstrip patches,” in Proc. Antennas Propag. published more than 100 journal and conference papers, and he is holder of
Soc. Int. Symp., Jul. 2001, pp. 486–489. three patents on array and reflectarray antennas.
[12] S. M. Duffy and S. D. Targonski, “Comparison of two flat reflector-type Prof. Encinar was a corecipient of the 2005 H. A. Wheeler Applications Prize
designs for dual-polarization, dual-band operation,” in Proc. Antennas Paper Award and the 2007 S. A. Schelkunoff Transactions Prize Paper Award,
given by IEEE Antennas and Propagation Society.
Propag. Soc. Int. Symp., Jul. 2001, pp. 288–291.
[13] T. A. Milligan, Modern Antenna Design, 2nd ed. Piscataway, NJ:
IEEE Press, 2005, pp. 447–473.
[14] D. G. Bodnar, “Lens antennas,” in The Handbook of Antenna Design.
London, U.K.: Peter Peregrinus, 1986, vol. 1 and 2, ch. 16, pp. 1–31. Mariano Barba was born in Murcia, Spain. He re-
[15] D. C. Chang and M. C. Huang, “Microstrip reflectarray antenna with ceived the Ingeniero de Telecomunicación and Ph.D.
offset feed,” Electron. Lett., vol. 28, no. 16, pp. 1489–1491, 1992. degrees from the Universidad Politécnica de Madrid
(UPM), Spain, in 1990 and 1996, respectively.
[16] E. Carrasco, B. Alfageme, and J. A. Encinar, “Design of a multilayer
From 1991 to 1994 he was with the Departamento
aperture-coupled cell used as phase-shifter in reflectarrays,” J. Int. Nice
de Electromagnetismo y Teoría de Circuitos at UPM,
Antennes, pp. 194–195, 2004. as a Postgraduate Researcher. During this time he de-
[17] E. Carrasco, M. Barba, and J. A. Encinar, “Reflectarray element veloped numerical methods and algorithms to charac-
based on aperture-coupled patches with slots and lines of variable terize and design microwave passive devices. From
length,” IEEE Trans. Antennas Propag., vol. 55, no. 3, pp. 820–825, 1994 to 2003 he was involved in the R&D depart-
Mar. 2007. ments of several companies and institutions devel-
[18] C. H. Wan and J. A. Encinar, “Efficient computation of generalized oping antennas and microwave circuits for space and terrestrial communication
scattering matrix for analyzing multilayer periodic structures,” IEEE applications. In 2003 he rejoined the Departamento de Electromagnetismo y
Trans. Antennas Propag., vol. 43, no. 11, pp. 1233–1242, Nov. 1995. Teoría de Circuitos as an Associate Professor. His research interests include the
[19] E. Carrasco, M. Barba, and J. A. Encinar, “Aperture-coupled reflec- analysis, characterization, and design of antenna and microwave circuits. He is
tarray element with wide range of phase delay,” Electron. Lett., vol. currently involved in the development and design of multibeam and reconfig-
42, no. 12, pp. 667–668, 2006. urable antennas.