See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/341112947

RF-MEMS Monolithic K and Ka Band Multi-State Phase Shifters as Building

Blocks for 5G and Internet of Things (IoT) Applications

Article  in  Sensors · May 2020

DOI: 10.3390/s20092612

CITATIONS READS

0 2

9 authors, including:

Jacopo Iannacci A. Bagolini

Fondazione Bruno Kessler Fondazione Bruno Kessler
178 PUBLICATIONS   1,002 CITATIONS    94 PUBLICATIONS   897 CITATIONS   

SEE PROFILE SEE PROFILE

Flavio Giacomozzi Elena Bochkova

Fondazione Bruno Kessler Institute of Fundamental Electronics
185 PUBLICATIONS   1,263 CITATIONS    21 PUBLICATIONS   39 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Integrated High Efficiency Class-E PAs View project

High quality factor resonances View project

All content following this page was uploaded by Jacopo Iannacci on 03 May 2020.

The user has requested enhancement of the downloaded file.

 sensors
Article
RF-MEMS Monolithic K and Ka Band Multi-State
Phase Shifters as Building Blocks for 5G and Internet
of Things (IoT) Applications
Jacopo Iannacci 1, * , Giuseppe Resta 1 , Alvise Bagolini 1 , Flavio Giacomozzi 1 ,
Elena Bochkova 2 , Evgeny Savin 2 , Roman Kirtaev 2 , Alexey Tsarkov 2 and Massimo Donelli 3
1 Center for Materials and Microsystems (CMM), Fondazione Bruno Kessler (FBK), Via Sommarive,
18-38123 Trento, Italy; resta@fbk.eu (G.R.); bagolini@fbk.eu (A.B.); giaco@fbk.eu (F.G.)
2 Bazovye Tekhnologii, LLC, 125124 Moscow, Russia; elena.bochkova@btlabs.ru (E.B.); e.savin@btlabs.ru (E.S.);
roman.kirtaev@phystech.edu (R.K.); a.tsarkov@btlabs.ru (A.T.)
3 Department of Information Engineering and Computer Science, University of Trento, 38123 Trento, Italy;
massimo.donelli@unitn.it
* Correspondence: iannacci@fbk.eu; Tel.: +39-0461-314-441




Received: 1 April 2020; Accepted: 2 May 2020; Published: 3 May 2020


Abstract: RF-MEMS, i.e., Micro-Electro-Mechanical Systems (MEMS) for Radio Frequency (RF)
passive components, exhibit interesting characteristics for the upcoming 5G and Internet of Things
(IoT) scenarios, in which reconfigurable broadband and frequency-agile devices, like high-order
switching units, tunable filters, multi-state attenuators, and phase shifters will be necessary to enable
mm-Wave services, small cells, and advanced beamforming. In particular, satellite communication
systems providing high-speed Internet connectivity utilize the K and Ka bands, which offer larger
bandwidth compared to lower frequencies. This paper focuses on two design concepts of multi-state
phase shifter designed and manufactured in RF-MEMS technology. The networks feature 4 switchable
stages (16 states) and are developed for the K and Ka bands. The proposed phase shifters are realized
in a surface micromachining RF-MEMS technology and the experimentally measured parameters are
compared with Finite Element Method (FEM) multi-physical electromechanical and RF simulations.
The simulated phase shifts at both the operating bands fit well the measured value, despite the
measured losses (S21) are larger than 5–7 dB if compared to simulations. However, such a non-ideality
has a technological motivation that is explained in the paper and that will be fixed in the manufacturing
of future devices.

Keywords: RF-MEMS; MEMS; RF passives; phase shifters; multi-state passive networks; 5G;
Internet of Things (IoT); wideband operability

1. Introduction
RF-MEMS, i.e., Micro-Electro-Mechanical Systems (MEMS) for Radio Frequency (RF) applications
have been around for a long time. First discussed in scientific papers after the mid-1990s [1,2],
MEMS-based RF passives quickly ignited huge expectations in terms of market breakthroughs,
primarily because of their remarkable and unprecedented electromagnetic characteristics, like wideband
operability, large reconfigurability/tunability, very-low loss, high-isolation, high Q-factor [3].
Despite the initial market expectations of RF-MEMS were disappointed, the context started to
change quite radically after 2010 when smartphones started to become increasingly more popular.
The increasing integration of devices in the handsets led to a progressive worsening of the quality
of communications [4,5]. Analog impedance matching tuners based on RF-MEMS technology were
the first in line solution available [6], which made possible solving the problem, and, on the other

Sensors 2020, 20, 2612; doi:10.3390/s20092612 www.mdpi.com/journal/sensors

Sensors 2020, 20, x FOR PEER REVIEW 2 of 14
Sensors 2020, 20, 2612 2 of 13

first in line solution available [6], which made possible solving the problem, and, on the other hand,
to score
hand, to the
score first
thesuccessful exploitation
first successful exploitation of RF-MEMS
of RF-MEMS in mass-market
in mass-market applications.
applications. Moreover,
Moreover, the
spread of RF-MEMS-based reconfigurable passive networks
the spread of RF-MEMS-based reconfigurable passive networks in the consumer segment is pulling in the consumer segment is pulling
applied research
applied researchand andengineering
engineering of of other
other components
components based based on Microsystem
on Microsystem technology,
technology, with
with robust
robust switches and switching units
switches and switching units first in line [7,8]. first in line [7,8].
Besides the
Besides the abovementioned
abovementionedcase, case,another
another intriguing
intriguing application
application field for for
field high-performance
high-performance RF-
RF-MEMS is gaining shape, which is the upcoming fifth generation of mobile communications, i.e., 5G,
MEMS is gaining shape, which is the upcoming fifth generation of mobile communications, i.e., 5G,
with all
with all its
its implications
implications and and spillovers
spillovers in in the
the Internet
Internet of of Things
Things (IoT)
(IoT) scenario.
scenario.
Active development
Active development of of 5G
5G technologies
technologies and and satellite
satellite communications
communications demand demand for for high-speed
high-speed
stable data transfer services from any location. Satellite systems
stable data transfer services from any location. Satellite systems providing high-speed Internet providing high-speed Internet
connectivity utilize K and Ka frequency bands offering a
connectivity utilize K and Ka frequency bands offering a large bandwidth compared to lower large bandwidth compared to lower
frequencies. Conventionally,
frequencies. Conventionally, signal signal reception
reception and and transmission
transmission are are carried
carried out out through
through two two separate
separate
frequency channels in the K and Ka frequency bands, which,
frequency channels in the K and Ka frequency bands, which, in turn, increase the demand for in turn, increase the demand for the
the
high-performance components
high-performance components intended
intended for for signal
signal control
control of of the
the transceiver
transceiver modules.
modules.
In light of this scenario, this work presents and discusses a couple design
In light of this scenario, this work presents and discusses a couple of of design concepts
concepts of RF-
of
RF-MEMS-based reconfigurable phase shifters. From the architectural point of view, we relied the
MEMS-based reconfigurable phase shifters. From the architectural point of view, we relied on on
True-Time
the True-Time Delay (TTD)
Delay topology,
(TTD) whichwhich
topology, is well-known
is well-knownin the literature. Given the Given
in the literature. field oftheapplication
field of
application mentioned above, we tailored the design of the network in such a way to have frequency
mentioned above, we tailored the design of the network in such a way to have the center the center
of operation
frequency at 19.5 GHz
of operation at 19.5andGHz 29.3
andGHz, i.e., in
29.3 GHz, i.e.,the K and
in the K and KaKaband,
band,respectively.
respectively. Concerning
Concerning
reconfigurability, we opted for 4 stages (i.e., 16 configurations), controlled
reconfigurability, we opted for 4 stages (i.e., 16 configurations), controlled by electrostatically driven by electrostatically driven
ohmic series RF-MEMS micro-relays. In particular, the K band
ohmic series RF-MEMS micro-relays. In particular, the K band phase shifter employs a classical phase shifter employs a classical
transversal clamped-clamped
transversal clamped-clampedswitch switchtopology.
topology.On Onthetheother
other hand,
hand, thethe
KaKa band
band phase
phase shifters,
shifters, givengiven
its
its higher operating frequency, imposes more stringent requirements
higher operating frequency, imposes more stringent requirements in terms of reduced network footprint. in terms of reduced network
footprint.
Having in Having
mind such in mind such
a driver, wea developed
driver, weand developed
implemented and implemented
an innovativeanin-line innovative in-line
longitudinal
longitudinalswitch
RF-MEMS RF-MEMS design, switch design, conjugating
conjugating the compactness the compactness
typical oftypical of cantilevered
cantilevered (single-
(single-hinged)
hinged) micro-relays, with the mechanical robustness and resilience
micro-relays, with the mechanical robustness and resilience of clamped-clamped solutions. of clamped-clamped solutions.
The paper
The paper is is arranged
arranged as as follows.
follows. AfterAfter thethe current
current section
section framing
framing the the evolution
evolution of of RF-MEMS
RF-MEMS
and the
and the opportunities
opportunitiesof ofthe
the5G 5GandandIoT IoTapplication
application fields, thethe
fields, following
following Section
Section2 will introduce
2 will introduce the
reconfigurable phase shifters design concepts, along with modeling
the reconfigurable phase shifters design concepts, along with modeling and simulations. Section 3 and simulations. Section 3 will
develop
will developa comprehensive
a comprehensive discussion
discussion of ofthethe RF-MEMS
RF-MEMSphase phaseshifters,
shifters,leveraging
leveraging the acquired
the acquired
experimental data. Finally, Section 4 will collect some
experimental data. Finally, Section 4 will collect some conclusive considerations.conclusive considerations.

2. RF-MEMS
2. RF-MEMS Phase
Phase Shifter:
Shifter: Design
Design Concept
Concept and
and Simulations
Simulations
The RF-MEMS
The RF-MEMS phase
phase shifters
shifters concepts
concepts operating
operating atat K and Ka
K and Ka frequency
frequency bands
bands areare designed
designed
capitalizing on True-Time Delay (TTD) technology [9–15] and are manufactured
capitalizing on True-Time Delay (TTD) technology [9–15] and are manufactured in the surface in the surface
micromachining technology
micromachining technology available
available at at Fondazione
Fondazione Bruno
Bruno Kessler
Kessler (FBK)
(FBK) inin Italy
Italy [16].
[16]. The
The resulting
resulting
phase delay is defined by the lengths of the transmission lines. By selecting different
phase delay is defined by the lengths of the transmission lines. By selecting different propagating propagating
signal paths
signal paths in
in the
the circuit,
circuit, thethe total
total phase
phase shift
shift can
can be
be tuned
tuned byby aa predefined
predefined value.
value. The
The designed
designed
phase shifters
phase shiftersconsist
consistofof4 4sections
sections (i.e.,
(i.e., 4 bits),
4 bits), connected
connected in series.
in series. EachEach section
section has RF
has two twopaths
RF paths
with
with different lengths, as reported by the schematic
different lengths, as reported by the schematic in Figure 1. in Figure 1.

Figure 1. Top view of the 4-bit True-Time Delay (TTD) phase shifter schematic topology. The upper
sections
Figure 1.correspond to the 4-bit
Top view of reference path with
True-Time a total
Delay (TTD)0◦ phase shift.
shifter schematic topology. The upper
sections correspond to the reference path with a total 0° phase shift.
Sensors 2020, 20, 2612 3 of 13

The total phase shift is controlled by the combination of the four sections path lengths, selected
using RF-MEMS ohmic
Sensors 2020, 20, x FORswitches
PEER REVIEW monolithically integrated into the network. Once the 3 ofswitches
14 are
actuated (i.e., pulled-in), the RF signal travels through the selected delay line. Each bit can provide an
The total phase shift is controlled by the combination of the four sections path lengths, selected
additionalusing
22.5RF-MEMS
degrees ohmic phaseswitches
shift atmonolithically
the operating frequency.
integrated into theThus, a 4Once
network. bits the
phase shifter
switches are provides
24 = 16 different
actuated phase delays:the0◦RF
(i.e., pulled-in), signal◦travels
, ±22.5 , ±45◦through
, ±67.5the◦ , ±90 ◦ , ±112.5
selected ◦ , ±135
delay line. Each◦bit
, ±157.5 ◦ , and 180◦ at
can provide
an additional
19.5 GHz and 29.3 GHz 22.5center
degreesfrequencies,
phase shift at theforoperating
the K and frequency.
Ka band Thus, a 4 bits phase
specific designshifter provides
variant, respectively.
24 = 16 different phase delays: 0°, ±22.5°, ±45°, ±67.5°, ±90°, ±112.5°, ±135°, ±157.5°, and 180° at 19.5
The designed phase shifter is based on 50 Ω Coplanar Waveguide (CPW) transmission line and
GHz and 29.3 GHz center frequencies, for the K and Ka band specific design variant, respectively.
the time delayThe switch
designed option is shifter
phase basedison series-type
based electrostatically
on 50 Ω Coplanar Waveguide (CPW)driventransmission
RF-MEMSline ohmic
and switches.
Thus, the phase
the timeshiftdelayis switch
controlled
optionby is the applied
based voltageelectrostatically
on series-type signal to each of the
driven four bits.
RF-MEMS ohmic The K band
switches.
phase shifter Thus,
is fitted the phase
with shift is
switches controlled by
described by the
theapplied
authors voltage signalFigure
in [17]. to each2aof the four bits.
shows The
a schematic view
K band phase shifter is fitted with switches described by the authors in [17]. Figure 2a shows a
of the transverse RF-MEMS switching unit integrated into the CPW transmission line.
schematic view of the transverse RF-MEMS switching unit integrated into the CPW transmission line.

(a) (b)

Figure 2. Figure 2. Schematic

Schematic of theoftransverse
the transverse
(a)(a)and
and longitudinal
longitudinal (b)(b)
Micro-Electro-Mechanical Systems Systems
Micro-Electro-Mechanical
(MEMS) membrane integrated into the Coplanar Waveguide (CPW) transmission line and employed
(MEMS) membrane integrated into the Coplanar Waveguide (CPW) transmission line and employed in
in the K and Ka band phase shifters designs, respectively.
the K and Ka band phase shifters designs, respectively.
The Ka band device is almost twice smaller in size, so the same switch geometry cannot be fitted
The Kain itband
due todevice is almost
its dimensions and twice smaller
space needed for in size, so One
integration. the of
same switch
the most geometry
compact cannot be fitted
configurations
in it due tois its
related to cantilever-type
dimensions geometry
and space needed(i.e., single-hinged
for integration.membrane)
One of [18].
theHowever, from a general
most compact configurations
point of view, cantilevered MEMS switches are not as robust as clamped-clamped designs [19,20]. To
is related to cantilever-type geometry (i.e., single-hinged membrane) [18]. However, from a general
avoid reliability issues, a compact longitudinal membrane with complex anchor geometry has been
point of view, cantilevered
designed, as reportedMEMS switches
in Figure are not aswith
2b. This membrane robust as clamped-clamped
optimized characteristics allowsdesigns
lowering[19,20]. To
avoid reliability issues,
the actuation a compact
voltage longitudinal
and reducing membrane
the area occupation, whilewith complex
still relying anchor geometry
on double-hinged MEMS has been
designed, suspended
as reported membrane solution,
in Figure 2b. as for the
This transverse micro-switch.
membrane with optimized characteristics allows lowering
Both the longitudinal and transverse switch geometries were optimized to achieve a wide
the actuation voltage and reducing the area occupation, while still relying on double-hinged MEMS
operating frequency range in terms of low loss and high isolation. Table 1 shows that in the operating
suspendedrange membrane
0–40 GHz,solution, as for
the simulated thearetransverse
losses less than −0.8micro-switch.
dB and isolation is better than −20 dB.
Both the longitudinal and transverse switch geometries were optimized to achieve a wide operating
frequency range in terms of low loss and high isolation. Table 1 shows that in the operating range
0–40 GHz, the simulated losses are less than −0.8 dB and isolation is better than −20 dB.
Structural deformations due to the actuation voltage were simulated for 1.5 µm membrane
thickness and stiffening rib [16] of 2.5 µm at varied air-gaps between the membrane and the actuation
pads. Table 2 reports the applied actuation voltage depending on the air-gap for both membranes.
The results of electromechanical simulations for the minimum gap of 2 µm are illustrated in
Figure 3, in which the actuation voltage (VPI ) is 28 V for the transverse MEMS membrane and 38 V for
the novel compact longitudinal design concept.

Table 1. Main simulated radio frequency (RF) characteristics for the transverse and longitudinal Single
Pole Single Throw (SPST) RF-MEMS switches previously shown in Figure 2.

Parameter Transverse SPST Longitudinal SPST

Frequency range, GHz 0–40 0–40
Actuation voltage, V 28 38
−30 @ 15 GHz −30 @ 10 GHz
Isolation, dB
−20 @ 35 GHz −20 @ 35 GHz
Insertion loss, dB −0.8 @ 40 GHz −0.12 @ 40 GHz
 −30 @ 15 GHz −30 @ 10 GHz
Isolation, dB
−20 @ 35 GHz −20 @ 35 GHz
Insertion loss, dB −0.8 @ 40 GHz −0.12 @ 40 GHz

Structural deformations due to the actuation voltage were simulated for 1.5 µ m membrane
thickness and stiffening rib [16] of 2.5 µ m at varied air-gaps between the membrane and the actuation
Sensors 2020, 20,pads.
2612 Table 2 reports the applied actuation voltage depending on the air-gap for both membranes. 4 of 13

Table 2. Actuation (pull-in) voltage at different air-gaps for the transverse and longitudinal RF-MEMS
membranes previously shown in Figure 2.
Table 2. Actuation (pull-in) voltage at different air-gaps for the transverse and longitudinal RF-MEMS
Air-gap [µm] VPI [V]–Transverse VPI [V]–Longitudinal
membranes previously shown in2Figure 2. 28 38
2.5 37 52
Air-Gap [µm] 2.7 VPI [V]–Transverse
42 60 VPI [V]–Longitudinal
2 28 38
The results2.5of electromechanical simulations 37for the minimum gap of 2 µ m 52
are illustrated in
Figure 3, in which
2.7 the actuation voltage (VPI) is 28
42V for the transverse MEMS membrane
60 and 38 V
for the novel compact longitudinal design concept.

(a) (b)
Figure 3. Electromechanical simulated results for the transverse (a) and longitudinal (b) RF-MEMS
Figure 3. Electromechanical simulated results for the transverse (a) and longitudinal (b) RF-MEMS
membrane geometries with 2 µ m air-gap. The color scale indicates the membrane deformation.
membrane geometries with 2 µm air-gap. The color scale indicates the membrane deformation.
The RF performance of the phase shifters employing the just discussed micro-relays is simulated
The RF performance
by means of Finiteof Element
the phase shifters
Method (FEM)employing the 4just
analysis. Figures anddiscussed
5 show the micro-relays
reflection (S11), is simulated
insertion loss (S21), and phase delay for several configurations of the K and Ka band phase shifters,
by means of Finite Element Method (FEM) analysis. Figures 4 and 5 show the reflection (S11), insertion
respectively. The simulated losses at 19.5 GHz are better than −2 dB in a 2 GHz frequency span, while
loss (S21), and phaseisdelay
reflection betweenfor−40several configurations
and −25 dB of the(KKband
at the center frequency andphase
Ka band
shifter;phase shifters,
see Figure 4). respectively.
The simulated losses at 19.5 GHz are better than −2 dB in a 2 GHz frequency span, while reflection is
between −40 and −25 dB at the center frequency (K band phase shifter; see Figure 4).
Sensors 2020, 20, x FOR PEER REVIEW 5 of 14

(a)

(b)

(c)

Figure 4. Simulated
Figure 4. Simulated reflection
reflection (a), transmission
(a), transmission (b),
(b), andphase
and phase shift
shift(c)(c)forfor
several configurations
several of
configurations of the
the K band RF-MEMS reconfigurable phase shifter (center frequency at 19.5 GHz).
K band RF-MEMS reconfigurable phase shifter (center frequency at 19.5 GHz).
On the other hand, the device working at 29.3 GHz has simulated losses better than −2.4 dB for
most configurations, despite at −90 and +90 phase shifts, −2.6 dB losses are observed at the maximum
limit of 2 GHz frequency range (Ka band phase shifter; see Figure 5).
 Figure 4. Simulated reflection (a), transmission (b), and phase shift (c) for several configurations of
the K band RF-MEMS reconfigurable phase shifter (center frequency at 19.5 GHz).

On the other hand, the device working at 29.3 GHz has simulated losses better than −2.4 dB for
most
Sensorsconfigurations,
2020, 20, 2612 despite at −90 and +90 phase shifts, −2.6 dB losses are observed at the maximum
5 of 13
limit of 2 GHz frequency range (Ka band phase shifter; see Figure 5).

(a)

Sensors 2020, 20, x FOR PEER REVIEW 6 of 14

(b)

(c)

Figure 5.
Figure Simulated reflection (a),
5. Simulated (a), transmission
transmission(b),
(b),and
andphase
phaseshift
shift(c)(c)
forfor
several configurations
several of the
configurations of
Ka band RF-MEMS reconfigurable phase shifter (center frequency at 29.3
the Ka band RF-MEMS reconfigurable phase shifter (center frequency at 29.3 GHz). GHz).

Onconclusive
As the other hand, the device
consideration working
of this at 29.3
section, GHz
it must behas simulatedthat
highlighted losses better
most than
of the −2.4
losses dBdue
are for
most
to configurations,
microswitches, despite
as it will be −90 and +90
at discussed phase
more shifts, −2.6
in technical dB losses
details laterare
in observed
this paper,at especially
the maximumfor
limit of 2 GHz frequency range (Ka band phase shifter; see Figure
what concerns the non-idealities exhibited by the tested physical samples. 5).
As conclusive consideration of this section, it must be highlighted that most of the losses are due
3.toExperimental
microswitches,Results
as it will
andbe Discussion
discussed more in technical details later in this paper, especially for what
concerns the non-idealities exhibited by the tested physical samples.
In this section, an extensive experimental characterization is reported, both for what concerns
the RF-MEMS
3. Experimental phase shifters
Results physical samples, as well as for the standalone micro-switches they
and Discussion
employ. The mentioned testing comprises the analysis of the RF characteristics and the
In this section, an extensive experimental characterization is reported, both for what concerns the
electromechanical behavior of the available RF-MEMS samples.
RF-MEMS phase shifters physical samples, as well as for the standalone micro-switches they employ.
The S-parameters (scattering parameters) are measured using a probe station equipped with
The mentioned testing comprises the analysis of the RF characteristics and the electromechanical
Ground-Signal-Ground (GSG) CPW microprobes (250 μm pitch), a Vector Network Analyzer (VNA)
behavior of the available RF-MEMS samples.
with an operating frequency range from 10 kHz up to 40 GHz and a voltage source. The VNA is
calibrated with the SOLT (Short Open Load Thru) method using an on-wafer GSG calibration kit. The
discussed phase shifters require actuating four DC signal pads and one ground (GND) pad. Thus, we
exploit a setup featuring five DC probes and two RF probes to fully characterize the physical devices.
The probes setup is shown in Figure 6.
Sensors 2020, 20, 2612 6 of 13

The S-parameters (scattering parameters) are measured using a probe station equipped with
Ground-Signal-Ground (GSG) CPW microprobes (250 µm pitch), a Vector Network Analyzer (VNA)
with an operating frequency range from 10 kHz up to 40 GHz and a voltage source. The VNA is
calibrated with the SOLT (Short Open Load Thru) method using an on-wafer GSG calibration kit. The
discussed phase shifters require actuating four DC signal pads and one ground (GND) pad. Thus, we
exploit a setup featuring five DC probes and two RF probes to fully characterize the physical devices.
The probes
Sensors setup
2020, 20, x FORisPEER
shown in Figure 6.
REVIEW 7 of 14
Sensors 2020, 20, x FOR PEER REVIEW 7 of 14

. .
Figure 6. Photograph of the measurement setup featuring both RF and DC microprobes.
Figure 6. Photograph of the measurement setup featuring both RF and DC microprobes.
In theFigure 6. Photograph
first place, of the measurement
the attention is concentratedsetup featuring both RF andRF-MEMS
on the standalone DC microprobes.
switching unit
electromagnetic characteristics.
In the first In this is
place, the attention regard, the transverse
concentrated and longitudinal
on the standalone RF-MEMSohmic micro-relays
switching unit
Inelectromagnetic
previouslythe discussed
first place, inthe attention
Section 2 and
characteristics. isreported
In this concentrated
regard, in on the
theFigures standalone
2 and
transverse RF-MEMS
and3 longitudinal
are measured up to
ohmic switching
40 GHz. unit
micro-relays The
electromagnetic
previously characteristics.
discussed in In
Section this
2 and regard,
reportedthein transverse
Figures 2 andand
3 longitudinal
are measured ohmic
up
following Figures 7 and 8 show the comparison of measured and simulated S-parameters of transverse to 40 micro-relays
GHz. The
previously discussed
following
and longitudinal Figures in7 Section
switches, and 2 andthe
8 show
respectively, reported in Figures
in comparison
the OFF andofON 2state.
and 3 and
measured are measured up to 40 GHz.
simulated S-parameters of The
transverse and longitudinal switches, respectively, in the OFF and ON
following Figures 7 and 8 show the comparison of measured and simulated S-parameters ofstate.
transverse and longitudinal switches, respectively, in the OFF and ON state.

(a) (b)

(a) (b)

(c) (d)
Figure 7. Measured versus simulated S-parameters characteristics of the transverse ohmic switch
Figure 7. Measured versus simulated S-parameters characteristics of the transverse ohmic switch
reported in Figures 2a and 3a with reference to (a) reflection (S11) in the OFF state; (b) isolation (S21)
reported in OFF
in the Figures 2a(c)
state; and 3a with(S11)
reflection reference to (a)
in the ON reflection
state; (S11) in the
(d) transmission OFF
(S21) state;
in the ON(b) isolation (S21) in
state.
(c) (S11) in the ON state; (d) transmission (S21) in the(d)
the OFF state; (c) reflection ON state.
Figure 7. Measured versus simulated S-parameters characteristics of the transverse ohmic switch
reported in Figures 2a and 3a with reference to (a) reflection (S11) in the OFF state; (b) isolation (S21)
in the OFF state; (c) reflection (S11) in the ON state; (d) transmission (S21) in the ON state.
Sensors 2020, 20, 2612 7 of 13
Sensors 2020, 20, x FOR PEER REVIEW 8 of 14

(a) (b)

(c) (d)
Figure
Figure8.8. Measured
Measuredversus
versussimulated
simulatedS-parameters
S-parameterscharacteristics
characteristicsof ofthe
thelongitudinal
longitudinalohmic ohmicswitch
switch
reported
reportedinin Figures
Figures 2b and 3b with reference
reference to
to(a)
(a)reflection
reflection(S11)
(S11)ininthe
theOFF
OFFstate;
state;(b)
(b)isolation
isolation(S21)
(S21)
in
in the
the OFF
OFF state;
state; (c)(c) reflection
reflection (S11)
(S11) inin the
the ONON state;
state; (d)(d) transmission
transmission (S21)
(S21) in in
thethe
ONON state.
state.

Concerningthe
Concerning theisolation
isolation(S21)
(S21) ininthetheOFFOFF state,
state, thethe simulated
simulated traces
traces predict
predict quitequite accurately
accurately the
the experimental behavior, as visible in Figures 7b and 8b. A
experimental behavior, as visible in Figures 7b and 8b. A more articulated discussion has to be more articulated discussion has to
be developed
developed for forthe theONON state
state of of
bothboth switchesasasthe
switches thedisagreement
disagreementbetween betweenmeasurements
measurements and and
simulationsisisrelevant.
simulations relevant.In Inparticular,
particular,the themeasured
measuredreflection
reflection(S11) (S11)isisworse
worsethanthansimulations
simulationsfor forboth
both
switches,as
switches, asemerged
emergedin inFigures
Figures7c 7cand
and8c, 8c, with
with particular
particularreference
referenceto to the
the latter
latter case
case (longitudinal
(longitudinal
switch).Relevantly,
switch). Relevantly,the thelosses
losses(S21)
(S21)ofofthe thephysical
physicalsamples
samplesare aresignificantly
significantlylarger largerthanthanexpected,
expected,
being
being them worse than simulations of an extent ranging from 0.2–0.3
worse than simulations of an extent ranging from 0.2–0.3 dB up to 0.8–0.9 dB. In dB up to 0.8–0.9 dB. In particular,
the measured
particular, the S21 of the transverse
measured S21 of the switch
transverse employed
switchin the K band
employed in phase is −0.8
shifterphase
the K band dB atis19.5
shifter −0.8GHz,
dB
while
at 19.5 the
GHz,simulated
while the prediction −0.1 dB (see
simulatedisprediction Figure
is −0.1 dB7d).
(seeThe experimental
Figure S21 of the longitudinal
7d). The experimental S21 of the
switch employed
longitudinal switchinemployed
the Ka band in thephase
Ka bandshifter is −1shifter
phase dB at is 29.3
−1 GHz
dB atwhile
29.3 GHzthe simulation predicts a
while the simulation
value slightly
predicts a value larger
slightly −0.1 than
thanlarger dB (see −0.1 Figure
dB (see 8d).Figure 8d).
Despitesuch
Despite suchaarather
rathersevere
severeperformance
performancedrift drift
inin terms
terms of of larger
larger attenuation,
attenuation, its its explanation
explanation is
is straightforward and the solution quite easy as the cause is related
straightforward and the solution quite easy as the cause is related to a known technology aspect. The to a known technology aspect.
The physical
physical samplessamples
availableavailable
at the at theof
time time
thisofpaper’s
this paper’s
writingwriting
are comingare coming
from the from
firstthe firstofwafer
wafer the
of the RF-MEMS fabrication batch processed in the facility. In
RF-MEMS fabrication batch processed in the facility. In such a wafer, which was considered a such a wafer, which was considered
a preliminary
preliminary test,
test, strict
strict control
control onon thethecleaning
cleaningofofthrough-oxide
through-oxidevias viastotothethealuminum
aluminumunderpassunderpass
(aftertheir
(after theiropening)
opening)was wasnot notperformed
performed[16]. [16].This
Thishappened
happenedfor foraatechnical
technicalproblem
problemand andled ledtotolarger
larger
contact resistance between gold and aluminum, detected when
contact resistance between gold and aluminum, detected when measuring a few test structures for measuring a few test structures for
processcontrol.
process control.This Thisnon-ideality
non-idealityalso alsoaffects
affectsS-parameters
S-parametersleading leadingto tothe
thebad
badperformance
performanceshown showninin
Figure 7c,d
Figures 7c,d and
and Figure
8c,d for8c,d thefor the switches
switches in the ONin the ON state.
state.
Inthe
In thenext
nextwafers
wafersof ofthe
thesame
samefabrication
fabricationbatch, batch,still
stilltotobebefinalized,
finalized,the theissue
issueofofviasviasopening
opening
wascorrectly
was correctlyperformed
performedand andwe weexpect
expectsignificantly
significantlylower lowervalues
valuesofofcontact
contactresistance
resistanceand, and,ininturn,
turn,
reduced losses. To provide an experimental indication of the low-losses
reduced losses. To provide an experimental indication of the low-losses typically achieved by the RF- typically achieved by the
RF-MEMS
MEMS technology
technology employed
employed for for implementing
implementing thethe phase
phase shifters
shifters here
here reported,it itcan
reported, canbe behelpful
helpful
referring to [8,19]. In such works, RF-MEMS series ohmic switches with through-oxide vias properly
Sensors 2020, 20, 2612 8 of 13

Sensors 2020, 20, x FOR PEER REVIEW 9 of 14

referring to [8,19]. In such works, RF-MEMS series ohmic switches with through-oxide vias properly
opened and cleaned showed measured values of the S21 parameter better than −1 dB up to 50 GHz
opened and cleaned showed measured values of the S21 parameter better than −1 dB up to 50 GHz
when conducting (ON state) and, more importantly, better than −0.3 dB up to 30 GHz.
when conducting (ON state) and, more importantly, better than −0.3 dB up to 30 GHz.
Of course, the employment of switching units with the increased losses just reported affects
Of course, the employment of switching units with the increased losses just reported affects
negatively the RF characteristics of the phase shifters. In this respect, one should bear in mind that,
negatively the RF characteristics of the phase shifters. In this respect, one should bear in mind that,
regardless of the particular configuration of 4 bits, the RF signal has to travel through eight ON switches.
regardless of the particular configuration of 4 bits, the RF signal has to travel through eight ON
Therefore, the S11 and S21 simulated characteristics previously shown in Figure 4a,b and Figure 5a,b
switches. Therefore, the S11 and S21 simulated characteristics previously shown in Figures 4a,b and
cannot be met by the physical samples, as it is going to be discussed in the following.
5a,b cannot be met by the physical samples, as it is going to be discussed in the following.
To conclude the discussion on the testing of standalone switches, it must be reported that the DC
To conclude the discussion on the testing of standalone switches, it must be reported that the
characterization of the pull-in behavior of the physical samples showed a satisfactory agreement with
DC characterization of the pull-in behavior of the physical samples showed a satisfactory agreement
the actuation voltages previously discussed in Tables 1 and 2. This consideration applies to all the
with the actuation voltages previously discussed in Tables 1 and 2. This consideration applies to all
tested replicas of the transversal and longitudinal RF-MEMS micro-relays, which showed variations of
the tested replicas of the transversal and longitudinal RF-MEMS micro-relays, which showed
the pull-in voltage always in the range of a few percent with respect to simulations. Further details are
variations of the pull-in voltage always in the range of a few percent with respect to simulations.
not provided for the sake of brevity, leaving proper room to the upcoming RF characterization of the
Further details are not provided for the sake of brevity, leaving proper room to the upcoming RF
phase shifters.
characterization of the phase shifters.
The 4-bit phase shifters, regardless of the operating frequency band, have four actuation pads for
The 4-bit phase shifters, regardless of the operating frequency band, have four actuation pads
each reference line, four pads for the respective four different phase delays and a single ground pad
for each reference line, four pads for the respective four different phase delays and a single ground
(labeled as GND). A layout top-view of both devices is provided in Figure 9.
pad (labeled as GND). A layout top-view of both devices is provided in Figure 9.

(a) (b)
Figure 9.
Figure 9. Top-view
Top-view layout
layout topology
topology of
of the
the (a)
(a) K
K band
band (19.5
(19.5 GHz
GHz center
center frequency)
frequency) and
and of
of the
the (b)
(b) Ka
Ka
band (29.3 GHz center frequency) RF-MEMS phase shifters design concepts.
band (29.3 GHz center frequency) RF-MEMS phase shifters design concepts.

The
The reference
referencephase phasepads
padsarearelabeled
labeledas asSS00 and
and placed
placed at at the
the top
top of
of the
the scheme.
scheme. The
The pads
pads named
named
SS22.5
22.5,, S
S4545, ,SS90,90,and
andSS180 routethe
180route theDCDCsignal
signaltotothe
thebranches
branches realizing
realizing thethe corresponding
corresponding phase delays.
The
The K K and
and Ka Ka bands
bands RF-MEMS
RF-MEMS phasephase shifters
shiftershave
haveaasilicon
siliconchip
chipfootprint
footprintof of10.5 mm×× 7.5
10.5mm 7.5 mm
mm and
and
7.5
7.5 mmmm × × 8.5 mm, respectively.
respectively.
The desired
The desired bitsbits combinations
combinationsare areobtained
obtainedimposing
imposingaa45 45 V V DC
DC signal
signal toto the corresponding
corresponding pads
pads
(i.e. above
(i.e., above pull-in) pull-in) and 0 V to the ground pad. pad. ForFor the
the sake of clarity, in the the following,
following, wewe show
show
measurements for
measurements for just
just aa few
few bits
bits combinations, namely 0, 45, 67.5, 90, and 180 degrees of phase shift.
Wemeasured
We measuredthe thereflection
reflection(S11),
(S11),insertion
insertion loss
loss (S21),
(S21), andand phase
phase delaydelay
in ain2 aGHz
2 GHz frequency
frequency range
range for
for K
the the and K andKa band Ka band operating
operating frequencies,
frequencies, takingtaking into account
into account the diversity
the diversity of the
of the two two proposed
proposed design
design concepts.
concepts. The S-parameters
The S-parameters measurements
measurements are reported
are reported in the following
in the following Figures 10Figures
and 11,10for
and
the11,
K
for the
and Ka K bandandphaseKa band phaserespectively.
shifters, shifters, respectively.
Sensors 2020, 20, 2612 9 of 13

Sensors 2020, 20, x FOR PEER REVIEW 10 of 14

(a)

(b)

(c)

Figure 10. Measured

Measured characteristics
characteristics of
of the reflection (a), transmission (b), and phase delay (c) for the
19.5 GHz center frequency RF-MEMS phasephase shifter
shifter (K
(K band).
band).

The measured reflection of the observed

observed configurations between −15
configurations ranges between −15 and −25 dB at the
−25 dB
frequency (see
center frequency (seeFigure
Figure10a),
10a),while
whilethe
thesimulated
simulated values
values areare comprised
comprised between
between −25−25 dB and
dB and −43
−43(see
dB dB (see Figure
Figure 4a). 4a).
More More relevantly,
relevantly, the the measured
measured transmission
transmission −6.8−6.8
is between
is between andand−8.6−8.6 dBthe
dB at at
the center
center frequency
frequency (see(see
FigureFigure
10b),10b),
whilewhile the simulations
the simulations showshow an S21
an S21 between
between −1.4−1.4 −1.8 −1.8
and and dB
dB (see
(see Figure
Figure 4b). 4b). On the
On the otherother hand,
hand, thethe measured
measured phase
phase delay
delay is is moreaccurate,
more accurate,asasititwill
will be
be discussed
shortly later.
For the Ka band network, the measured S11 ranges between −13 and −17 dB at the center frequency
(see Figure 11a), while the simulated values are comprised between −14 dB and −35 dB (see Figure 5a).
The measured transmission is between −7 and −9.5 dB at the center frequency (see Figure 11b), while
the simulations show an S21 between −1.7 and −2.4 dB (see Figure 5b). On the other hand, as for the
previous network, the measured phase delay is more accurate.
Sensors 2020, 20, 2612 10 of 13
Sensors 2020, 20, x FOR PEER REVIEW 11 of 14

(a)

(b)

(c)

Figure 11.
Figure 11. Measured
Measured characteristics
characteristics of
of the
the reflection
reflection (a),
(a), transmission
transmission (b),
(b), and
and phase
phase delay
delay (c)
(c) for
for the
the
29.3 GHz center frequency RF-MEMS phase shifter (Ka
29.3 GHz center frequency RF-MEMS phase shifter (Ka band). band).

For
For what
the Ka concerns the characteristics
band network, the measured of reflection
S11 rangesand between
transmission
−13 ofandboth
−17the
dBphase
at theshifters
center
designs,
frequency the(see
influence
Figureof the switches
11a), while thenon-idealities
simulated values previously reportedbetween
are comprised leads to a−14
significant
dB and −35worsening
dB (see
of the performance
Figure of the network,
5a). The measured transmission which is particularly
is between −7 andsevere
−9.5 dBfor
atthe
the losses. Nonetheless,
center frequency (seethe fact
Figure
that
11b),the worsening
while of experimental
the simulations show an S21 S-parameters
between −1.7 is rather independent
and −2.4 on the
dB (see Figure 5b).particular
On the othernetwork
hand,
configuration
as for the previousand switch
network,design corroborates
the measured the delay
phase explanation
is morethat it is due to the additional contact
accurate.
resistance of theconcerns
For what switches.the
Eventually, as discussed
characteristics before,
of reflection thetransmission
and technologicalofnon-ideality was already
both the phase shifters
solved
designs, in the next wafersofstill
influence theto switches
be finalized.
non-idealities previously reported leads to a significant
In conclusion,
worsening despite the large
of the performance of thedivergence
network, which between experimental
is particularly and nominal
severe characteristics,
for the losses. Nonetheless,the
measured
the fact thatphasethedelay configuration
worsening exhibits a S-parameters
of experimental satisfactory match with independent
is rather the simulations, on as
thehighlighted
particular
network
by configuration
the phase and switch
difference plots reported design corroborates
in Figure 12, for a the
veryexplanation
narrow range thataround
it is due
thetoKthe
and additional
Ka band
contactfrequencies
center resistance of the switches.
(Figure Eventually, as discussed before, the technological non-ideality was
12a,b, respectively).
already solved in the next wafers still to be finalized.
In conclusion, despite the large divergence between experimental and nominal characteristics,
the measured phase delay configuration exhibits a satisfactory match with the simulations, as
highlighted by the phase difference plots reported in Figure 12, for a very narrow range around the
K and Ka band center frequencies (Figure 12a,b, respectively).
Sensors 2020, 20, 2612 11 of 13
Sensors 2020, 20, x FOR PEER REVIEW 12 of 14

(a)

(b)

Figure
Figure 12.
12. The
Thephase
phasedifference
difference between
between simulations
simulations and
and measurements
measurements at
at the
the center
center frequencies
frequencies for
for
the
the (a)
(a) K
K band
band and
and (b)
(b) Ka band RF-MEMS phase shifters.

4. Conclusions
4. Conclusions
RF-MEMS, i.e.
RF-MEMS, i.e., Microsystems
Microsystems (MEMS)(MEMS) based based Radio
Radio Frequency
Frequency passive
passive components,
components, exhibitexhibit
pronounced characteristics in terms of high performance, tunability/reconfigurability,
pronounced characteristics in terms of high performance, tunability/reconfigurability, and frequency and frequency
agility, making
agility, makingthem theman interesting solution
an interesting for the upcoming
solution for the fifth generation
upcoming fifthof mobile communications,
generation of mobile
i.e., 5G, with all its implications and spillovers in the Internet of Things (IoT). Active
communications, i.e., 5G, with all its implications and spillovers in the Internet of Things (IoT). Active development of 5G
development of 5G technologies and satellite communication demands for high-speed stable from
technologies and satellite communication demands for high-speed stable data transfer services data
any location. Satellite systems providing high-speed Internet connectivity
transfer services from any location. Satellite systems providing high-speed Internet connectivity utilize K and Ka frequency
bands K
utilize offering
and Kaa frequency
larger bandwidth compared
bands offering to lower
a larger frequencies
bandwidth bands.to lower frequencies bands.
compared
This work presented and discussed a couple of designof concepts
This work presented and discussed a couple of design concepts RF-MEMS-based reconfigurable
of RF-MEMS-based
phase shifters working
reconfigurable in the K working
phase shifters and Ka bands,
in thewith 19.5Ka
K and GHz and 29.3
bands, withGHz19.5center
GHzfrequency,
and 29.3 GHz respectively.
center
Both networks feature 4 reconfigurable stages and realize 16 different phase
frequency, respectively. Both networks feature 4 reconfigurable stages and realize 16 different delays by means onphase
series
ohmic by
delays RF-MEMS
means on switches. In particular,
series ohmic RF-MEMS the phase shifter
switches. design concept
In particular, the phase working in the Ka
shifter design band
concept
employs a novel in-line switch solution, which allows conjugating the
working in the Ka band employs a novel in-line switch solution, which allows conjugating the compactness of cantilevered
micro-relays with
compactness the robustness
of cantilevered of the clamped-clamped
micro-relays with the robustness configuration.
of the clamped-clamped configuration.
The characteristics and performance of the RF-MEMS
The characteristics and performance of the RF-MEMS phase phase shifters and and
shifters standalone switches
standalone were
switches
extensively discussed, capitalizing on Finite Element Method (FEM)
were extensively discussed, capitalizing on Finite Element Method (FEM) simulations and simulations and experimental
testing of the testing
experimental S-parameters.
of the S-parameters.
Due to
Due toaatechnology-related
technology-related non-ideality,
non-ideality, thethe ON-state
ON-state resistance
resistance of theof switches
the switches was larger
was larger than
than the nominal value. This led to losses around 0.8–0.9 dB at K and
the nominal value. This led to losses around 0.8–0.9 dB at K and Ka center frequencies, while Ka center frequencies, while
the
the simulations predicted 0.1–0.2 dB. As the discussed phase shifters,
simulations predicted 0.1–0.2 dB. As the discussed phase shifters, regardless of the selected regardless of the selected
configuration, always
configuration, alwaysemployemploy eight cascaded
eight cascadedON switches, the impact
ON switches, theonimpact
the overall on RFthecharacteristics
overall RF
is important. The measured networks exhibited additional losses
characteristics is important. The measured networks exhibited additional losses (S21) as (S21) as large as 5–7 dB as
large at 5–7
the
twoatcenter
dB the twofrequencies when compared
center frequencies to FEM simulations.
when compared However,
to FEM simulations. the reason
However, thefor such for
reason a scarce
such
performance is known and identified at the technology level, and the next
a scarce performance is known and identified at the technology level, and the next fabricated samples fabricated samples will
show better performance. Finally, despite the large divergence from the nominal
will show better performance. Finally, despite the large divergence from the nominal specifications, specifications, the
measured
the measured phase
phaseshift configurations
shift configurations exhibited
exhibited limited
limiteddifferences
differences whenwhen compared
compared to simulations.
to simulations.
Sensors 2020, 20, 2612 12 of 13

Author Contributions: Conceptualization, E.S.; data curation, J.I. and E.B.; funding acquisition, E.S.; investigation,
G.R., A.B., F.G., E.B., R.K., A.T., and M.D.; methodology, J.I., E.S., R.K., and A.T.; supervision, J.I.; validation, E.B.;
visualization, J.I.; writing—original draft, J.I.; writing—review and editing, J.I., G.R., A.B., F.G., E.B., E.S., and M.D.
All authors have read and agreed to the published version of the manuscript.
Funding: Concerning the activities carried out by the authors affiliated with Bazovye Tekhnologii, LLC, Moscow,
this work has been supported by the Ministry of Education and Science of the Russian Federation under the
Federal target program “Research and Development in Priority Directions of the Russian Science and Technology
Complex in 2014–2020.”
Acknowledgments: The authors would like to thank the Shared Facilities Center of the Moscow Institute of
Physics and Technology (MIPT) for making available the equipment necessary to carry out the S-parameters
measurements reported in this work.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Brown, E.R. RF-MEMS switches for reconfigurable integrated circuits. IEEE Trans. Microw. Theory Tech. 1998,
46, 1868–1880. [CrossRef]
2. Shen, S.-C.; Feng, M. Low actuation voltage RF MEMS switches with signal frequencies from 0.25 GHz
to 40 GHz. In Proceedings of the International Electron Devices Meeting 1999, Washington, DC, USA,
5–8 December 1999; pp. 689–692. [CrossRef]
3. Rebeiz, G.M. RF MEMS: Theory, Design, and Technology, 1st ed.; John Wiley & Sons: Hoboken, NJ, USA, 2003;
pp. 1–485.
4. Iannacci, J. RF-MEMS: An enabling technology for modern wireless systems bearing a market potential still
not fully displayed. Springer Microsyst. Tech. 2015, 21, 2039–2052. [CrossRef]
5. Iannacci, J. RF-MEMS Technology for High-Performance Passives: The Challenge of 5G Mobile Applications, 1st ed.;
IOP Publishing: Bristol, UK, 2017; pp. 1–166.
6. Dubuc, D.; Grenier, K.; Iannacci, J. RF-MEMS for smart communication systems and future 5G applications.
In Smart Sensors and MEMS, 2nd ed.; Nihtianov, S., Luque, A., Eds.; Woodhead Publishing: Cambridge, UK,
2018; pp. 499–539.
7. Zahr, A.H.; Zhang, L.Y.; Dorion, C.; Deveautour, A.; Beneteau, A.; Stefanini, R.; Blondy, P. RF-MEMS Switches
for Millimeter-Wave Applications. In Proceedings of the European Microwave Conference in Central Europe
(EuMCE), Prague, Czech Republic, 13–15 May 2019; pp. 336–338.
8. Iannacci, J. RF-MEMS technology as an enabler of 5G: Low-loss ohmic switch tested up to 110 GHz. Elsevier
Sens. Actuators A Phys. 2018, 279, 624–629. [CrossRef]
9. Barker, N.S.; Rebeiz, G.M. 1998 Distributed MEMS true-time delayphase shifters and wide band switches.
IEEE Trans. Microw. Theory Tech. 1998, 46, 1881–1889. [CrossRef]
10. Hacker, J.B.; Mihailovich, R.E.; Kim, M.; DeNatale, J.F. A Ka-band 3-bit RF MEMS true-time-delay network.
IEEE Trans. Microw. Theory Tech. 2003, 51, 305–308. [CrossRef]
11. Kim, M.; Hacker, J.B.; Mihailovich, R.E.; DeNatale, J.F. A DC-to-40 GHz four-bit RF MEMS true-time delay
network. IEEE Microw. Wirel. Compon. Lett. 2001, 11, 56–58. [CrossRef]
12. Lakshminarayanan, B.; Weller, T.M. Optimization and implementation of impedance-matched true-time-delay
phase shifters on quartz substrate. IEEE Trans. Microw. Theory Tech. 2007, 55, 335–342. [CrossRef]
13. Koul, S.K.; Dey, S. RF MEMS True-Time-Delay Phase Shifter. In Micro and Smart Devices and Systems,
1st ed.; Vinoy, K.J., Ananthasuresh, G.K., Pratap, R., Krupanidhi, S.B., Eds.; Springer: Berlin, Germany, 2014;
pp. 467–485. [CrossRef]
14. Dey, S.; Koul, S.K. Design and development of a CPW-based 5-bit switched-line phase shifter using inline
metal contact MEMS series switches for 17.25 GHz transmit/receive module application. IOP J. Micromech.
Microeng. (JMM) 2013, 24, 1–25. [CrossRef]
15. Nordquist, C.D.; Dyck, C.W.; Kraus, G.M.; Sullivan, C.T.; Austin IV, F.; Finnegan, P.S.; Balance, M.H. Ku-Band
Six-Bit RF MEMS Time Delay Network. In Proceedings of the IEEE Compound Semiconductor Integrated
Circuits Symposium, Monterey, CA, USA, 12–15 October 2008; pp. 1–4. [CrossRef]
16. Giacomozzi, F.; Mulloni, V.; Colpo, S.; Iannacci, J.; Margesin, B.; Faes, A. A Flexible Fabrication Process for
RF MEMS Devices. Rom. J. Inf. Sci. Technol. (ROMJIST) 2011, 14, 259–268.
 Sensors 2020, 20, 2612 13 of 13

17. Savin, E.; Bochkova, E.; Kirtaev, R.; Iannacci, J. RF MEMS packaged switch evaluation tests up to 25 GHz.
Springer Microsyst. Technol. 2019, 25, 2191–2195. [CrossRef]
18. Liu, Y.; Liu, J.; Yu, B.; Liu, X. A compact single-cantilever Multicontact RF-Mems switch with enhanced
reliability. IEEE Microw. Wirel. Compon. Lett. 2018, 28, 191–193. [CrossRef]
19. Iannacci, J.; Huhn, M.; Tschoban, C.; Pötter, H. RF-MEMS Technology for 5G: Series and Shunt Attenuator
Modules Demonstrated up to 110 GHz. IEEE Electron. Dev. Lett. 2016, 37, 1336–1339. [CrossRef]
20. Savin, E.A.; Chadin, K.A.; Kirtaev, R.V. Design and manufacturing of X-band RF MEMS switches. Springer
Microsyst. Technol. 2018, 24, 2783–2788. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

View publication stats