This article has been accepted for inclusion in a future issue of this journal.

Content is final as presented, with the exception of pagination.

IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES 1

Electromagnetic Energy Harvesting

Using Full-Wave Rectification
Faruk Erkmen, Member, IEEE, Thamer S. Almoneef, Graduate Student Member, IEEE, and
Omar M. Ramahi, Fellow, IEEE
(Invited Paper)

Abstract— This paper presents a new approach to realizing despite its common use in power electronics [18]–[21]. Fur-
full-wave rectification for electromagnetic energy harvesting. thermore, the reported applications have various limitations.
Instead of using one antenna to feed one rectifier circuit, References [18] and [21] focused on the matching and con-
we propose utilizing two antennas connected to one rectifier to
form a full-wave rectenna. This balanced configuration allows version efficiency of the full-wave rectifier, but did not report
the received power to be rectified and transferred to a dc the overall rectenna performance. Hosain et al. [19] reported
load between two antennas, making it very easy to channel the a detailed full-wave rectifier design integrated with a PIFA
harvested power in rectenna arrays. The proposed concept is antenna. However, this study was for a specific application
demonstrated in the microwave regime using an array of full- and did not comment on scalability to an array setting where
wave rectennas optimized for 2.45 GHz. The full-wave rectenna
array is compared with an array of half-wave rectennas that channeling the rectified dc current can be challenging. A full-
occupy the same footprint and are optimized to maximize wave bridge rectifier attached to a single folded dipole antenna
power absorption at the same frequency. Measurements showed was reported recently where only the rectification efficiency of
that the proposed full-wave rectifier performed better than the the rectifier alone was studied [20].
half-wave rectifier and achieved 74% radiation-to-dc conversion Designing a highly efficient rectenna is very important
efficiency. Here, without loss of generality, the proposed concept
is demonstrated using T-matched dipole antennas at a specific for EM energy harvesting. However, in order to achieve
frequency, but the concept can be applied to other antenna types practical adoption of the technology, it is equally important
and other frequencies. to have array implementations with efficient dc channel-
Index Terms— Array, channeling, dipole antenna, energy har- ing mechanisms. With the available ambient power levels,
vesting, full-wave rectification, rectenna, rectifier, Schottky diode. individual rectennas are limited in terms of the amount of
power they can collect. (Several recent studies investigated
I. I NTRODUCTION the available RF power densities in various parts of the
world [22]–[24].) Therefore, converting ambient RF energy
T HE idea of collecting electromagnetic (EM) energy and
converting it into various forms of electrical power dates
back to Tesla’s experiments in the early 20th century [1],
into practical dc power can be possible only using large
rectenna arrays. Various different antenna types have been
followed by researchers in Japan and USA in subsequent studied such as dipoles, bow-ties, patches, rings, SRR loaded
decades [2], [3]. The first rectenna for efficient reception and antennas, and spiral antennas [15], [16], [25]–[30]. In terms
rectification of microwave power was developed in the early of the rectification mechanism, most of the reported rectenna
1960s by Raytheon Corporation [4]. The introduction of semi- arrays use half-wave rectification. Some designs utilized volt-
conductor diodes and the invention of Schottky diodes were age doubler circuits and stacked differential structures to
significant developments toward the realization of practical enhance the dc output voltage [26], [31]. Only a few examples
rectenna designs [4]–[9]. implemented full-wave rectification in rectenna arrays with
The basic operation of EM energy harvesting relies on very limited details provided on the rectenna efficiency as a
two key elements: the receiving antenna, or EM collec- system [9].
tor, and the rectifying circuit. The rectifier can be con- In this paper, we focus on developing a rectenna array
structed using a single-series mounted diode, a single-shunt system that utilizes full-wave rectification to maximize ac to
mounted diode, or a diode bridge [10]. Among these designs, dc power conversion efficiency and thus the efficiency of the
the single diode configuration serving as half-wave rectifier entire energy harvesting system. We introduce a new concept
is the simplest and most common configuration [11]–[17]. where a full-wave rectification circuit is differentially fed
Diode bridge rectifier or full-wave rectifier, on the other by two separate antennas. This structure allows the received
hand, has very few examples in RF or microwaves literature power to be rectified and transferred to a dc load between
two antennas. We first validated the concept by designing and
Manuscript received October 12, 2016; revised January 24, 2017; accepted fabricating a rectenna array that consists of T-matched dipole
February 2, 2017.
The authors are with the Department of Electrical and Computer Engi- antennas and the proposed full-wave rectifiers. Experimental
neering, University of Waterloo, Waterloo, ON N2L 3G1, Canada (e-mail: results are provided showing the radiation-to-dc conversion
[email protected]; [email protected]). efficiency of the fabricated rectenna. We then compared the
Color versions of one or more of the figures in this paper are available
online at http://ieeexplore.ieee.org. proposed full-wave rectenna to that of half-wave rectennas
Digital Object Identifier 10.1109/TMTT.2017.2673821 operating at the same frequency and occupying the same
0018-9480 © 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

2 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES

Fig. 1. Schematic illustrating the concept of full-wave rectification using two

identical dipole antennas. The distance between the two antennas is electrically
small and not shown to scale.

physical footprint. Finally, the dc channeling feature of the

proposed topology is demonstrated in an array of full-wave
rectennas where the harvested dc power is channeled into a Fig. 2. Equivalent schematic of the proposed full-wave rectenna showing
single resistive load for the entire panel. the direction of dc current. (a) During the positive half cycle. (b) During the
negative half cycle.
II. T HEORY OF O PERATION
There are different implementations of a rectenna that as dependent ac sources, and the proposed full-wave rectenna
include single or multiple diodes together with matching system becomes equivalent to a bridge rectifier connected to
circuits and filters [10]. In its most simple form, a rectifier two dependent ac sources on both sides as shown in Fig. 2.
needs a diode, a load resistor, and a storage element to During the positive half cycle, diodes D2 and D3 would turn
smoothen the output voltage or current. The storage element ON with inductor L2 providing the dc current return path.
can be either a capacitor in shunt or an inductor in series Similarly, during the negative half cycle, diodes D1 and D4
with the load. A half-wave rectifier is a simple and low-cost would turn ON with inductor L1 providing the dc current
construction, which requires very few components. Its main return path. The inductors L1 and L2 act as dc short and
limitation is the fact that it can utilize only one half cycle RF open between the two antennas thus completing the closed
of the ac signal. On the other hand, full-wave rectifiers can circuit for dc operation while isolating the antennas from each
take advantage of both half cycles of the ac signal, which can other for RF operation. L1 and L2 also serve as smoothing
lead to improved efficiencies. However, full-wave rectification filters to minimize the ripples of the rectified dc current.
requires two diodes to be turned ON compared with only one The working principle of the proposed configuration is very
in half-wave rectification. The full-wave rectenna concept we similar to the traditional full-wave bridge rectifier. The main
propose here is illustrated using the schematic shown in Fig. 1 difference is that the proposed rectenna utilizes two separate
where a full-wave bridge is implemented between two dipole antennas differentially feeding the full-wave rectifier, which
antennas. Diode pairs are connected to the antenna terminals, allows the received RF power by the antennas to be rectified
and a resistive load is connected between them. Here, dipole and transferred to a dc load between the antennas. This feature
antennas are chosen for simplicity, but the concept can be can be utilized for alternative dc channeling schemes in planar
applied to other types of antennas. The ends of the dipoles and stacked rectenna arrays. If the two antennas were replaced
are tied together through ac isolating inductors to provide a with a single antenna, V1 and V2 in Fig. 2 would merge
dc path as will be shown below. to become one dependent ac voltage source and the whole
When EM radiation is received by the antennas, ac voltage configuration would become a traditional full-wave rectenna.
is generated across their terminals with a sinusoidal waveform. For ideal full-wave rectification, V 1 and V 2, the dependent
The generated voltage depends on the antenna parameters as ac voltages replacing the antennas in the equivalent model,
well as the properties of the incident wave such as frequency, should be identical to each other. However, this condition is
polarization, and angle of incidence. Antenna parameters such possible only if identical antennas are exposed to the same
as radiation efficiency and directivity determine how well the incident wave properties. If the incoming wave has an off-
antenna captures the EM radiation and generates the ac voltage normal angle with respect to the plane of the two antennas,
at its terminals. The impedance matching between the antenna there would be a phase difference between the generated
and the rectifier determines how much of the generated ac voltage sources due to the distance between the two antennas
power can be transferred to the rectifier and how much will even if the antennas were identical. In such cases, the phase
be reflected. The rectifier’s RF-to-dc conversion efficiency difference can be minimized by maintaining an electrically
determines how much of the input power can be converted to small distance between the antennas. Moreover, if the antennas
dc power at its load. The antennas in Fig. 1 can be modeled were not identical, the generated voltages could be different
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

ERKMEN et al.: EM ENERGY HARVESTING USING FULL-WAVE RECTIFICATION 3

in both amplitude and phase. At any given time, the generated

voltages can turn ON only two of the diodes, while the
other two are in reverse bias. The full-wave rectification
mechanism is possible as long as there is enough potential
difference to turn ON the two diodes during both half cycles.
Overall, the power conversion efficiency can be expected to
degrade for off-normal incidence or for nonidentical antenna Fig. 3. Optimized antenna dimensions of the T-matched dipole for full-wave
configurations. rectification at 2.45 GHz (all in mm): a = 44, b = 3, c = 4, d = 1, e = 3,
f = 21, and g = 1.
III. F ULL -WAVE R ECTENNA D ESIGN
The rectifier and antenna are the two primary functional
blocks of a rectenna system. These two blocks are gener-
ally designed separately and integrated through a matching
network. In this paper, we conceived the rectenna as an
integral system, and thus the rectification circuitry was not
only matched to the antenna (as if they were two isolated
stages) to ensure optimal power transfer but the dc channeling
was also integrated as part of the entire system rather than only
a receiving stage. Schottky diodes from Avago HSMS-286x
family were used because of their low turn-on voltage and high
detection sensitivity [32]. Dual diodes in SOT-23 packages of
HSMS-2863 and HSMS-2864 were chosen for ease of fabri-
cation. For the inductor selection, the inductance and the self-
resonant frequency (SRF) were the two main selection criteria.
High inductance value is desired for better smoothing function, Fig. 4. Fabricated panel of full-wave rectenna array (FWR-Panel). The
and better RF isolation between the antennas. High SRF is rectifier circuit placed at the bottom layer is shown in the inset.
also desired to avoid parasitic effects of the inductor package.
However, the inductance is inversely proportional with the Rectenna designs often require a matching circuit because
SRF and these two criteria narrow down the inductor selection. in most cases the antennas and rectifiers are optimized sep-
Murata’s 0603 size LQW18AN47NG00D was chosen having arately and most antennas cannot be matched to the rectifier
47 nH inductance, 0.29  dc resistance, and 2.6 GHz SRF [33] impedance easily [19]. In this paper, we adopted the T-matched
(47 nH was the highest inductance available in the high dipole antenna since it provides excellent control over its input
quality factor LQW series that has an SRF higher than the impedance by optimization of its dimensions [34]–[37].
intended operating frequency of 2.45 GHz). As explained The full-wave rectenna array was designed using CST
in Fig. 2, the proposed full-wave rectenna concept requires Microwave Studio [38]. The dipole antenna dimensions and
the two antennas to be connected to each other through the spacing between array elements were optimized to max-
inductors at each end. Using a single inductor would require imize the received power from an incident plane wave with
physical traces to be connected to the antennas affecting their linear polarization and with a normal angle of incidence to the
radiated performance. To avoid this effect, the connection was array panel. It should be noted that the implementation of the
achieved using two inductors at the antenna ends and a PCB full-wave rectifier did not affect the spacing between antennas,
trace between them. Inductors have high impedance at the and did not increase the optimum size of the rectenna struc-
frequency of operation and act as an open circuit. Hence, ture. The distance between array elements was mainly deter-
both antennas were essentially disconnected from the PCB mined by the mutual coupling effects between antennas. The
trace and from each other during the RF operation, making it T-sections of the dipoles were optimized through parametric
possible to design and optimize their antenna characteristics simulations in order to achieve the desired impedance char-
individually. acteristics that conjugate match the measured rectifier input
The full-wave rectifier was built on Rogers RO4003 board, impedance. Fig. 3 shows the optimized antenna dimensions.
and its input impedance was measured using Agilent E-5071C This approach minimized the mismatch losses while eliminat-
vector network analyzer (VNA). The impedance behavior is ing the need for a matching network.
dependent on the rectifier load and input power. Therefore, The rectenna array was fabricated on a Rogers RO4003 sub-
the rectifier impedance was characterized over a range of strate with 1.52 mm thickness and 35 μm copper cladding
input power and load resistance by adjusting the VNA port as shown in Fig. 4. The full-wave rectifier was implemented
power settings and the load resistance. The impedance curve between each antenna pair. In the array configuration, the ver-
at 5 dBm input power and 1 k load resistance was determined tical and horizontal spacings between the antenna elements
to be most suitable to be matched by the antenna impedance, (see Fig. 4) were 20 and 5 mm, respectively. The antenna
and was selected to be used for antenna optimization. This terminals were connected to the rectifier leads through plated
selections was revisited later where the radiated measurement vias. The central antenna pair and the load in between were
results are discussed. used for the measurements.
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

4 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES

Fig. 5. Differential input impedance measurement setup. (a) Custom

calibration kit. (b) W.FL to SMA adaptor. (c) Calibrated reference plane of
the logical differential port. (d) Input impedance measurement of the T-match
dipole antenna.

In order to verify the conjugate match condition, the input Fig. 6. Measured impedance plots of the full-wave rectifier and the antenna.
impedance of the central antenna was measured using the dif- Frequency range is 1.5–4 GHz. Good conjugate match is obtained without a
ferential probe method with a custom-built calibration kit. For matching network.
balanced antennas, such as dipoles, impedance measurements
require special techniques where equal and opposite feed TABLE I
currents are needed from the test equipment [39], [40]. There I MPEDANCE M ATCHING OF THE F ULL -WAVE R ECTENNA
are several techniques that can be used to avoid this issue,
but the differential probe method was found most suitable
for this paper due to its practicality [39]–[42]. This technique
places the calibration reference plane directly at the measured
port, which eliminates the need for postprocessing such as
de-embedding or conversion calculations [43], [44]. The
custom-built calibration kit was fabricated on Rogers
RO4003 substrate as shown in Fig. 5(a). The substrate thick-
ness was 1.524 mm. The calibration kit included open, short,
50  load, and through. Traces in the calibration kit were
kept very short and designed as short coplanar waveguides
without ground plane. Two 100  resistors were used to
achieve a differential 50  matched load, and negligible where Pdc,col. is the dc power collected at the load and PRad.,av.
impedance jumpers were used to achieve short termination. is the total incident radiated power available at the physical
Keysight E5071C was used as the VNA. Two ports of the VNA area of the antenna [19], [28], [46]. ηRad-dc can be written as
were combined to compose a logical differential port. Hirose
W.FL series miniature coaxial cables were used as extension ηRad-dc = ηRad-ac ηac-dc (2)
to regular semirigid coaxial cables. W.FL series offers very
low profile mated height, small footprint, and extreme light where ηRad-ac is the radiation-to-ac efficiency representing the
weight, which makes it an ideal choice for delicate antenna effectiveness of the antennas to capture EM radiation and
measurements [45]. Fig. 5(b) and (c) shows the final differen- ηac-dc represents the conversion efficiency of the rectification
tial probe used for the impedance measurements. The input circuitry.
impedance of the central antenna was measured using the The efficiency of the full-wave rectenna was evaluated in an
differential port setup as illustrated in Fig. 5(d). Measured array implementation rather than a single standalone structure
antenna impedance together with the full-wave rectifier input because antennas perform differently when positioned in close
impedance is presented in Fig. 6. Measured impedances and proximity to adjacent antennas. Mutual coupling between them
the return loss between the antenna and the rectifier are plays a significant role in determining the individual antenna
tabulated in Table I. As can be seen, very good matching was parameters [46]–[48]. This effect was taken into account using
obtained around 2.45 GHz. an array configuration and evaluating the dc power converted
through a single central element. The obtained results represent
IV. R ADIATED M EASUREMENTS the per-unit-cell performance of the array. A good estimate of
The rectenna performance was evaluated using the the total dc power collected by the array is the product of the
radiation-to-dc power conversion efficiency ηRad-dc defined as dc power collected by the central element and the number
of cells in the array, assuming an efficient dc channeling
Pdc,col.
ηRad-dc = (1) mechanism. Of course, the larger the array, the more accurate
PRad.,av. this estimate becomes.
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

ERKMEN et al.: EM ENERGY HARVESTING USING FULL-WAVE RECTIFICATION 5

TABLE II
L INK B UDGET PARAMETERS AT 2.45 GHz

Fig. 7. Schematic showing the setup for measurement of the radiation-to-dc

conversion efficiency. antenna to the rectifier. Therefore, to obtain the efficiency
in (1), the available incident power PRad.,av. was determined
by applying the link budget calculations [19]
PRad.,av. = SRad.,av. A p (3)
1
SRad.,av. = EIRP (4)
4πd 2
EIRP = Ps − L c + G a + G h (5)
where EIRP is the effective isotropic radiated power of
the transmitter and L fs = 1/4πd 2 is the distance depen-
dency, or the free space path loss of the EM energy expressed
with inverse square law. The source power Ps was known
from the signal generator. The losses of the coaxial cables
were measured using the VNA. The amplifier gain G a was
measured with a power meter and the horn antenna gain G h
was obtained from its datasheet. The source power and the
frequency were varied during the measurements to find the
peak efficiency. The transmitter horn antenna gain was also
frequency dependent. Table II provides the definitions of the
Fig. 8. Measurement setup inside an anechoic chamber. link budget parameters and their values at 2.45 GHz.
Fig. 9(a) shows the rectenna efficiency versus load resis-
tance. The peak occurred at 1 k load that is consistent
The radiated measurement setup is illustrated in Fig. 7. with the rectifier design provided above. A full-wave rectifier
A photograph of the measurements setup inside an anechoic with load resistance 1 k was taken as a reference while the
chamber is shown in Fig. 8. A signal generator was used antenna was optimized to match this impedance.
as the source. Output of the signal generator was amplified Fig. 9(b) shows the rectenna efficiency versus incident
and fed into a standard gain horn antenna in such a way power. The peak occurred when the incident power density
that the electric field of the incident wave was parallel to the SRad.,av. was 0.23 mW/cm2 . By multiplying this value with the
dipole arms. The full-wave rectenna panel was placed 1 m rectenna physical area A p used for the efficiency calculation,
away from the horn antenna. Considering the design frequency we can obtain the available power for the rectenna to be
of 2.45 GHz and the largest antenna dimension of 44 mm, PRad.,av = 4.5 mW (6.5 dBm). This was also consistent with
the Fraunhofer distance was calculated as 0.032 m and the the rectifier design because the reference impedance used for
measurement distance of 1 m satisfied the far-field conditions. antenna optimization was obtained by measuring the full-wave
Measurements were carried out with varying load resistance, rectifier with 5 dBm input power.
incident power, and frequency. The rectified dc voltage across Fig. 9(c) shows the rectenna efficiency versus frequency
the load was recorded using a digital multimeter and the while keeping the load resistance and the incident power
collected dc power Pdc,col. was calculated for all test cases. at their optimum values. A peak radiation-to-dc conversion
Physically measuring the received power at the antenna efficiency of 74% was recorded at 2.55 GHz. Considering
terminals was not feasible because this would require the that the rectenna array was originally optimized for 2.45 GHz
antenna to be matched to the test equipment’s 50  port and with good impedance matching with the rectifier, the frequency
that would contradict one of the key features of this paper: dependency of the rectenna efficiency was found to be in good
to eliminate the matching network by directly matching the agreement with design objectives. The T-matched dipoles were
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

6 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES

circuit only where the RF power is fed into the rectifier ter-
minal (generally via an SMA connection) and the rectified dc
power is measured at the output load resistor. It indicates what
percentage of the rectifier’s input RF power can be converted
into dc output power at its load. In this paper, we designed
the whole rectenna system with a novel rectifying topology
and measured the radiation-to-dc conversion efficiency that
includes receiving the incident EM radiation, transferring the
captured power to the rectifier, and converting it into dc power
at the load. Radiation-to-dc conversion efficiency is a metric
for the rectenna as a system, indicating what percentage of
the available EM power (radiated) on a given surface area
can be collected and converted into dc power at the system
load. Unlike the RF-to-dc conversion efficiency, radiation-to-
dc conversion efficiency takes the antenna performance and
the physical area into account when determining the overall
performance.
Efficiencies involving the physical areas of rectennas have
been previously reported. For instance, 53% efficiency with
an array of split ring resonators was reported in [29], while
in [51], a near-unity efficiency was reported employing a
ground plane backed metasurface. However, both of these
studies measured the collected power in terms of time average
power (ac) and they did not include rectifiers. In [52], a large
rectenna array was presented, but its overall efficiency was
measured using the transmitted power instead of available
power on the array surface. In [11], [31], and [53], similar
radiation-to-dc conversion efficiency definitions were used
reporting 38%, 38%, and 44.1% overall efficiencies for power
densities as low as 33, 0.030, and 0.041 W/m2 , respectively.

V. C OMPARISON W ITH H ALF -WAVE R ECTENNAS

Although the experimental results showed that the new
rectenna topology is highly efficient, it is imperative to com-
pare it with the half-wave rectenna array design to demonstrate
the value added by the additional complexity involving full-
wave rectification. Therefore, for meaningful comparison, two
panels of rectenna arrays were designed using a simple half-
wave rectifier circuit that includes a diode, a load resistor,
and an inductor, all in series. The same components from
the full-wave rectification design were used. The first panel
was designed to reflect the traditional λ/2 spacing between the
antenna elements (HWR-Panel1), and the second was identical
to the full-wave rectenna array in terms of distances between
the antenna elements (HWR-Panel2). Both rectenna arrays
used T-matched dipole antennas. Only the T-sections were
Fig. 9. Measured radiation-to-dc conversion efficiency ηRad-dc of the full- optimized for each panel keeping the main antenna length and
wave rectenna system. (a) As a function of the load resistance. (b) As a width constant for a fair comparison.
function of the incident power. (c) As a function of the frequency. Following the design steps described above, the half-wave
rectifier was built on Rogers RO4003 board and its input
impedance was measured. T-matched dipole antennas were
found to be highly sensitive to fabrication tolerances and even then optimized to match the rectifier impedance. Both panels
small deviations in the trace widths could be responsible for were fabricated on Rogers RO4003 boards. Fig. 10(a) shows
the frequency shift observed in Fig. 9(c). the HWR-Panel1, and Fig. 10(b) shows the HWR-Panel2.
We emphasize that the measured 74% efficiency should not Excellent matching was obtained with return loss greater than
be compared with the widely reported RF-to-dc conversion 10 dB for both panels over the frequency range of interest.
efficiencies in [20], [26], [27], [30], [49], and [50]. The The half-wave rectenna panels were tested following the
RF-to-dc conversion efficiency is a measure of the rectifier steps described above. The optimum load resistance for each
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

ERKMEN et al.: EM ENERGY HARVESTING USING FULL-WAVE RECTIFICATION 7

Fig. 11. Measured radiation-to-dc conversion efficiency ηRad-dc of the

fabricated rectennas. (a) As a function of the incident power. (b) As a function
of the frequency.
Fig. 10. Fabricated half-wave rectenna panels with the designated unit
cell area shown using dashed lines. (a) HWR-Panel1: 3 × 3 layout with
λ/2 spacing. (b) HWR-Panel2: 3×6 layout similar to full-wave rectenna panel
in FWR-Panel. The HWR-Panel2 array used half-wave rectifier, while the
FWR-Panel array used the proposed full-wave rectifier. The
measurements shown in Fig. 11 verify that the full-wave recti-
panel was determined where peak efficiency was recorded. fication indeed performs better than the half-wave rectification
Using these values, both panels were tested with respect to with an increase in the peak ηRad-dc of more than 10%.
incident power and frequency. Voltage readings were recorded As a limitation, the rectennas in this paper generally
from the central rectennas and the efficiencies were calculated required higher power densities than the reported available
using the physical areas shown with dashed lines in Fig. 10, ambient levels to reach their peak efficiencies [23]. However,
which represent unit cells. there are still many applications where the power densities
Measured efficiencies of the half-wave rectennas compared are well within the measured levels in this paper. Outdoor
with the previously measured full-wave rectenna are provided locations closer to mobile network towers or indoor locations
in Fig. 11. Several observations can be made based on these closer to Wi-Fi routers can be given as examples.
results. HWR-Panel1 and HWR-Panel2 have identical half-
wave rectifiers and essentially the same antenna elements, but
VI. C HANNELING THE DC P OWER IN
HWR-Panel2 demonstrates much higher efficiency. Since the
F ULL -WAVE R ECTENNA A RRAY
main difference is the panel layout, this efficiency improve-
ment can be attributed to the denser array configuration All the measurements so far were performed at the central
in HWR-Panel2. This is expected because the rectenna in element of the arrays to obtain a unit cell characteristics. In this
HWR-Panel1 has twice the physical footprint, but harvests section, we designed and fabricated a full-wave rectenna array
only 22% more power. This comparison experimentally veri- to measure its overall panel performance and to demonstrate
fies that utilizing denser arrays can increase the rectenna effi- the ease of dc channeling with the proposed rectenna topology.
ciency significantly. Comparing HWR-Panel2 and FWR-Panel, Similar to the FWR-Panel and HWR-Panel2, 18 T-matched
we note that the array layouts were identical and both dipole elements were used in a 3 × 6 configuration. Instead
structures used essentially the same antenna elements. of using a load resistor for every rectifier, the dc current
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

8 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES

Fig. 12 demonstrates the actual array implementation of the

proposed rectenna. The antenna dimensions were optimized
for the unit cell, which is fairly accurate for the central
antennas but not for the ones on the periphery of the panel.
Thus, for the peripheral rectennas, larger mismatch losses
would be expected between the antennas and the rectifiers.
This is the main reason for the efficiency of the panel to be
lower than the previously measured unit cell rectenna. The
rectified dc power from each rectenna element was combined
Fig. 12. Full-wave rectenna array where all the harvested dc power is in parallel, which resulted in a relatively low load resistance
collected using a single load. (a) Top surface showing the antennas (directly
facing the incoming radiation). (b) Bottom surface showing the rectifiers and of 66 . For practical applications, the load impedance of the
the dc channeling traces. rectenna array can be treated as a design parameter and can be
optimized to match a desired value. This can be achieved by
either a series or parallel combination of the dc current, or by
optimizing the individual rectenna parameters.
VII. C ONCLUSION
A novel EM radiation-to-dc power conversion system is
proposed realizing full-wave rectification. This balanced con-
figuration allows the received power to be rectified and trans-
ferred to a dc load between two antennas. The concept is
demonstrated at 2.45 GHz using T-matched dipole antennas.
The rectifier was built first, and the antenna dimensions
were optimized to match the measured rectifier impedance,
eliminating the need for a matching network.
Radiated measurements showed that the proposed full-wave
rectenna achieved 74% radiation-to-dc conversion efficiency,
which was higher than similarly constructed half-wave recten-
nas. In addition to its high efficiency, the proposed topology
allowed the rectified dc current to flow from one antenna to
another through a resistive load. This feature was utilized in
a rectenna array to demonstrate the ease of dc channeling to
a single load.
R EFERENCES
[1] M. Cheney, Tesla: Man Out of Time. New York, NY, USA: Simon and
Schuster, 2011.
[2] H. Yagi and S. Uda, “On the feasibility of power transmission by
electric waves,” in Proc. 3rd Pan-Pacific Sci. Congr., Tokyo, Japan,
Oct./Nov. 1926, pp. 1305–1313.
[3] “Electric light without current,” Literary Dig., vol. 112, no. 3, pp. 30–31,
Jan. 1932.
[4] W. C. Brown, “The history of power transmission by radio waves,”
IEEE Trans. Microw. Theory Techn., vol. MTT-32, no. 9, pp. 1230–
1242, Sep. 1984.
[5] W. C. Brown, “The amplitron, a super power microwave generator,”
Electron. Prog., vol. 5, no. 1, pp. 1–5, Jul. 1960.
[6] W. C. Brown, “Electronic and mechanical improvement of the receiving
Fig. 13. Measured radiation-to-dc conversion efficiency (ηRad-dc ) of the terminal of a free-space microwave power transmission system,” NTRS,
full-wave rectenna array with dc power channeled to a single load. (a) As a NASA, Chicago, IL, USA, Tech. Rep. CR-135194, PT-4964, Aug. 1977.
function of the incident power. (b) As a function of the frequency. [7] W. C. Brown and J. F. Triner, “Experimental thin-film, etched-circuit
rectenna,” in IEEE MTT-S Int. Microw. Symp. Dig., Dallas, TX, USA,
Jun. 1982, pp. 185–187.
was channeled to flow through a single load for the entire [8] P. Koert, J. Cha, and M. Machina, “35 and 94 GHz rectifying antenna
panel. A two-layer RO4003 substrate was used for fabrication systems,” in Proc. SPS Power Space, Gif-sur-Yvette, France, Aug. 1991,
where the rectifier circuits and the dc channeling traces were pp. 541–547.
[9] S. S. Bharj, R. Camisa, S. Grober, F. Wozniak, and E. Pendleton, “High
routed at the bottom layer, while the antennas were at the top efficiency C-band 1000 element rectenna array for microwave powered
layer directly facing the incoming radiation. Fig. 12 shows the applications,” in IEEE MTT-S Int. Microw. Symp. Dig., vol. 1. Jun. 1992,
fabricated full-wave rectenna array. pp. 301–303.
[10] V. Marian, B. Allard, C. Vollaire, and J. Verdier, “Strategy for microwave
As shown in Fig. 13, radiated measurements showed that energy harvesting from ambient field or a feeding source,” IEEE Trans.
peak ηRad-dc of 52% was obtained at 66  load resistance and Power Electron., vol. 27, no. 11, pp. 4481–4491, Nov. 2012.
0.14 mW/cm2 incident power at the panel surface. Total panel [11] H. Takhedmit, L. Cirio, F. Costa, and O. Picon, “Transparent rectenna
and rectenna array for RF energy harvesting at 2.45 GHz,” in Proc.
area used for efficiency calculations was 170.4 cm2 , which 8th Eur. Conf. Antennas Propag. (EuCAP), Hague, The Netherlands,
encompasses all of the 18 rectennas. Apr. 2014, pp. 2970–2972.
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

ERKMEN et al.: EM ENERGY HARVESTING USING FULL-WAVE RECTIFICATION 9

[12] T. Oka, T. Ogata, K. Saito, and S. Tanaka, “Triple-band single- [36] K. R. Demarest and D. D. Deavours, “Limitations of the Uda model
diode microwave rectifier using CRLH transmission line,” in Proc. for T-match antennas,” Prog. Electromagn. Res., vol. 113, pp. 1–15,
Asia–Pacific Microw. Conf., Sendai, Japan, Nov. 2014, pp. 1013–1015. Jan. 2011.
[13] S. D. Assimonis and A. Bletsas, “Energy harvesting with a low-cost [37] T. S. Northup, “Designing a planar t-match antenna to have a 2nd-order
and high efficiency rectenna for low-power input,” in Proc. IEEE Chebyshev band-pass filter frequency response for the purpose of UHF
Radio Wireless Symp. (RWS), Newport Beach, CA, USA, Jan. 2014, RFID,” Ph.D. dissertation, Dept. Elect. Eng. Comput. Sci., Univ. Kansas,
pp. 229–231. Lawrence, KS, USA, 2008.
[14] U. Olgun, C. C. Chen, and J. L. Volakis, “Wireless power harvest- [38] CST. Studio Suite CST Computer Simulation Technology AG, accessed
ing with planar rectennas for 2.45 GHz rfids,” in Proc. (URSI) Int. on Jan. 1, 2015. [Online]. Available: http://www.cst.com
Symp. Electromagn. Theory (EMTS), Berlin, Germany, Aug. 2010, [39] T. Koskinen, H. Rajagopalan, and Y. Rahmat-Samii, “Impedance mea-
pp. 329–331. surements of various types of balanced antennas with the differen-
[15] J. Heikkinen and M. Kivikoski, “Low-profile circularly polarized rectify- tial probe method,” in Proc. IEEE Int. Workshop Antenna Technol.,
ing antenna for wireless power transmission at 5.8 GHz,” IEEE Microw. Santa Monica, CA, USA, Mar. 2009, pp. 1–4.
Wireless Compon. Lett., vol. 14, no. 4, pp. 162–164, Apr. 2004. [40] R. Meys and F. Janssens, “Measuring the impedance of balanced
[16] J. Heikkinen and M. Kivikoski, “A novel dual-frequency circularly antennas by an S-parameter method,” IEEE Antennas Propag. Mag.,
polarized rectenna,” IEEE Antennas Wireless Propag. Lett., vol. 2, no. 1, vol. 40, no. 6, pp. 62–65, Dec. 1998.
pp. 330–333, Jan. 2003. [41] C. A. Balanis, Antenna Theory: Analysis and Design. Hoboken, NJ,
[17] J.-S. Sun, R.-H. Chen, S.-K. Liu, and C.-F. Yang, “Wireless power USA: Wiley, 2012.
transmission with circularly polarized rectenna,” Microw. J., vol. 2, [42] K. D. Palmer and M. W. van Rooyen, “Simple broadband measurements
pp. 1–15, Jan. 2011. of balanced loads using a network analyzer,” IEEE Trans. Instrum.
[18] B. Merabet et al., “A 2.45-GHz localized elements rectenna,” in Proc. Meas., vol. 55, no. 1, pp. 266–272, Feb. 2006.
3rd IEEE Int. Symp. Microw., Antenna, Propag. EMC Technol. Wireless [43] J. Grosinger, C. Mecklenbrauker, and A. L. Scholtz, “UHF RFID
Commun., Beijing, China, Oct. 2009, pp. 419–422. transponder chip and antenna impedance measurements,” in Proc. 3rd
[19] M. K. Hosain, A. Z. Kouzani, M. F. Samad, and S. J. Tye, “A miniature Int. EURASIP Workshop RFID Technol, Cartagena, Spain, Sep. 2010,
energy harvesting rectenna for operating a head-mountable deep brain pp. 43–46.
stimulation device,” IEEE Access, vol. 3, pp. 223–234, Apr. 2015. [44] J. Grosinger, L. W. Mayer, C. F. Mecklenbrauker, and A. L. Scholtz,
[20] M. Ito et al., “High efficient bridge rectifiers in 100MHz and 2.4 GHz “Input impedance measurement of a dipole antenna mounted on a
bands,” in Proc. IEEE Wireless Power Transf. Conf. (WPTC), Jeju City, car tire,” in Proc. Int. Symp. Antennas Propag., Bangkok, Thailand,
South Korea, May 2014, pp. 64–67. Oct. 2009, pp. 1175–1178.
[21] K. Itoh, “RF bridge rectifier and its good possibility for wireless power [45] Hirose W.FL Datasheet, accessed on Jan. 1, 2016. [Online]. Available:
transmission systems,” in Proc. IEEE Int. Symp. Radio-Freq. Integr. https://www.hirose.co.jp
Technol. (RFIT), Sendai, Japan, Aug. 2015, pp. 226–228. [46] O. M. Ramahi, T. S. Almoneef, M. Alshareef, and M. S. Boybay,
[22] R. Vyas, H. Nishimoto, M. Tentzeris, Y. Kawahara, and T. Asami, “Metamaterial particles for electromagnetic energy harvesting,” Appl.
“A battery-less, energy harvesting device for long range scavenging Phys. Lett., vol. 101, no. 17, p. 173903, Oct. 2012.
of wireless power from terrestrial tv broadcasts,” in IEEE MTT-S Int. [47] H. Kamoda, M. Hanazawa, S. Kitazawa, H. Ban, N. Kukutsu, and
Microw. Symp. Dig., Montreal, QC, Canada, Jun. 2012, pp. 1–3. K. Kobayashi, “Design of rectenna array panel taking into account
[23] M. Pinuela, P. D. Mitcheson, and S. Lucyszyn, “Ambient RF energy mutual coupling for RF energy harvesting,” in Proc. IEEE Radio
harvesting in urban and semi-urban environments,” IEEE Trans. Microw. Wireless Symp. (RWS), Newport Beach, CA, USA, Jan. 2014, pp. 61–63.
Theory Techn., vol. 61, no. 7, pp. 2715–2726, Jul. 2013. [48] H. Kamoda, M. Hanazawa, S. Kitazawa, H. Ban, and K. Kobayashi,
[24] M. Pinuela, D. C. Yates, P. D. Mitcheson, and S. Lucyszyn, “London “Mutual coupling effect on rectenna array for RF energy harvesting,”
RF survey for radiative ambient RF energy harvesters and efficient in Proc. Eur. Microw. Conf. (EuMC), Nuremberg, Germany, Oct. 2013,
dc-load inductive power transfer,” in Proc. 7th Eur. Conf. Antennas pp. 503–506.
Propag. (EuCAP), Gothenburg, Sweden, Apr. 2013, pp. 2839–2843. [49] Y. Q. Wang and X. X. Yang, “Design of a high-efficiency circularly
[25] A. Mavaddat, S. H. M. Armaki, and A. R. Erfanian, “Millimeter-wave polarized rectenna for 35 GHz microwave power transmission system,”
energy harvesting using 4×4 microstrip patch antenna array,” IEEE in Proc. Asia–Pacific Power Energy Eng. Conf., Shanghai, China,
Antennas Wireless Propag. Lett., vol. 14, pp. 515–518, Feb. 2015. Mar. 2012, pp. 1–4.
[26] Y.-J. Ren and K. Chang, “5.8-GHz circularly polarized dual-diode [50] H. Sun and W. Geyi, “A new rectenna with all-polarization-receiving
rectenna and rectenna array for microwave power transmission,” IEEE capability for wireless power transmission,” IEEE Antennas Wireless
Trans. Microw. Theory Techn., vol. 54, no. 4, pp. 1495–1502, Jun. 2006. Propag. Lett., vol. 15, pp. 814–817, 2016.
[27] Y.-J. Ren and K. Chang, “Bow-tie retrodirective rectenna,” Electron. [51] M. E. Badawe and O. M. Ramahi, “Metasurface for near-unity electro-
Lett., vol. 42, no. 4, pp. 191–192, Feb. 2006. magnetic energy harvesting and wireless power transfer,” in Proc. IEEE
[28] J. A. Hagerty, F. B. Helmbrecht, W. H. McCalpin, R. Zane, Int. Symp. Antennas Propag. (APSURSI), Jun. 2016, pp. 609–610.
and Z. B. Popovic, “Recycling ambient microwave energy with broad- [52] J. Ye, C. Yang, and Y. Zhang, “Design and experiment of a rectenna
band rectenna arrays,” IEEE Trans. Microw. Theory Techn., vol. 52, array base on GaAs transistor for microwave power transmission,” in
no. 3, pp. 1014–1024, Mar. 2004. Proc. IEEE Int. Conf. Microw. Millim. Wave Technol. (ICMMT), vol. 1.
[29] T. S. Almoneef and O. M. Ramahi, “Can split-ring resonators be viable Jun. 2016, pp. 323–326.
for electromagnetic energy harvesting?” in Proc. IEEE Antennas Propag. [53] T. Matsunaga, E. Nishiyama, and I. Toyoda, “5.8-GHz stacked differ-
Soc. Int. Symp. (APSURSI), Memphis, TN, USA, Jul. 2014, pp. 424–425. ential rectenna suitable for large-scale rectenna arrays with dc connec-
[30] J. O. McSpadden, L. Fan, and K. Chang, “Design and experiments of tion,” IEEE Trans. Antennas Propag., vol. 63, no. 12, pp. 5944–5949,
a high-conversion-efficiency 5.8-GHz rectenna,” IEEE Trans. Microw. Dec. 2015.
Theory Techn., vol. 46, no. 12, pp. 2053–2060, Dec. 1998.
[31] T. Sakamoto, Y. Ushijima, E. Nishiyama, M. Aikawa, and I. Toyoda,
“5.8-GHz series/parallel connected rectenna array using expandable
differential rectenna units,” IEEE Trans. Antennas Propag., vol. 61,
no. 9, pp. 4872–4875, Sep. 2013. Faruk Erkmen (M’10), photograph and biography not available at the time
[32] AvagoTech. Avago HSMS-286x Datasheet, accessed on of publication.
Oct. 21, 2015. [Online]. Available: http://www.avagotech.com/products/
wireless/diodes/schottky/hsms-286k
[33] Murata. Murata LQW18an72ng00d Datasheet, accessed on
Jan. 10, 2016. [Online]. Available: http://www.murata.com/
Thamer S. Almoneef (GS’10), photograph and biography not available at
[34] T. S. Almoneef, H. Sun, and O. M. Ramahi, “A 3-D folded dipole
the time of publication.
antenna array for far-field electromagnetic energy transfer,” IEEE Anten-
nas Wireless Propag. Lett., vol. 15, pp. 1406–1409, Apr. 2016.
[35] J. Choo, J. Ryoo, J. Hong, H. Jeon, C. Choi, and M. M. Tentzeris,
“T-matching networks for the efficient matching of practical RFID tags,”
in Proc. Eur. Microw. Conf. (EuMC), Rome, Italy, Sep./Oct. 2009, Omar M. Ramahi (S’86–M’90–SM’00–F’09), photograph and biography not
pp. 5–8. available at the time of publication.