IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS—II: EXPRESS BRIEFS, VOL. 67, NO.

1, JANUARY 2020 87

A Reconfigurable Rectifier With Optimal Loading

Point Determination for RF Energy Harvesting
From −22 dBm to −2 dBm
Zizhen Zeng , Student Member, IEEE, Johan J. Estrada-López , Student Member, IEEE,
Mohamed A. Abouzied , Member, IEEE, and Edgar Sánchez-Sinencio, Life Fellow, IEEE

Abstract—This brief introduces a wide working range rectifier

with reconfigurable stage to accommodate with input power from
−22 dBm to −2 dBm and maintain high rectifier efficiency.
An auxiliary rectifier is designed for self-start-up functions. The
measured rectifier peak efficiency is 39% at −5 dBm, and sensi-
tivity efficiency is 7.5% at −22 dBm in static state. Self-start-up
performance difference is less than 4% in the normal work-
ing range. Compared to conventional fixed stage rectifier, our
design achieves 2× (+19%) peak efficiency improvement and
high-efficiency range (ηRect > 20%) is improved by 14 dB. And
a novel fractional open-circuit voltage approximation method is
also proposed to determine the optimized loading point in wide
input range to help in maximum power point tracking (MPPT).
Optimal fraction number is 0.5 in simulation and 0.6 is verified
in measurement. Compared to other designs in literature, our
proposed method shows better accuracy, simplicity, and avoiding
redundancy over working range from −22 dBm to −2 dBm, with
measured MPPT accuracy over 87%.
Index Terms—RF, energy harvesting, reconfigurable rectifier,
maximum power point tracking (MPPT), fractional open circuit
voltage (FOCV).
I. I NTRODUCTION
ECENTLY, RF Energy Harvesting (RF-EH) technol-
R ogy is arousing more and more attention where the
other ambient energy source, like solar, thermal, vibration
are absent. Moreover, RF energy harvester only needs an
antenna to harvest ambient Radio Frequency energy, which
is well accommodating with existing wireless communication
systems. However, ambient RF energy source could be variant
along with its power level. For example, the ambient GSM or
Wi-Fi average power is usually smaller than −30dBm [1], [2],
Manuscript received October 29, 2018; revised January 19, 2019; accepted
February 11, 2019. Date of publication February 14, 2019; date of current
version December 26, 2019. This work was supported in part by Intel, in part
by TI, in part by Silicon Labs, in part by CONACYT, and in part by UADY.
This brief was recommended by Associate Editor H. Sekiya. (Corresponding
author: Zizhen Zeng.)
Z. Zeng and E. Sánchez-Sinencio are with the Department of Electrical and
Computer Engineering, Texas A&M University, College Station, TX 77843
USA (e-mail: zzzeng@tamu.edu).
J. J. Estrada-López is with the Department of Electrical and Computer
Engineering, Texas A&M University, College Station, TX 77843 USA, and
also with the Universidad Autónoma de Yucatán, Mérida 97000, Mexico.
M. A. Abouzied is with Qualcomm, San Diego, CA, USA.
Color versions of one or more of the figures in this paper are available
online at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TCSII.2019.2899338
1549-7747 c 2019 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.
Fig. 1. RF energy harvesting system with dc/dc converter.
where rectenna with antenna array is employed to improve sen-
sitivity; the ambient energy from downlink devices can achieve
peak power to −20 dBm and even −10 dBm highly based on
user behaviors [1]; for far-field energy transfer or RFID appli-
cations, the available power could be as high as 0 dBm in
some extreme case [3]. Therefore, to design an RF energy
harvester that can accommodate with various input power lev-
els and maintain the optimal efficiency is very challenging.
Moreover, the rectifier voltage requirement and efficiency are
direct trade-off parameters. Typical rectenna design uses one-
stage rectifier to get high efficiency without considering output
voltage requirement. Conventional RFID rectifier design usu-
ally employs relative large stage number to ensure high output
voltage but efficiency and working range are compromised.
Compared to other high-efficiency energy harvesters [4], [5],
efficiency of RF energy harvester is still too low. To better
adjust in different RF energy harvesting applications, using
dc/dc converter to improve sensitivity with regulation func-
tions is shown in recent work [6]–[10], using reconfigurable
rectifier structure [11], [12] or employing different rectification
path for different input power [13].
Figure 1 shows a conventional RF harvester with dc/dc
converter and the bottleneck is still on the rectifier effi-
ciency itself. The typical efficiency of a rectifier is less than
20% in microwatt power conditions. What’s worse, rectifier’s
output voltage is changing with the input available power,
which could be below 0.2 V for low input power and over
3.0 V for high input power. Designing a dc/dc converter
with such large input voltage range would be a challenge
especially for the limited power budget. Therefore, differ-
ent up/down and both types of dc/dc converters have been
proposed for RF energy harvesting and it increases design
complexities [6]–[10]. Meanwhile, it is also very difficult to
maintain the rectifier’s optimal efficiency in such a wide input

Authorized licensed use limited to: Universiti Kebangsaan Malaysia. Downloaded on December 14,2021 at 08:58:20 UTC from IEEE Xplore. Restrictions apply.
88 IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS—II: EXPRESS BRIEFS, VOL. 67, NO. 1, JANUARY 2020
Fig. 2. Proposed 8-stage reconfigurable rectifier schematic.
power range. Therefore, a high-efficiency rectifier with effec-
tive control scheme for a wide input power range is vitally
important to improve the overall system end-to-end efficiency.
This brief is to design a wide input power range RF-EH
system with an accommodated optimal loading point deter-
mination method. Section II will show (A) Circuit design
and analysis; (B) Conventional method for optimal loading
and (C) Proposed FOCVA method. Measurement results are
summarized in Sections III and IV concludes this brief.
II. P ROPOSED S OLUTION AND I MPLEMENTATION
Figure 2 shows the reconfigurable rectifier design. The rec-
tifier topology is based on Dickson rectifier with thin-oxide
native NMOS. 8 stages are chosen to provide 8, 4, 2 and 1-
stage configuration by digital control. In low power mode, all
the stages are cascading to generate higher voltage by increas-
ing output resistance. In high power mode, all the stages are in
parallel to generate a current combination with smaller output
resistance. The dual path multiplexer provides the path to next
stage or bypass it to final output. To achieve self-start up and
self-sustain functions in low voltage, an auxiliary two-stage
rectifier is designed to power up the controller internally. To
handle large voltage stress, thick-oxide native NMOS is used
in auxiliary rectifier.
Compared to other similar configuration designs in [11]
and [12], which both require certain output voltage at 1 V
with large output capacitor and complicate configure method,
this brief can achieve much lower start-up voltage as 0.4 V,
more simple and wide range method to detect optimal load-
ing point, generating moderate output voltage range suitable
for various RF energy harvesting application interfaced with
dc/dc converter. An off-chip controller, AT-mega MCU is used
to control the rectifier and evaluate the performance of the
proposed MPPT method. The MCU power is excluded in this
brief because the MCU is only used for design validation and
the system can be implemented with fully-integrated solution.
A. Reconfigurable Rectifier Output Impedance Analysis
All the rectifier configurations are shown in figure 3, all the
multiplexer switches are simplified as parasitic resistance. Reff
represents the effective output resistance of the RF-DC rectifier
for optimal power extraction. Neff is the effective rectifier stage
Authorized licensed use limited to: Universiti Kebangsaan Malaysia. Downloaded on December 14,2021 at 08:58:20 UTC from IEEE Xplore. Restrictions apply.
Fig. 3. Reconfiguration output resistance analysis.
number. RF port is input interfaced with matching network
and antenna. The output is connected to 20 pF on-chip capac-
itor and off-chip loading resistance. Although Reff itself highly
depends on technologies and design parameters [14], [15], it
can be controlled effectively by rectifier stage Neff . Simply,
each rectifier single cell could be regarded as dc voltage source
with internal resistance Rint . Here we assume Rint is constant
to provide straightforward intuition for analysis. The open-
circuit output voltage along with effective Reff is changed by
parallel/cascade configuration. Normally Rint is ranging from
10 to 100 K therefore the parasitic Ron is negligible. For dif-
ferent configurations from fixed N0 = 8 stage, Neff equals to
1, 2, 4 and 8. And the Reff will be Rint /8, Rint /2, Rint ∗ 2, and
Rint ∗ 8 respectively, which means the output power is higher
with smaller Neff with similar output voltage. For example
when Neff = 4, two 4-stage rectifiers are connected in parallel
therefore the effective Reff is Rint ∗ 2.
Although rectifier operation is nonlinear, it can be approxi-
mated as resistive source for certain working range [7]. Such
approximation is still valid for wide range operation because
each rectifier configuration only covers for limited working
range. Ideally, when the loading resistance is close to rectifier
output impedance, the optimal loading point appears near half
of the open-circuit voltage. Therefore the MPPT is achieved
in wide range operation by adjusting the equivalent loading
resistance through the cascading dc/dc converter.
To better explain the effect on efficiency, our assumption is
(1) All the transistors and capacitors are identical. (2) Body
effect is ignored. (3) Parasitical resistance and capacitor due
for reconfiguration multiplexer are ignored. (4) Stage rectifier
cell performance is independent of output current and output
voltage. (5) Rectifier input impedance from RF port is similar
among different configurations [14]–[16]:
C
VOUT = Neff ∗ ( ∗ VRF − Vth − Vloss ) (1)
C + Cpar
2
VOUT
Rload
ηRect = 2
(2)
VOUT
Rload + Ploss ∗ Neff
ZinStage
ZinRect = (3)
N0
C is the stage capacitor in each stage rectifier and Cpar is
representing the parasitic capacitor. VRF is the voltage swing
seen in rectifier’s input. Vth is the transistor threshold voltage
and Vloss is representing voltage drop due to loading current
ZENG et al.: RECONFIGURABLE RECTIFIER WITH OPTIMAL LOADING POINT DETERMINATION 89

and leakage current. ηRect is RF-DC power conversion effi-

ciency of rectifier and our design focus is only on ηRect because
it is the dominant parameter in total efficiency. Typically, VOUT
is maximized in certain stage number Neff and VRF to get max-
imum efficiency by decreasing Cpar , Vth , and Vloss . Although
the internal switch implementation will increase Ploss , Neff is
still the dominant factors for ηRect . This also explains why
designers prefer smaller Neff for target output voltage or power
to get better efficiency. And when output voltage is exceeding Fig. 4. Comparison of optimal loading and VOUT in 8-stage and 1-stage.
required voltage level, decreasing Neff by reconfiguration can
maximize the efficiency. ZinRect in (3) is the input impedance
from the RF input port. It is reverse proportional to N0 because
all the stage rectifier is equivalent to parallel connection to the
RF port. As long as each rectifier is connected to the RF port
through the coupling capacitor, this N0 will remain unchanged
in equation (3). As pointed out in [17], ZinRect is also depen-
dent on output voltage VOUT . Therefore, changing Neff to limit
VOUT changes can also help to suppress the changes of ZinRect .
Our design starting point is based on equation (1) and (2),
with target input power and output voltage requirement, the
efficiency is maximized by sizing and selection of stage num-
ber. And then the working range is extended thanks to the
reconfigurable stage number. Fig. 5. FOCVA efficiency accuracy (simulation).

C. Fractional Open-Circuit Voltage Approximation (FOCVA)

B. Conventional Optimal Loading Determination
Conventionally, the optimal loading resistance for RF
energy harvester is difficult to model or predict because of
its non-linear nature. And its value together with the ratio to
open circuit voltage VOC is also changing with input power,
which makes it is more difficult for tracking [6]. Although
observed in [7], the optimal loading resistance is almost con-
stant, it is only valid in a very limited input power range.
Therefore, in various RF energy harvester publications, differ-
ent loading resistance is used to optimize for different input
power level [18], [19]. However, the methodology to choose
such optimal resistance is still missing. The optimal loading
point is even more difficult to track for wide input power range
and in [10], a time-based MPPT for buck-booster is used to
estimate the input power and overcome the nonlinear effect
but it requires large capacitor array and long integration time
to achieve wide range operation. All the previous MPPT meth-
ods are proposed rather for limited operation range or required
complicated circuitry and control. We observed each rectifier
configuration is working only for certain limited input power
range, therefore a constant loading resistance for each rectifier
configuration can be used. Then each limited working range
can be combined without overlap to generate a large input
power range with the help of reconfiguration.
Figure 4 shows the simulation results of the optimal output
voltage and corresponding loading resistance for the 8-stage
and 1-stage configurations. By sweeping VRF from 0.2 V to
0.5 V, the equivalent input power is similar, ranging from 5 to
50 μW. The optimal loading resistance is changed up to 67%
with fixed stage design, causing the complexity for maximum
power point tracking. However, if the VRF range is limited
(VRF = 0.2 to 0.3 V in 8-stage, 0.4 to 0.5 V in 1-stage),
the optimal loading resistance changes will be minimized,
therefore the fixed optimal loading can be used.
To determine the optimal loading point without introducing
too much complexity, a novel FOCV Approximation method
is proposed for wide input power range operation. Although
FOCV is not typically used in RF energy harvesting applica-
tions, analysis and simulation results indicate its effectiveness
with such quasi-resistive approximation.
As shown in Figure 5, simulation results show a ratio of
0.5 is a very good approximation to get maximum output effi-
ciency across different input power, overall efficiency accuracy
is above 95% compared to maximum efficiency point. Another
observation is that optimal loading resistance is unchanged
with such method, which also justified our previous state-
ment that in small input power range, the optimal loading
resistance is almost constant. Efficiency accuracy here is the
ratio between ηPCE of fractional VOC method (VO /VOC = 0.5)
and captured maximum ηPCE through different loadings. Test-
bench is also including a L-C matching network and the
inductor Q is 30, which shows the proposed FOCVA Method
can also work along with fixed matching network.
III. M EASURED R ESULTS
The whole system is fabricated in the IBM-130nm process
and the reconfigurable rectifier layout area is 450μm*70μm
including two-stage auxiliary rectifier. The digital controller
for switch configuration is 80μm*10μm. A reference design
of 8-stage fixed rectifier is also fabricated for comparison.
The reference design layout is 180μm*60μm with identical
stage rectifier cell. Figure 6 shows the test bench and design
layout photograph. To evaluate the real absorbed power, effec-
tive input power is calibrated with S11 parameters by Vector
Network Analyzer. ηRect with external fixed power source is
plotted. And then external power source is placed with the
internal VAUX , representing the self start-up performance.

Authorized licensed use limited to: Universiti Kebangsaan Malaysia. Downloaded on December 14,2021 at 08:58:20 UTC from IEEE Xplore. Restrictions apply.
90 IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS—II: EXPRESS BRIEFS, VOL. 67, NO. 1, JANUARY 2020
Fig. 6. Circuit test-bench and design chip micrograph.
Fig. 7. Rectifier Zin real and imaginary part measurement.
Fig. 8. (a)Rectifier VOUT (b) FOCVA MPPT accuracy.
In figure 7, the Zin real and imaginary part is plotted from
S11 parameter measurement at 820 MHz with different config-
urations. The typical value of Zin is 2.3 + 15.9∗j ohm because
of parasitic inductance from bonding wire, LQFP package and
FR4 PCB trace. Both the real and imaginary part change with
the loading. It explains why the rectifier with fixed matching
efficiency degrades in higher loading current conditions. And
it also shows that the reconfiguration design can suppress this
S11 shifting in a larger input power range thus will suppress
the matching loss.
Figure 8a shows the rectifier output voltage VOUT when
fixed loading is applied within each configuration. The out-
put voltage is ranging from 0.3 V to 1.0 V. It is exceeding
the target output voltage of 0.3 to 0.72 V (ratio of 0.6)
because, in the high-power range, the optimal resistance from
FOCVA is decreased. However, for a better comparison, recti-
fier performance with fixed loading is used considering the
efficiency absolute difference is <3% compared to optimal
resistance.
Figure 8b summarizes our proposed FOCVA MPPT accu-
racy performance in target input power. The maximum error
compared to 0.6 is 0.08, thus the overall accuracy is over
86.7%, where MPPT accuracy is defined as the ratio of
the real measured fractional value of the optimal point to
the predefined ratio 0.6, indicating the effectiveness of our
proposed method. Rectifier open-circuit output voltage VOC
Authorized licensed use limited to: Universiti Kebangsaan Malaysia. Downloaded on December 14,2021 at 08:58:20 UTC from IEEE Xplore. Restrictions apply.
Fig. 9. Rectifier power conversion efficiency comparison.
is kept in 0.5 to 1.2 V by controlling the Neff stage num-
ber, then different loading resistance is selected for VOUT to
reach the target fractional value. Measurement results show
that VOUT = VOC ∗ 0.6 is a good estimation value to deter-
mine the optimal loading from −22 to −2 dBm, 100X ranging
from 6.3 μW to 630 μW. The maximum voltage for the pro-
cess is 1.2 V and if maximum voltage requirement is relaxed,
this FOCVA method and the rectifier working range, could
be further increased to higher power level. The reason frac-
tional ratio changes from 0.5 in the simulation because here the
real input power is calibrated by S11 parameter to characterize
rectifier performance, assuming ideal matching with tracking
functions as [17]. The loading resistance is almost constant
with every configuration excluding some extreme case.
Figure 9 shows the measurement results of rectifier power
conversion efficiency among fixed-stage reference and recon-
figurable design with or without self start-up. External Vdd
performance is used to compare because the VAUX can be
replaced with stable voltage source in steady-stage in dc/dc
converter involved applications and the power consumption of
the controller is negligible compared to output power, mainly
dominant by the leakage current. The fixed stage rectifier is
identical to 8-stage rectifier without any switches for a fair
comparison. For reference design’s output, the requirement is
relaxed to be 0.3 to 1.2 V without detecting VOC .
Measurement results show our proposed reconfigurable rec-
tifier can work with input power from −22 dBm to −2 dBm.
Self-start-up performance difference is less than 2% except
for low-power range (−22 to −20 dBm) and extreme high-
power range (−6 to −2 dBm). The first case is because, in
low input power, the auxiliary rectifier output VAUX is not high
enough to minimize switch conduction resistance. The second
case is because the leakage current through ESD diode is more
severe with higher VAUX . However, from −20 dBm to −2 dBm
working range, the whole efficiency difference is less than 4%.
Each reconfigurable stage has very small overlap region with
its adjacent states with proposed FOCVA, which avoids state
redundancy compared to other similar counterparts without
any clear boundaries. To have better performance in low-power
range, more auxiliary stages can be used but efficiency loss is
more severe in high-power range.
Compared with the reference design, the efficiency for
fixed stage design is only better in extreme case smaller than
ZENG et al.: RECONFIGURABLE RECTIFIER WITH OPTIMAL LOADING POINT DETERMINATION
TABLE I
L ITERATURE C OMPARISON AND S UMMARY
−20 dBm. Overall, our proposed reconfigurable design is bet-
ter from −15 dBm to −2 dBm. The comparison shows that our
proposed design can achieve 2X peak efficiency improvement
(+19%) with a high-efficiency range extension (+14 dB),
sacrificing little sensitivity performance (<3%). It means our
purposed reconfigurable design is extending the working range
significantly and it can be configured to work with different
applications, from microwatt ambient energy harvest, upper
link energy harvest, and potential wireless power transfer or
short distance RFID applications.
Table I summarizes the overall performance and com-
parison in literature. This brief shows best sensitivity in
−22 dBm and higher peak power efficiency of 39% compared
to [11] in static state performance. Overall peak efficiency
is improved by 19% and high efficiency working range is
extended by 14 dB compared to reference design. Compared
to work [3], [13], this brief can achieve both wide input range
and peak efficiency improvement simultaneously. For dc/dc
regulator assisted energy harvesting applications, this brief can
provide a moderate output voltage from 0.3 to 1.0 V with
fixed loading resistance for each configuration, and 0.3 to
0.72 V with FOCVA ratio = 0.6, improving rectifier efficiency
and relieve following booster voltage gain requirements. For
example, for a typical 1.2 V supply voltage requirement, it
only needs boost gain from 1.67 to 4, while other work in
comparison will need up and down voltage converter.
IV. C ONCLUSION
A RF-EH system based on reconfigurable rectifier is
designed for wide input power range and maintain high effi-
ciency from −22 dBm to −2 dBm. The final output is
designed to be 0.3 to 0.7 V with FOCVA and suitable for
following boost converter to perform MPPT. Compared to
conventional Dickson rectifier, this design peak efficiency
equals to 39%, with 19% peak efficiency improvement and the
high-efficiency range >20% is from −16 dBm to −2 dBm.
With the help of reconfigurable structures, Fractional Open-
Circuit Voltage Approximation Method (FOCVA) is proposed
for optimal loading determination in a wide input power
range. Simulation and measurement results show our proposed
method can achieve good MPPT accuracy in the whole range
from −22 dBm to −2 dBm, with MPPT accuracy over 86%.
Authorized licensed use limited to: Universiti Kebangsaan Malaysia. Downloaded on December 14,2021 at 08:58:20 UTC from IEEE Xplore. Restrictions apply.
91
Compared existing MPPT method in literature, our method
can achieve simplicity and accuracy simultaneously. In future
work, a dc/dc converter could be design with MPPT functions
based on our proposed FOCVA method to maximize the power
extraction.
R EFERENCES
[1] K. Mimis, D. R. Gibbins, S. Dumanli, and G. T. Watkins, “The ant
and the elephant: Ambient RF harvesting from the uplink,” IET Microw.
Antennas Propag., vol. 11, no. 3, pp. 386–393, Feb. 2017.
[2] S. Shen, C.-Y. Chiu, and R. D. Murch, “A dual-port triple-band L-probe
microstrip patch rectenna for ambient RF energy harvesting,” IEEE
Antennas Wireless Propag. Lett., vol. 16, pp. 3071–3074, 2017.
[3] A. K. Moghaddam et al., “A 73.9%-efficiency CMOS rectifier using a
lower DC feeding (LDCF) self-body-biasing technique for far-field RF
energy-harvesting systems,” IEEE Trans. Circuits Syst. I, Reg. Papers,
vol. 64, no. 4, pp. 992–1002, Apr. 2017.
[4] Q. Wan, Y.-K. Teh, Y. Gao, and P. K. T. Mok, “Analysis and design
of a thermoelectric energy harvesting system with reconfigurable array
of thermoelectric generators for IoT applications,” IEEE Trans. Circuits
Syst. I, Reg. Papers, vol. 64, no. 9, pp. 2346–2358, Sep. 2017.
[5] X. Liu, L. Huang, K. Ravichandran, and E. Sánchez-Sinencio, “A highly
efficient reconfigurable charge pump energy harvester with wide harvest-
ing range and two-dimensional MPPT for Internet of Things,” IEEE J.
Solid-State Circuits, vol. 51, no. 5, pp. 1302–1312, May 2016.
[6] K. R. Sadagopan, J. Kang, Y. Ramadass, and A. Natarajan, “A 960pW
co-integrated-antenna wireless energy harvester for WiFi backchan-
nel wireless powering,” in Proc. IEEE Int. Solid-State Circuits Conf.
(ISSCC), San Francisco, CA, USA, 2018, pp. 136–138.
[7] T. Paing, J. Shin, R. Zane, and Z. Popovic, “Resistor emulation approach
to low-power RF energy harvesting,” IEEE Trans. Power Electron.,
vol. 23, no. 3, pp. 1494–1501, May 2008.
[8] G. C. Martins and W. A. Serdijn, “An RF energy harvester with MPPT
operating across a wide range of available input power,” in Proc. IEEE
Int. Symp. Circuits Syst. (ISCAS), 2018, pp. 1–5.
[9] T.-C. Huang et al., “A battery-free 217 nW static control power buck
converter for wireless RF energy harvesting with α-calibrated dynamic
on/off time and adaptive phase lead control,” IEEE J. Solid-State
Circuits, vol. 47, no. 4, pp. 852–862, Apr. 2012.
[10] X. Hua and R. Harjani, “A 5uW-5mW input power range, 0-3.5V out-
put voltage range RF energy harvester with power-estimator-enhanced
MPPT controller,” in Proc. IEEE Custom Integr. Circuits Conf.,
San Diego, CA, USA, 2018, pp. 1–4.
[11] M. A. Abouzied, K. Ravichandran, and E. Sánchez-Sinencio, “A
fully integrated reconfigurable self-startup RF energy-harvesting system
with storage capability,” IEEE J. Solid-State Circuits, vol. 52, no. 3,
pp. 704–719, Mar. 2017.
[12] Z. Zeng et al., “A WLAN 2.4-GHz RF energy harvesting system
with reconfigurable rectifier for wireless sensor network,” in Proc.
IEEE Int. Symp. Circuits Syst. (ISCAS), Montreal, QC, Canada, 2016,
pp. 2362–2365.
[13] Y. Lu et al., “A wide input range dual-path CMOS rectifier for RF
energy harvesting,” IEEE Trans. Circuits Syst. II, Exp. Briefs, vol. 64,
no. 2, pp. 166–170, Feb. 2017.
[14] J. Yi, W.-H. Ki, and C.-Y. Tsui, “Analysis and design strategy of UHF
micro-power CMOS rectifiers for micro-sensor and RFID applications,”
IEEE Trans. Circuits Syst. I, Reg. Papers, vol. 54, no. 1, pp. 153–166,
Jan. 2007.
[15] M. A. Abouzied and E. Sánchez-Sinencio, “Low-input power-level
CMOS RF energy-harvesting front end,” IEEE Trans. Microw. Theory
Techn., vol. 63, no. 11, pp. 3794–3805, Nov. 2015.
[16] S. Shieh and M. Kamarei, “Transient input impedance modeling of rec-
tifiers for RF energy harvesting applications,” IEEE Trans. Circuits Syst.
II, Exp. Briefs, vol. 65, no. 3, pp. 311–315, Mar. 2018.
[17] M. Stoopman, S. Keyrouz, H. J. Visser, K. Philips, and W. A. Serdijn,
“Co-design of a CMOS rectifier and small loop antenna for highly sen-
sitive RF energy harvesters,” IEEE J. Solid-State Circuits, vol. 49, no. 3,
pp. 622–634, Mar. 2014.
[18] L.-X. Liu et al., “An ultra-low-power integrated RF energy harvesting
system in 65-nm CMOS process,” Circuits Syst. Signal Process., vol. 35,
no. 2, pp. 421–441, Feb. 2016.
[19] L. Liu et al., “A dual band RF energy harvester with hybrid threshold
voltage self-compensation,” J. Circuits Syst. Comput., vol. 25, no. 6,
Jun. 2016, Art. no. 1650055.