6904 IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, VOL. 69, NO.

9, SEPTEMBER 2020

Effects of Different Modulation Techniques on

Charging Time in RF Energy-Harvesting System
Mustafa Cansiz , Dogay Altinel , and Gunes Karabulut Kurt , Senior Member, IEEE

Abstract— Radio-frequency (RF) energy harvesting is a defense, building automation, security, industrial monitoring,
promising technology that obtains energy from electromagnetic agriculture, and data center. Commercial RF energy-harvesting
waves. As an alternative energy source, this modern technology module from Powercast Company [6] can charge AA or AAA
can provide power wirelessly enabling the practical use of
battery-free devices and it is also expected to be widely used size batteries. Besides, RF energy-harvesting technology can
in future applications. Many studies have examined the effects also provide operational autonomy for wireless sensor net-
of different modulation techniques on power or efficiency in RF works and Internet of Things. It can benefit from sensors to
energy-harvesting systems. In this article, the effects of different obtain data, such as mineral content, humidity, and temperature
modulation techniques, such as amplitude shift keying (ASK), of the soil. As an agricultural application, wireless sensor
quadrature amplitude modulation (QAM), phase-shift keying
(PSK), and frequency shift keying (FSK) on charging time, were networks were placed in an outdoor field dispersedly and these
measured and theoretically analyzed in detail. The measurement sensors were used to monitor the pests [7]. Moreover, an LCD
system was designed using a signal generator, an RF energy- display was powered wirelessly using this technology [8].
harvesting circuit, and other auxiliary devices. The signal gener- An RF energy-harvesting circuit mainly consists of an
ator produced four different modulated signals as transmit power antenna, a rectifier, a voltage multiplier, and an energy storage
signals from 1 to 10 dBm at the interval of 1 dBm, and then,
the RF energy-harvesting circuit was fed with these power signals. device. Antenna is a transducer that converts electric field
According to the measurement results, four different modulation strength to voltage difference or vice versa. Rectifier is the
techniques charged the RF energy-harvesting circuit at different most important part of the RF energy-harvesting circuit that
times. It was observed that the 4-ASK modulation technique converts RF power to dc power, and it has a significant impact
had the shortest charging time in the range of 1–10 dBm on system efficiency. Voltage multiplier provides a higher
when comparing to QAM, PSK, and FSK modulation techniques,
and thus, it was evaluated to be the most suitable modulation output dc voltage level to drive a sensor or an energy storage
technique for the RF energy-harvesting system among these device. Finally, energy storage devices, such as supercapaci-
modulation techniques in terms of charging time. tors or batteries, are used to store the harvested energy. Before
Index Terms— Charging time, energy harvesting, measure- designing an RF energy-harvesting circuit, power densities
ment, modulation technique, radio frequency (RF). of electromagnetic fields in the ambient environment must
be measured and investigated. The most powerful frequency
I. I NTRODUCTION band or bands are preferred for the RF energy-harvesting
circuits. For this purpose, many electromagnetic spectral mea-
T HE number of electronic devices that make our lives
easier is increasing with the rapid development of tech-
nology. These electronic devices need energy to operate.
surements were conducted in various countries [9], [10].
In the literature, there are many studies on RF energy
Solar [1], thermal gradients [2], mechanical vibrations [3], and harvesting [11]–[13]. The amount of harvested energy is
electromagnetic waves [4], [5] are significant energy sources. directly proportional to the conversion efficiency that reflects
Energy harvesting as an alternative energy source has recently the performance of RF energy-harvesting circuits [14]. In some
attracted more attention. In addition, electromagnetic or RF studies [15]–[17], the conversion efficiency was improved by
energy harvesting is a promising technology to meet especially utilizing signals with different waveforms. In [16], signals with
the energy needs of low-power devices. In this context, this various waveforms, such as orthogonal frequency-division
technology is used in many practical application fields, such as multiplexing (OFDM), Gaussian white noise, and chaotic
signals, were evaluated at 433 MHz. When compared with
Manuscript received October 12, 2019; revised February 7, 2020; accepted the single-carrier signal, it was determined that the waveforms
February 18, 2020. Date of publication March 4, 2020; date of current version
August 11, 2020. The Associate Editor coordinating the review process was with a high peak-to-average power ratio (PAPR) increased the
Helko van den Brom. (Corresponding author: Mustafa Cansiz.) conversion efficiency of the rectifier. Similarly, the conversion
Mustafa Cansiz is with the Department of Electrical and Electron- efficiency values were compared in the presence of single-
ics Engineering, Dicle University, 21280 Diyarbakır, Turkey (e-mail:
[email protected]). tone, OFDM, white noise, and chaotic signals, where the
Dogay Altinel is with the Department of Electrical and Electronics Engi- chaotic signal showed the best performance in [17]. RF
neering, Istanbul Medeniyet University, 34700 Istanbul, Turkey (e-mail: signals can also be utilized as multisine or more generally
[email protected]).
Gunes Karabulut Kurt is with the Department of Electronics and Commu- multitone signals in RF energy-harvesting systems [18]–[20].
nication Engineering, Istanbul Technical University, 34469 Istanbul, Turkey Based on simulations and experimental studies, it was proved
(e-mail: [email protected]). that multitone signals can improve the conversion efficiency
Color versions of one or more of the figures in this article are available
online at http://ieeexplore.ieee.org. for the same average power. The waveforms of multitone
Digital Object Identifier 10.1109/TIM.2020.2978360 signals can be optimized to obtain high performance in RF
0018-9456 © 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See https://www.ieee.org/publications/rights/index.html for more information.

Authorized licensed use limited to: Gebze Technical University. Downloaded on May 02,2023 at 08:29:20 UTC from IEEE Xplore. Restrictions apply.
CANSIZ et al.: EFFECTS OF DIFFERENT MODULATION TECHNIQUES ON CHARGING TIME IN RF ENERGY-HARVESTING SYSTEM 6905

energy-harvesting systems [21]–[23]. It was shown that power- an energy harvester with a wireless channel between them.
optimized waveforms increased the effective distance in radio- Denoting the transmitted signal, the channel gain, and the
frequency identification (RFID) systems. A waveform-aware noise as x(t), h(t), and n(t), respectively, the received signal
RF energy harvester was designed to optimize its performance y(t) is expressed as
depending on the excited signals in [24]. RF signals with
widely used modulation techniques, including amplitude shift y(t) = h(t) α(β, d) x(t) + n(t) (1)
keying (ASK), phase-shift keying (PSK), GSMK, and quadra- where α(β, d) is the path loss depending on the path loss
ture amplitude modulation (QAM), have also been investigated exponent β and the distance d between the transmitter and
for RF energy-harvesting systems [25]–[30]. It was stated that the energy harvester. n(t) can be considered as negligible for
the complementary cumulative distribution function (CCDF) energy harvesting [37]. Assuming narrowband block fading
and the PAPR were important indicators of the conversion effi- during transmission (h(t) = h) for given distance and path
ciency. CCDF curves are used to characterize the power of RF loss exponent values (α(β, d) = α), the received power Pr is
signals. The conversion efficiency increases with the increase stated as
in PAPR of input RF signals. In [31], a modulation technique
based on two-tone signals was proposed for simultaneous Pr = ηt |h|2 α2 |x|2 (2)
wireless information and power transfer (SWIPT) systems.
The detailed information on the effects of wireless power where |·|2 represents the power of related parameter and ηt
waveforms, such as single-tone and multitone modulated sig- is the RF energy transmitting efficiency. Its value changes
nals and other different waveforms, is available in [32]. according to the modulation techniques.
The energy obtained from the RF signals can be used On the one hand, for a CW signal, the transmit power
to charge energy storage units, such as rechargeable batter- remains constant (Pt = |x|2 ), and the change in the received
ies or supercapacitors. The powerful signals charge the energy power is mainly determined by the channel power gain.
storage unit of an RF energy-harvesting circuit in less time Depending on the CW signal, the received power becomes
than weak signals. In [33], it was shown that the measured Pr = ηt |h|2 α2 Pt . (3)
charging time increased with the increase in distance which
caused the signal power to weaken. In [34], the charging time On the other hand, for a modulated signal, the transmit
in RF energy-harvesting systems was investigated to derive power changes symbol by symbol depending on the associ-
statistical expressions based on the system parameters. In [35], ated modulation techniques, such as ASK, QAM, PSK, and
an experiment was conducted by using continuous RF signals FSK. Let S denote the symbol constellation of a modulation
with constant power to charge a supercapacitor, and the charg- technique, including M symbols. The symbol constellation is
ing time distribution was analyzed. In [36], it was theoretically expressed as
shown that the selected modulation technique based on a
S = {x1 , x2 , . . . , xi , . . . , xM } (4)
symbol constellation changed the power distribution of the
received signal and correspondingly the charging time for the where xi is the ith symbol of a modulation technique,
RF energy-harvesting circuit. Although these studies provide i = 1, 2, . . . , M . From the perspective of RF energy har-
information about the charging time in RF energy harvesting, vesting, the symbol constellation includes K subsets repre-
they do not give detailed information about the effect of sented by Sk , k = 1, 2, . . . , K. Each subset is defined as
modulation techniques on the charging time and there is a Sk = {xi ∈ S| |xi | = Pi = Pk }, where Pi is the
lack of experimental studies in the literature. transmit power of the ith symbol and Pk is the transmit
How often the sensors plugged into an RF energy-harvesting power of symbols with the same (the kth) power value. In M -
circuit send the measured data are directly related to the ASK, M -QAM, M -PSK, and M -FSK modulation techniques,
charging time. In this context, charging time is essential for an the amplitude and shape of symbols determine the power of
RF energy-harvesting system. In this article, it was aimed to symbols. The amplitudes of symbols remain constant in M -
measure and investigate the effects of four different modula- PSK and M -FSK. The symbols of M -ASK and M -QAM can
tion techniques, such as ASK, QAM, PSK, and frequency shift be expressed as
keying (FSK) on the charging time in the RF energy-harvesting 
system. It was determined that each modulation technique Ai gi (t), M -ASK
xi (t) = (5)
charged the energy storage unit at different times. Therefore, (AIi + jAQi ) gi (t), M -QAM
each modulation technique may have different charging time
values. In Section II, a theoretical analysis of the system is where gi (t) is the signal pulse shape and Ai , AIi , and AQi
given. In Section III, the measurement system is presented in are the amplitudes. Based on these amplitudes, the transmit
detail. In Section IV, the charging time and received power power of the ith symbol is expressed for M -ASK as
of four different modulation techniques are experimentally 3Pat
analyzed. Conclusions are presented in Section V. Pi = Ai 2 (6)
M2 − 1
II. T HEORETICAL A NALYSIS for M -QAM as
In the most general sense, the system model of RF energy 3Pat
Pi = (A2 + A2Qi ) (7)
harvesting can be considered as a pair of a transmitter and 2(M − 1) Ii

Authorized licensed use limited to: Gebze Technical University. Downloaded on May 02,2023 at 08:29:20 UTC from IEEE Xplore. Restrictions apply.
6906 IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, VOL. 69, NO. 9, SEPTEMBER 2020

and for M -PSK and M -FSK as

Pi = Pat (8)
where P at denotes the average
K transmit power of symbols,
M
Pat = i=1 Pi /M = k=1 M k Pk /M . The parameter of
Mk is the cardinality of the kth subset Sk . The subsets of
symbol constellations and their cardinalities change based on
the modulation techniques. For 16-QAM modulation, as an
example, there are three subsets according to the radii of
symbols and their cardinalities are 4, 8, and 4 in the order
of power values. Depending on the ith symbol of a modulated
signal, the received power is expressed as
Pr,i = ηt |h|2 α2 Pi (9)
and the harvested power becomes
Ph,i = ηh (ηt |h|2 α2 Pi − Ps ) (10) Fig. 1. Measurement system.

where Ps denotes the power sensitivity level of RF energy this relationship, the charging time (Tc ) can be expressed as
harvester. For the received power values below Ps , it is not γ
possible to harvest energy in practical RF energy-harvesting Tc = (14)
Ph
circuits. ηh is the RF energy-harvesting efficiency. It is equal to
where γ is the conversion coefficient. It is modeled depending
zero for negative values of (Pr,i − Ps ) and changes between
on the capacitor parameters. The conversion coefficient can be
zero and one as a function of power and frequency (fi ) of
defined as
the ith symbol, 0 ≤ ηh (Pr,i , fi ) ≤ 1. Considering a transmit
probability for each symbol, the harvested power from all Cc Vc 2 Dc
γ= (15)
symbols is stated as 2
M where Dc is the discharge depth of capacitor. Thus, the average

Ph = pi ηh (ηt |h|2 α2 Pi − Ps ) (11) charging time (Tac ) becomes
i=1 Cc Vc 2 Dc
Tac = M . (16)
where pi is the transmit probability of the ith symbol. 2 pi ηh (ηt |h|2 α2 Pi − Ps )
i=1
Both (10) and (11) give the instantaneous values
of Ph,i and Ph . In these equations, the channel power
gain is a random variable (z = |h|2 ). It can be modeled for III. M ATERIALS AND M ETHOD
power depending on the fading effect of the channel. In this In this section, the impacts of four different modulation
case, the average harvested power for modulated signals can techniques on the charging time of the RF energy-harvesting
be calculated by using the probability density function (PDF) circuit were measured and recorded. Besides, the measurement
of z (fz (z)) as system was described in detail and the collection of the
M  ∞ measurement data was explained.
Pah = pi ηh (ηt z α2 Pi − Ps )fz (z)dz. (12)
i=1 0 A. Measurement
This integral takes the value of zero for the values between The measurement system consists of NI PXIe-1082 from
0 and Ps /α2 Pi of z due to ηh = 0. As a transmission cable National Instruments [38], P2110 Powerharvester module from
between the transmitter and the energy harvester is used in Powercast Company [6], WSN-Eval-01 wireless sensor board,
the measurements of this article, the distribution of z is not XLP 16-bit development board, and PICtail daughter card. All
considered in detail. measurement devices are shown in Fig. 1.
In practical RF energy-harvesting systems, the amount of NI PXIe-1082 was used to generate four different modulated
harvested energy is not sufficient to power a wireless device signals (ASK, QAM, PSK, and FSK) at 915-MHz carrier fre-
continuously. Thus, it is necessary to use energy storage quency. These modulation techniques are the most widely used
devices, such as rechargeable batteries or supercapacitors. The in radio communication. Pseudo-noise (PN) sequence order is
energy stored in a capacitor (Ec ), when fully charged, can be the same for four different modulated signals generated by the
calculated in terms of the capacity of capacitor (Cc ) as signal generator. M values for ASK, QAM, PSK, and FSK
were selected as 4, 16, and 32. According to the charging
Cc Vc 2
Ec = (13) time, each modulation technique was compared for different
2 M values, and then, different modulation techniques were
where Vc is the charging voltage. The charging time (Tc ) is compared with each other. In other words, the charging time
inversely proportional to the harvested power [34]. Based on values of the 4-ASK, 16-ASK, and 32-ASK were compared

Authorized licensed use limited to: Gebze Technical University. Downloaded on May 02,2023 at 08:29:20 UTC from IEEE Xplore. Restrictions apply.
CANSIZ et al.: EFFECTS OF DIFFERENT MODULATION TECHNIQUES ON CHARGING TIME IN RF ENERGY-HARVESTING SYSTEM 6907

TABLE I
C HARGING T IME M EASUREMENTS FOR D IFFERENT M ODULATION T ECHNIQUES

with each other, and then, they were compared with the other Instruments [38], and then, the relationship between charging
modulation techniques. The modulated signal as a transmit time and received power was explained in detail.
power signal was directly connected to the P2110 Powerhar-
vester module with a coaxial cable (right circuit on the table as B. Collecting Data
shown in Fig. 1). The transmit power signal from 1 to 10 dBm When the P2110 Powerharvester module stores sufficient
at the interval of 1 dBm fed the P2110 Powerharvester module, energy, it provides power for the WSN-Eval-01 wireless sensor
and then, the harvested dc output power was recorded. board, which sends the data to the access point. Then, when the
The P2110 Powerharvester module was used as an RF data reach the XLP 16-bit development board via the PICtail
energy-harvesting circuit, and it collected power of the modu- daughter card wirelessly, the XLP 16-bit development board
lated signal generated by NI PXIe-1082. This module enables calculates the packet number, time, and time difference (dT)
efficient energy harvesting between 902- and 928-MHz fre- between the consecutive packets.
quency band. Moreover, this module can also harvest energy The data obtained from the XLP 16-bit development board
by connecting an antenna to its SMA connector. However, that is connected to a computer with a cable can be dis-
the output of the signal generator is directly connected to played via HyperTerminal. Configuration of the HyperTermi-
the input of the RF energy-harvesting circuit with the coaxial nal should be as baud rate: 19 200, data bits: 8 bits, parity:
cable so that it results in minimal power loss. This circuit none stop bits: 1 bit, and flow control: hardware. For each
can harvest power down to −11.5-dBm input power level. modulated signal, 100 packets per each power level (dBm)
On the other hand, a WSN-Eval-01 wireless sensor board from 1 to 10 dBm were recorded on the computer via the
was plugged into the P2110 Powerharvester module, as shown HyperTerminal. Data containing temperature, light, humidity,
in Fig. 1. The WSN-Eval-01 wireless sensor board can sense RSSI, node, and TX ID was obtained from the wireless sensor
temperature, light, and humidity. It also has an interface for board (WSN-Eval-01). In addition, data containing packet
external sensors. The wireless sensor board sends the data, number, time, and time difference was taken from the access
such as temperature, light, humidity, received signal strength point (PICtail daughter card).
indicator (RSSI), node, and transmitter (TX) ID to the access
point which is placed on the XLP 16-bit development board. IV. R ESULTS AND D ISCUSSION
The XLP 16-bit development board includes Microchip’s When compared to each other, the RF input power signals
PIC24F microcontroller unit (MCU) (left circuit on the table with different modulation techniques have different ampli-
as shown in Fig. 1). The XLP 16-bit development board can tudes, and hence, they have different power values in the
receive data up to eight-node IDs wirelessly at the same time time domain. The characteristic of the modulation technique
and maintain a time counter for each node ID. Moreover, directly affects the charging time. Therefore, the charging time
the PICtail daughter card was plugged into the XLP 16-bit varies depending on the modulation technique, not random
development board. The PICtail daughter card as an access instantaneous power levels. For this reason, the charging time
point has 2.4 GHz, IEEE 802.15.4 radio module, and it values of the P2110 Powerharvester module may vary accord-
receives the data from the WSN-Eval-01 wireless sensor board. ing to each modulation technique. The P2110 Powerharvester
Apart from the charging time, the received power values module stores the harvested energy in a capacitor. When the
of the modulated signals were also measured and recorded. P2110 Powerharvester module reaches the sufficient energy
The channel bandwidth for each modulation technique was threshold during the energy harvesting, it will send the data
specified as 4 MHz. The total received power was measured to the XLP 16-bit development board that calculates the time
by the RF signal analyzer soft front panel from National and time difference.

Authorized licensed use limited to: Gebze Technical University. Downloaded on May 02,2023 at 08:29:20 UTC from IEEE Xplore. Restrictions apply.
6908 IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, VOL. 69, NO. 9, SEPTEMBER 2020

Fig. 2. Charging time for ASK and PSK modulation techniques: M = 4, 16, and 32. (a) 4-, 16-, and 32-ASK. (b) 4-, 16-, and 32-PSK.

as the RF input power level of the P2110 Powerharvester

module increased, the charging time decreased. In other words,
the more the RF input power level, the shorter the charging
time. Besides, no measurement sample could be obtained
for the FSK modulation technique at 1 dBm during 3 h.
It also makes this measurement impossible when considering
100 measurement samples at 1 dBm, and hence, it was shown
as not available (NA) in Table I. For this reason, measurement
samples were taken at 2 dBm for 4-, 16-, and 32-FSK
modulation techniques.
Compared with the other modulation techniques in Table I,
it is found that the ASK modulation technique has the
shortest charging time. In the ASK modulation technique,
the message signals directly affect the amplitude and power
of the carrier signals. Therefore, the power of the carrier
signal changes the charging time of the RF energy-harvesting
Fig. 3. Charging time for 32-ASK, 32-QAM, 32-PSK, and 32-FSK.
circuit. Besides, when the charging time values of the ASK
modulation technique in terms of M values are investigated,
The time difference between the consecutive packets was it is observed that charging time reduces when the M value
evaluated as charging time. For each RF input power level of the ASK modulation technique decreases. In other words,
(dBm), 100 packets were taken as measurement samples for as the M value decreases in the ASK modulation technique,
a more precise calculation. Then, the charging time for each the charging time becomes shorter. In the QAM modulation
modulation technique was calculated as the average of the technique, the message signals impact the amplitude and
time differences of the consecutive 100 packets. Moreover, phase of the carrier signals. The charging time values of the
100 measurement samples were taken for each RF input power 16-QAM and the 32-QAM are almost the same, and
level and 1000 measurement samples were evaluated per each the charging time values of the 4-QAM are shorter
M value (4, 16, and 32). Therefore, 3000 measurement sam- than them. Therefore, the effect of amplitude is observed
ples were taken for each modulation technique, and in total, for the M = 4 value. On the other hand, in the PSK and FSK
12 000 measurement samples were analyzed in this article. modulation techniques, the message signals affect the phase
As shown in Table I, the unit of transmit power level is dBm and frequency of the carrier signals, respectively. The charg-
and the unit of the charging time is second. ing time values for 4-, 16-, and 32-PSK are approximately
According to the transmit power, the charging time values the same. Similarly, the charging time values for 4-, 16-,
for each modulation and M value were calculated, as shown and 32-FSK are almost the same. In addition, in the PSK and
in Table I. Each modulation technique has the longest charging FSK modulation techniques, it is observed that the charging
time at the lowest transmit power level. On the contrary, time decreases when the transmit power increases. However,
it has the shortest charging time at the highest transmit power it should be stated that M values do not affect the charging
level. The shortest charging time was evaluated as 1.03 s time for the PSK and FSK modulation techniques. The effects
at 10 dBm for the 4-ASK modulation technique, whereas the of M values on charging time are clearly seen in Fig. 2(a)
longest charging time was calculated as 148.88 s at 2 dBm for the ASK modulation technique. On the contrary, Fig. 2(b)
for 32-FSK modulation technique. It was determined that shows that M values have no effects on charging time for the

Authorized licensed use limited to: Gebze Technical University. Downloaded on May 02,2023 at 08:29:20 UTC from IEEE Xplore. Restrictions apply.
CANSIZ et al.: EFFECTS OF DIFFERENT MODULATION TECHNIQUES ON CHARGING TIME IN RF ENERGY-HARVESTING SYSTEM 6909

Fig. 4. Power spectrum and received power for four modulation techniques at 2-dBm transmit power. (a) 32-ASK. (b) 32-QAM. (c) 32-PSK. (d) 32-FSK.

PSK modulation technique. It was determined that M values TABLE II

had the most effects on the ASK modulation technique and, R ECEIVED P OWER M EASUREMENTS
then, on the QAM modulation technique, but they had no
effects on PSK and FSK modulation techniques.
When the charging time values of four modulation
techniques are considered, the biggest difference in their
charging time values is seen in 32-ASK, 32-QAM,
32-PSK, and 32-FSK. Fig. 3 shows both the measured values
by means of the measurement system and the theoretical val-
ues calculated using (16). As presented in Section II, the charg-
ing time can be calculated based on the parameters of an
RF energy-harvesting system. The parameters are the transmit
power of symbols, the transmit probability of symbols, the RF
energy transmitting efficiency, the channel gain, the path
loss, the power sensitivity level, the RF energy-harvesting
efficiency, the charging voltage, the capacity of the capacitor,
and the discharge depth of capacitor. These parameters of the
measurement system for different modulation techniques are
given in (16), and they determine the system performance.
The differences between the modulation techniques are clearly levels of the carrier signals in the channel. The received power
displayed in the figure in terms of charging time. The main rea- of each modulation technique influences the charging time of
son for these differences in charging time values is the power the RF energy-harvesting circuit. According to the transmit

Authorized licensed use limited to: Gebze Technical University. Downloaded on May 02,2023 at 08:29:20 UTC from IEEE Xplore. Restrictions apply.
6910 IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, VOL. 69, NO. 9, SEPTEMBER 2020

Fig. 5. Received power distribution for (a) 32-ASK and (b) 32-QAM for 2-dBm transmit power.

power, the received power values for CW, 32-ASK, 32-QAM, V. C ONCLUSION
32-PSK, and 32-FSK are shown in Table II. The difference
We investigated the effects of different modulation tech-
between the transmit power and the received power for each
niques on the charging time of an RF energy-harvesting
modulation technique is called a loss. As shown in Table II,
system. A measurement system was set up to measure the
the minimum loss between the transmit power and the received
charging time for ASK, QAM, PSK, and FSK modulation
power is observed in CW modulation, and then, the losses
techniques. The measurements showed that the type of modu-
vary according to the modulation type. The highest received
lation techniques and their M values affected the charging time
power from 1 to 6 dBm is seen in the 32-ASK. Moreover, the
values in different amounts. When four modulation techniques
32-PSK has the highest received power at 7-dBm transmit
were compared to each other, the ASK modulation technique
power level. Furthermore, the 32-FSK has the highest received
performed better than others for all M values (4, 16, and 32).
power from 8 to 10 dBm. The biggest difference among the
The value of M had the greatest effect on the ASK modulation
charging time values is seen at 2-dBm transmit power level.
technique and then on the QAM modulation technique, but
At 2-dBm transmit power level, the received power values
the value of M did not affect the PSK and FSK modulation
of the 32-ASK, 32-QAM, 32-PSK, and 32-FSK are shown
techniques in terms of the charging time. As a result, the
in Fig. 4. The received bandwidth for each modulation tech-
4-ASK modulation technique was evaluated to be the most
nique is 4 MHz, and the center frequency of the carrier signal
suitable modulation technique among these modulations for
is 915 MHz. As shown in Fig. 4, the received power values
the charging time of the RF energy-harvesting system.
of the 32-ASK, 32-QAM, 32-PSK, and 32-FSK are 1.30,
1.10, 0.92, and 0.63 dBm, respectively. In addition to that, at
2-dBm transmit power level, the charging time values of R EFERENCES
the 32-ASK, 32-QAM, 32-PSK, and 32-FSK are 14.49,
[1] P. Jaffe and J. McSpadden, “Energy conversion and transmission mod-
20.40, 33.92, and 148.88, respectively. In the light of ules for space solar power,” Proc. IEEE, vol. 101, no. 6, pp. 1424–1437,
these data, it is seen that there is a direct relationship Jun. 2013.
between the charging time and the received power. Moreover, [2] A. R. M. Siddique, R. Rabari, S. Mahmud, and B. V. Heyst, “Thermal
energy harvesting from the human body using flexible thermoelectric
the received power distributions are also analyzed accord- generator (FTEG) fabricated by a dispenser printing technique,” Energy,
ing to the modulation type. When the received power dis- vol. 115, pp. 1081–1091, Nov. 2016.
tributions of the 32-ASK, 32-QAM, 32-PSK, and 32-FSK [3] A. C. Turkmen and C. Celik, “Energy harvesting with the piezoelectric
material integrated shoe,” Energy, vol. 150, pp. 556–564, May 2018.
were examined over 1000 samples, the largest variation was [4] H. Sun, Y.-x. Guo, M. He, and Z. Zhong, “Design of a high-efficiency
observed in the 32-ASK and less variation was seen in 2.45-GHz rectenna for low-input-power energy harvesting,” IEEE Anten-
the 32-QAM, as shown in Fig. 5(a) and (b), respectively. nas Wireless Propag. Lett., vol. 11, pp. 929–932, 2012.
[5] A. Collado and A. Georgiadis, “Conformal hybrid solar and electro-
On the other hand, the received power distributions of the magnetic (EM) energy harvesting rectenna,” IEEE Trans. Circuits Syst.
32-PSK and 32-FSK remained almost unchanged. As an I, Reg. Papers, vol. 60, no. 8, pp. 2225–2234, Aug. 2013.
RF energy harvester has different efficiency values at dif- [6] (2018). Powercast Corporation. [Online]. Available: http://www.
powercastco.com
ferent power levels, the received power distributions also [7] K. Guha, “RF energy harvesting in agriculture,” in Proc. 8th All India
have a direct effect on charging time. The results of the Peoples’ Technol. Congr., 2011, pp. 1–5.
received power distributions of the four different modula- [8] J. Bito, S. Kim, M. Tentzeris, and S. Nikolaou, “Ambient energy harvest-
ing from 2-way talk-radio signals for ‘smart’ meter and display appli-
tion techniques also support the effects of M values on cations,” in Proc. IEEE Antennas Propag. Soc. Int. Symp. (APSURSI),
charging time. Jul. 2014, pp. 1353–1354.

Authorized licensed use limited to: Gebze Technical University. Downloaded on May 02,2023 at 08:29:20 UTC from IEEE Xplore. Restrictions apply.
CANSIZ et al.: EFFECTS OF DIFFERENT MODULATION TECHNIQUES ON CHARGING TIME IN RF ENERGY-HARVESTING SYSTEM 6911

[9] M. Pinuela, P. D. Mitcheson, and S. Lucyszyn, “Ambient RF energy [32] A. Boaventura, D. Belo, R. Fernandes, A. Collado, A. Georgiadis,
harvesting in urban and semi-urban environments,” IEEE Trans. Microw. and N. B. Carvalho, “Boosting the efficiency: Unconventional waveform
Theory Techn., vol. 61, no. 7, pp. 2715–2726, Jul. 2013. design for efficient wireless power transfer,” IEEE Microw. Mag., vol. 16,
[10] M. Cansiz, T. Abbasov, M. B. Kurt, and A. R. Celik, “Mobile mea- no. 3, pp. 87–96, Apr. 2015.
surement of radiofrequency electromagnetic field exposure level and [33] D. De Donno, L. Catarinucci, and L. Tarricone, “RAMSES: RFID
statistical analysis,” Measurement, vol. 86, pp. 159–164, May 2016. augmented module for smart environmental sensing,” IEEE Trans.
[11] Y. Shi, J. Jing, Y. Fan, L. Yang, Y. Li, and M. Wang, “A novel Instrum. Meas., vol. 63, no. 7, pp. 1701–1708, Jul. 2014.
compact broadband rectenna for ambient RF energy harvesting,” AEU- [34] D. Altinel and G. K. Kurt, “Statistical models for battery recharging
Int. J. Electron. Commun., vol. 95, pp. 264–270, Oct. 2018. time in RF energy harvesting systems,” in Proc. IEEE Wireless Commun.
[12] S. Agrawal, R. D. Gupta, M. S. Parihar, and P. N. Kondekar, Netw. Conf. (WCNC), Apr. 2014, pp. 636–641.
“A wideband high gain dielectric resonator antenna for RF energy har- [35] D. Mishra, S. De, and K. R. Chowdhury, “Charging time characterization
vesting application,” AEU-Int. J. Electron. Commun., vol. 78, pp. 24–31, for wireless RF energy transfer,” IEEE Trans. Circuits Syst. II, Exp.
Aug. 2017. Briefs, vol. 62, no. 4, pp. 362–366, Apr. 2015.
[13] A. P. Sample, D. J. Yeager, P. S. Powledge, A. V. Mamishev, and [36] D. Altinel and G. Karabulut Kurt, “Energy harvesting from multiple RF
J. R. Smith, “Design of an RFID-based battery-free programmable sources in wireless fading channels,” IEEE Trans. Veh. Technol., vol. 65,
sensing platform,” IEEE Trans. Instrum. Meas., vol. 57, no. 11, no. 11, pp. 8854–8864, Nov. 2016.
pp. 2608–2615, Nov. 2008. [37] H. Chen, Y. Li, J. Luiz Rebelatto, B. F. Uchoa-Filho, and B. Vucetic,
[14] M. Cansiz, D. Altinel, and G. K. Kurt, “Efficiency in RF energy harvest- “Harvest-then-cooperate: Wireless-powered cooperative communica-
ing systems: A comprehensive review,” Energy, vol. 174, pp. 292–309, tions,” IEEE Trans. Signal Process., vol. 63, no. 7, pp. 1700–1711,
May 2019. Apr. 2015.
[15] A. Collado and A. Georgiadis, “Improving wireless power transmission [38] National Instruments. Accessed: Oct. 3, 2019. [Online]. Available:
efficiency using chaotic waveforms,” in IEEE MTT-S Int. Microw. Symp. http://www.ni.com
Dig., Jun. 2012, pp. 1–3.
[16] A. Collado and A. Georgiadis, “Optimal waveforms for efficient wireless
power transmission,” IEEE Microw. Wireless Compon. Lett., vol. 24,
no. 5, pp. 354–356, May 2014.
[17] A. Georgiadis, A. Collado, and K. Niotaki, “Rectenna design and signal
Mustafa Cansiz received the B.S. degree in elec-
optimization for electromagnetic energy harvesting and wireless power
trical and electronics engineering from Karadeniz
transfer,” IEICE Trans. Electron., vol. 98, no. 7, pp. 608–612, 2015.
Technical University, Trabzon, Turkey, in 2002,
[18] A. Boaventura, N. B. Carvalho, and A. Georgiadis, “The impact of multi- the M.Sc. degree in electrical and electronics engi-
sine tone separation on RF-DC efficiency,” in Proc. IEEE Asia–Pacific neering from Dicle University, Diyarbakır, Turkey,
Microw. Conf., Nov. 2014, pp. 606–609. in 2010, and the Ph.D. degree in electrical and elec-
[19] A. S. Boaventura and N. B. Carvalho, “Maximizing DC power in tronics engineering from Inonu University, Malatya,
energy harvesting circuits using multisine excitation,” in IEEE MTT-S Turkey, in 2016.
Int. Microw. Symp. Dig., Jun. 2011, pp. 1–4. From 2005 to 2011, he worked in the electron-
[20] M. H. Ouda, P. Mitcheson, and B. Clerckx, “Optimal operation of mul- ics and telecommunication industry. Since 2011, he
titone waveforms in low RF-power receivers,” in Proc. IEEE Wireless has been with Dicle University. His research areas
Power Transf. Conf. (WPTC), Jun. 2018, pp. 1–4. include the measurement of electromagnetic exposure and radio-frequency
[21] C. R. Valenta and G. D. Durgin, “Rectenna performance under power- energy-harvesting systems.
optimized waveform excitation,” in Proc. IEEE Int. Conf. (RFID),
Apr. 2013, pp. 237–244.
[22] C. R. Valenta, M. M. Morys, and G. D. Durgin, “Theoretical
energy-conversion efficiency for energy-harvesting circuits under power-
optimized waveform excitation,” IEEE Trans. Microw. Theory Techn.,
vol. 63, no. 5, pp. 1758–1767, May 2015. Dogay Altinel received the B.S. degree in electron-
[23] M. S. Trotter, J. D. Griffin, and G. D. Durgin, “Power-optimized ics engineering from Hacettepe University, Ankara,
waveforms for improving the range and reliability of RFID systems,” in Turkey, in 1992, and the M.Sc. and Ph.D. degrees
Proc. IEEE Int. Conf. (RFID), Apr. 2009, pp. 80–87. in telecommunications engineering from Istanbul
[24] J. F. Ensworth, S. J. Thomas, S. Y. Shin, and M. S. Reynolds, Technical University, Istanbul, Turkey, in 2014 and
“Waveform-aware ambient RF energy harvesting,” in Proc. IEEE Int. 2019, respectively.
Conf. (IEEE RFID), Apr. 2014, pp. 67–73. He worked in the electronics and telecommunica-
[25] G. Andia Vera, A. Georgiadis, A. Collado, and S. Via, “Design of a tion industry for many years from 1992 to 2012.
2.45 GHz rectenna for electromagnetic (EM) energy scavenging,” in He has been working at Istanbul Medeniyet Uni-
Proc. IEEE Radio Wireless Symp. (RWS), Jan. 2010, pp. 61–64. versity, Istanbul, since 2012. His research interests
[26] G. Fukuda, S. Yoshida, Y. Kai, N. Hasegawa, and S. Kawasaki, “Eval- include the areas of communication theory, wireless
uation on use of modulated signal for microwave power transmission,” networks, and energy harvesting systems.
in Proc. 44th Eur. Microw. Conf., Oct. 2014, pp. 425–428.
[27] H. Sakaki, K. Nishikawa, S. Yoshida, and S. Kawasaki, “Modulated
scheme and input power impact on rectifier RF-DC efficiency for
WiCoPT system,” in Proc. Eur. Microw. Conf. (EuMC), Sep. 2015,
pp. 60–63.
[28] G. Andia Vera, D. Allane, A. Georgiadis, A. Collado, Y. Duroc, and Gunes Karabulut Kurt (Senior Member, IEEE)
S. Tedjini, “Cooperative integration of harvesting RF sections for passive received the B.S. degree (Hons.) in electronics
RFID communication,” IEEE Trans. Microw. Theory Techn., vol. 63, and electrical engineering from Bogazici University,
no. 12, pp. 4556–4566, Dec. 2015. Istanbul, Turkey, in 2000, and the M.A.Sc. and Ph.D.
[29] J. Blanco, F. Bolos, A. Collado, and A. Georgiadis, “RF-energy har- degrees in electrical engineering from the University
vesting and wireless power transfer efficiency from digitally modulated of Ottawa, Ottawa, ON, Canada, in 2002 and 2006,
signals,” in Proc. IEEE 15th Medit. Microw. Symp. (MMS), Nov. 2015, respectively.
pp. 1–4. From 2005 to 2006, she was with TenXc Wireless,
[30] F. Bolos, J. Blanco, A. Collado, and A. Georgiadis, “RF energy Ottawa. From 2006 to 2008, she was with Edgewater
harvesting from multi-tone and digitally modulated signals,” IEEE Trans. Computer Systems Inc., Ottawa. From 2008 to 2010,
Microw. Theory Techn., vol. 64, no. 6, pp. 1918–1927, Jun. 2016. she was with Turkcell Research and Development
[31] M. Rajabi, N. Pan, S. Claessens, S. Pollin, and D. Schreurs, “Modulation Applied Research and Technology, Istanbul. Since 2010, she has been with
techniques for simultaneous wireless information and power transfer Istanbul Technical University, Istanbul, where she is currently a Professor.
with an integrated rectifier–receiver,” IEEE Trans. Microw. Theory She is also with Carleton University, Ottawa, as a Visiting Professor and also
Techn., vol. 66, no. 5, pp. 2373–2385, May 2018. appointed as an Adjunct Research Professor.

Authorized licensed use limited to: Gebze Technical University. Downloaded on May 02,2023 at 08:29:20 UTC from IEEE Xplore. Restrictions apply.