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

Practical Waveform-to-Energy Harvesting Model

and Transmit Waveform Optimization for RF
Wireless Power Transfer Systems
Nachiket Ayir , Member, IEEE, Taneli Riihonen , Senior Member, IEEE, and Mikko Heino , Member, IEEE

Abstract— The received radio-frequency (RF) power in far- sensors from becoming hazardous environmental waste after
field RF wireless power transfer (WPT)—with or without their batteries die and also avoiding the hassle of replacing the
simultaneous information transfer—is minuscule due to large batteries given the sheer size of an IoT network. An RF WPT-
propagation loss in wireless media. In such scenarios, adapting to
the receiver characteristics by transmit waveform optimization enabled sensor would comprise a rectenna (a portmanteau
is essential for maximizing the harvested direct current (dc) for the combination of a receiver antenna and an RF diode-
and, thus, the end-to-end efficiency of an RF WPT system. The based rectifier) to convert the incident RF energy to direct
receiver efficiency in RF WPT is governed by the RF-to-dc current (dc) energy, and possibly store it temporarily in its
efficiency of the rectifier as well as the impedance mismatch supercapacitor. While an RF WPT transmitter usually operates
at the antenna and load. In this article, we study the receiver
efficiency for any fixed load and, subsequently, present a novel in a single band, the sensors should practically be able to
rectifier model that relates the average harvested dc power to harvest energy from multiple such RF bands to maximize their
the distribution, that is, the histogram, of the instantaneous dc output.
power levels of the RF signal’s envelope over time. The proposed While the focus of RF WPT is to energize the receiver
waveform-to-energy harvesting (EH) model enables us to antici- sensors through RF radiation, the spectrum could be leveraged
pate the average harvested dc power for any waveform, including
communication signals as well, given the knowledge of the power- better by utilizing it to transport information as well as power
level distribution. Consequently, we conduct rigorous waveform together, thus giving rise to the concept of RF SWIPT. The
optimization to maximize the average harvested dc power and energy and information receivers could either be colocated or
determine the digital baseband signal at the transmitter that does separate depending on the application and device complexity.
so, namely prove that a pulsed tone at appropriate frequency RF SWIPT can also leverage the networking paradigms of
is optimal for RF WPT. We present a multiband test-bed for
determining the receiver efficiency for any digital baseband wireless-powered communication and backscatter networking,
waveform. The efficacy of the proposed model is corroborated where energy is transferred in the downlink and the transceiver
through experiments as well as simulations, which confirm that it spends it to transmit information in the uplink [1].
is operational as well as accurate in practice and that single-sine
pulses yield higher efficiency than basic multisine waveforms, A. Motivation
while a pulsed phase shift keying (PSK) is preferable for
simultaneous wireless information and power transfer (SWIPT). The state-of-the-art WPT technologies enable energy trans-
fer over short-range (about an inch), mid-range (about 1–2 m),
and far-field. All these competing technologies have their own
Index Terms— Multiband test-bed, rectifier model, simultane-
ous wireless information and power transfer (SWIPT), waveform benefits and limitations [3], [4], [5], [6]. Moreover, each of the
optimization, wireless power transfer (WPT). three technologies has commercially available products. While
far-field RF WPT that is in the scope of this work provides
I. I NTRODUCTION the benefits of mobility and longer operation range over the
other two, it comes with additional complexities too.

F AR-FIELD radio-frequency (RF) wireless power transfer

(WPT) and simultaneous wireless information and power
transfer (SWIPT) are eminent research topics [1], [2] with
Especially, RF WPT suffers from immense over-the-air
propagation losses resulting in only a small fraction of the
actual transmitted power reaching the receiver. This neces-
potential applications in consumer and industrial Internet-of- sitates optimizing receiver performance in RF WPT systems
Things (IoT). RF WPT alone could provide the multitude of to reduce the disadvantage. Consequently, a majority of the
IoT sensors with a replenishable energy source, thus averting research on RF WPT systems has thus far focused on tech-
Manuscript received 24 January 2023; revised 6 April 2023; accepted nical problems associated with the energy harvesting (EH)
16 May 2023. This work was supported in part by the Finnish Cultural receiver [7]. The performance of the receiver in RF WPT
Foundation Grant and in part by the Academy of Finland under Grant is characterized by its efficiency, which is the ratio of the
341489/346622. (Corresponding author: Nachiket Ayir.) dc
The authors are with the Faculty of Information Technology and Com- average harvested dc power (Pout ) to the average RF input
RF
munication Sciences, Tampere University, 33720 Tampere, Finland (e-mail: power (Pin ).
nachiket.ayir@tuni.fi; taneli.riihonen@tuni.fi; mikko.heino@tuni.fi). The receiver in RF WPT generally comprises a receiv-
Color versions of one or more figures in this article are available at
https://doi.org/10.1109/TMTT.2023.3284261. ing antenna, a matching network, a rectifier, and a load.
Digital Object Identifier 10.1109/TMTT.2023.3284261 To optimize the receiver performance in RF WPT, it is
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 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

first essential to have an accurate mathematical model of an to the average harvested dc power [8], [27], [28], [29] and
RF energy harvester. It is already well established in the rely on curve-fitting based on the measurement data. While
existing literature [7] that linear EH models cannot capture Boshkovska et al. [27] present a saturation model and utilizes
the true behavior of a diode-based rectifier, while nonlinear a sigmoid function for fitting, Chen et al. [8] present the
EH models can characterize the rectifier behavior in different receiver efficiency as a heuristic expression that is a function
power regimes [8]. In this article, we present a novel nonlinear of PinRF . Similarly, two nonlinear EH models are introduced
rectifier model that relies on the characteristics of the input in [28] where, in both cases, the output power is a polynomial
waveform. We review the existing nonlinear EH models in fraction of PinRF , whereas Xu et al. [29] present the output
Section I-B for reference. power as a simple polynomial of PinRF . Such models are very
useful for determining power allocation at the transmitter [30],
[31], but not at all for waveform optimization.
B. Literature Review To be able to design a waveform suitable for RF WPT
dc
While the practical applications of RF WPT are slowly or RF SWIPT, the rectifier models need to characterize Pout
on the rise, there is already a significant amount of aca- in terms of the parameters of the transmit signal. In [32]
demic research material available on RF WPT, RF SWIPT, and [33], the diode in the rectifier is modeled by approx-
and RF EH. These include waveform analysis, end-to-end imating the Shockley equation for a single diode, whereas
system performance analysis, rectifier modeling and design, Abeywickrama et al. [34] present a nonlinear current–voltage
waveform optimization, beamforming algorithms, and link- model of the rectifier’s diode. These models, however, do not
level analysis [2], [7], [9], [10], [11], [12], [13]. In some cases, incorporate the input impedance mismatch at the receiver,
the presented results vary depending on the part of the RF which varies with input power [13]. These models are specifi-
WPT system under consideration. For instance, experimen- cally utilized for determining the optimal amplitudes [34] and
dc
tal results in [14] reveal that multisine signals yield higher phases [32] for multisine waveforms for maximizing Pout in
receiver efficiency than orthogonal frequency-division multi- RF WPT.
plexing (OFDM) signals, chaotic signals, harmonic signals, Furthermore, based on the mentioned power-to-power and
and so on. However, while considering the nonlinearities at waveform-to-power rectifier models, the problem of waveform
the transmitter, the experiments in [9] and [15] suggest that design with complete [33], [35], partial [36], [37], [38], [39],
high peak-to-average power ratio (PAPR) multisine waveforms and no [40] knowledge of the channel state information (CSI)
fare poorly in terms of the end-to-end efficiency of the system. is available in the literature. In the case of a complete absence
However, as mentioned above, a majority of the research of CSI, the transmitter could resort to transmit diversity [40].
work caters to the receiver side of RF WPT, which includes Additionally, the problem of communication and signal design
rectenna design, mathematical modeling, waveform design, for optimizing the end-to-end transmission in RF WPT is
and measurement setups. A comprehensive survey of the presented in [41].
various rectenna topologies employed in RF WPT and RF EH Moving on to measurement setups, a GNU Radio-based
is presented in [16]. The type of receiving antenna to be used prototype test-bed for evaluating the end-to-end (i.e., dc-to-dc)
varies based on the application, frequency band, and so on. performance of any digital waveform for RF WPT was
While miniature on-chip antennas are employed in biomedical showcased in [9]. Based on pilot-based channel estimation
applications [17], [18] of RF WPT, wideband or dual-band and subsequent waveform optimization, the first closed-loop
antennas are preferred for RF EH applications [19], [20], [21]. WPT prototype was presented in [22] and further improved
Furthermore, the physical design of a rectenna operating at in [42]. These implementations introduce CSI feedback to
2.4 GHz, by designing Schottky diode-based voltage rectifiers, the transmitter, which is utilized to optimize the amplitude
is presented in [22] and [23]. Similarly, Duy et al. [24] present of multisine tones. However, the CSI feedback is attained
the rectenna designed specifically for the 5.8-GHz band. over a wired medium. The literature has several other RF
The receiver efficiency in RF WPT is dependent on sev- WPT test-beds for tasks such as charging a single super-
eral parameters. In practice, a designed rectenna would have capacitor with multiple ambient sources [43], determining
peak receiver efficiency at a certain frequency and lesser the digital modulation with the lowest charging time [44],
at other RF inputs. It is observed in [16] that antennas devising charging protocols to regulate multiple power sources
that operate at the lower end of the spectrum have higher and multiple energy harvesters [45], determining the opti-
RF-to-dc efficiency. Another important parameter that affects mal orientation between transmit and receive antennas [46],
the design of rectennas and the resultant receiver efficiency studying antenna directivity as a function of source–sensor
is the impedance mismatch between the receiving antenna and distance [47]. Furthermore, blind adaptive beamforming tech-
the rectifier network [13], [25], which again varies with the niques to direct the RF energy toward the harvester were
input signal power and frequency. The combined impact of studied and implemented in a multiple-input single-output
variation in input power, PAPR, and load resistance on the WPT test-bed in [48]. The beamformer weights are updated
receiver efficiency in RF WPT is presented in [26]. based on the feedback from the energy harvester to enhance
Besides physical design, there are significant theoreti- the dc output. With a different approach, the concept in [49]
cal contributions to the mathematical modeling of energy employs distributed beamforming with multiple power sources
harvesters. A majority of these are parametric power-to-power charging a single receiver. In this case, beamforming aligns the
models that relate the average input power of the RF waveform phases of RF waves from various transmitters at the receiver,
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

AYIR et al.: PRACTICAL WAVEFORM-TO-EH MODEL AND TRANSMIT WAVEFORM OPTIMIZATION 3

to obtain a multisine waveform with high PAPR at the receiver 2) The proposed rectifier model accurately predicts the
input. average harvested dc power for any waveform with
Coming to SWIPT test beds, an software-defined radio a known power-level distribution function, that is,
(SDR)-based system was presented in [50] to study histogram in other words. The model evaluates
rate–energy tradeoffs in SWIPT systems. Separate signals the instantaneous rectification efficiency based on
were employed for WPT and information transfer, and they continuous-wave (CW) data and utilizes it in conjunction
were superimposed for transmission based on time-switching with the distribution function to determine the average
and power-splitting schemes. The performance of multisine harvested dc power for a particular waveform. We exam-
and modulated OFDM signals for SWIPT was evaluated. ine the receiver efficiency of different RF WPT and
Another SWIPT test-bed to study the performance of multisine RF SWIPT waveforms using the proposed model to
frequency-shift keying (FSK) signals was showcased in [51]. demonstrate its applicability.
The information is encoded in the frequency spacing of the 3) We use the proposed rectifier model to perform gen-
sinusoids and thus can be decoded at the receiver without eral transmit-side waveform optimization to maximize
the need for a local oscillator. This encoding method allows the average harvested dc power at the receiver while
the use of high-PAPR signals for achieving high receiver effi- accounting for a transmit power constraint, available
ciency and high throughput. Both the aforementioned SWIPT channel bandwidth, and frequency-selective channel
test-beds utilize an ideal vector signal transceiver (VST) as a gains.
transmitter and information decoder, with the former utilizing 4) We present a multiband test-bed to evaluate the receiver
an external power amplifier (PA) as well. A brief overview of efficiency of any digital waveform. Using this test-
more such test-beds is available in [52]. bed, we experimentally verify the predictions of the
proposed rectifier model for various waveforms, includ-
C. Contributions and Organization ing information-carrying phase shift keying (PSK) and
The nonlinear waveform-to-power models [32], [33], [34], quadrature amplitude modulation (QAM) signals as well
dc
[35], [36] relate Pout to the incoming RF waveform ŷ(t) and as the optimized waveforms for a large number of
are thus suitable for waveform optimization. While Boaventura different load resistors, varying average input RF power,
and Carvalho [32] utilize their model only for analyzing the six frequency bands, and three harvester circuits.
RF WPT performance with different multisine waveforms, the The remainder of this article is organized as follows.
waveform optimization in [33], [34], [35], and [36] is limited In Section II, we present the system model employed in this
to the design of amplitudes and phases. The optimization work. Afterward, we present the novel waveform-to-power
thereof does not involve the design of the frequencies of the rectifier model in Section III and, using it, design the optimal
baseband signal, but instead assumes the baseband signal to be waveform for RF WPT in Section IV. The experimental
a multisine with evenly spaced frequencies. Consequently, the and simulation results that verify the precision of our recti-
state-of-the-art designs are incapable of yielding an arbitrary fier model and the subsequent optimization are presented in
waveform that corresponds to the global optimum as they are Section V. Finally, we present the conclusion of this work in
restricted to a multisine signal with a not-so-large number of Section VI.
tones. Naturally, the bandwidth constraint is also neglected.
Additionally, while [32], [33], [34], [35], and [36] are seminal II. S YSTEM M ODEL
works, models thereof are devised with a few simplifications In general, an RF WPT/SWIPT system comprises a digital
that may not necessarily be very practical. First, these models transceiver, an external PA, and an RF EH receiver (with or
assume perfect input impedance matching at the receiver as without information transceiver capability). The system also
they do not include a matching network between the receiving incorporates a CSI feedback mechanism so that the wave-
antenna and the rectifier circuit, which is usually present in form at the transmitter could be optimized over a frequency-
commercial energy harvesters [53]. Second, the models are selective channel. The digital transceiver creates x(t) that is
derived from a single-diode rectifier circuit. transmitted as x̂(t) at center frequency f c , propagated over
Overall, there is a need for a general waveform-to-power a wireless channel ĥ(t), and received as ŷ(t). The modeling
rectifier model and nonsimplified waveform optimization. applies to both WPT and SWIPT, but this article is presented in
In this article, we overcome the aforementioned research gaps the context of EH from ŷ(t) despite it may contain information
through the following novel contributions. too.
1) We present a novel rectifier model that expresses the Without loss of generality, we assume that the digital
average harvested dc power in terms of the power baseband signal to be transmitted is a complex-valued N -tone
distribution of the envelope signal of the incident RF multisine waveform given by
waveform. The proposed parametric model is inde-
pendent of the underlying matching network, rectifier x(t) = x I (t) + j x Q (t)
circuit, as well as load impedance architecture and thus XN
 
is applicable for any EH receiver circuit as soon as the = An exp j2π f n t + jφn (1)
parameter fitting is performed based on simulation or n=1

measurement data. Thus, it takes inherently into account where An , f n , and φn represent the amplitude, the base-
the impedance mismatch at the receiver input and output. band frequency, and the phase of the nth tone, respectively.
 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

A multisine waveform can approximate (or, if N → ∞, uncorrelated. For a frequency-selective slow-fading channel,
exactly represent) an interval of any signal x(t) when inter- the average received RF power can be denoted as
preted as a Fourier series, which justifies the assumption. N
X
The corresponding RF signal at the transmitter can be PinRF = A2n |Hn |2 . (8)
expressed as n=1

√ XN After rectification, the input signal ŷ(t) generates an output

x̂(t) = 2 An cos(2π( f c + f n )t + φn ) (2) voltage vout (t) across the load resistor R L . The harvested dc
n=1 dc
voltage (Vout ) is the root mean square (rms) value of vout (t),
given as
and the averagetransmitted RF power is given as s
1 T 2
Z
N
T T dc
v (t)dt.
Z Z
1 1 X Vout = (9)
RF
Pout = x̂ 2 (t)dt = |x(t)|2 dt = A2n (3) T 0 out
T 0 T 0 n=1
Correspondingly, the instantaneous output power is given as
where T is the time period or interval duration of the modulat- ([vout
2
(t)]/R L ) and the harvested dc power is given as
ing signal. The result of (3) applies to multisines irrespective dc 2


Vout
of individual An , f n , or φn (so long as f n are distinct for dc
Pout = . (10)
all n) as a direct consequence of Parseval’s theorem. We let RL
RF Finally, the overall receiver efficiency is computed as
PTx denote the transmit power limit, such that Pout ≤ PTx .
The RF signal traverses a wireless medium with channel P dc
response ĥ(t) and arrives at the energy harvester as η = out
PinRF
dc 2
ŷ(t) = ĥ(t) ∗ x̂(t) + ŵ(t)


Vout
= PN . (11)
≈ ĥ(t) ∗ x̂(t) R L n=1 A2n |Hn |2
X N
Ân cos 2π( f c + f n )t + φ̂ n


= (4) III. N OVEL R ECTIFIER M ODEL
n=1
In this section, we propose a novel waveform-to-EH model
where the receiver noise ŵ(t) can be considered to be negligi- for an RF EH receiver.1 The proposed model relates the
ble from the EH perspective since the noise power level is way average harvested dc power to the characteristics of the
below the sensitivity threshold of RF energy harvesters [53], input RF waveform, and not to just its average RF input
[55]. The receiver noise is similarly neglected in other such power like typical nonlinear models. The proposed model
research works on RF WPT, assiduously summarized in [7]. incorporates the impact of the matching network along with
The impact of the channel on the individual multisine the rectifier circuit, for a given load resistor. In compari-
tones is represented as Ân = An |Hn | and φ̂ n = φn + ̸ Hn . son with the existing nonlinear RF EH models, the novelty
The baseband signal for the received RF signal ŷ(t) can be of the proposed waveform-to-EH model is demonstrated in
described as Table I.
Let us first briefly explain the notion behind this rectifier
N
X model: Given a baseband-modulated RF signal, where f c is
Ân cos 2π f n t + φ̂ n


y(t) = (5) sufficiently higher than the highest | f n |, the individual cycles
n=1
(or a few of them) can approximately be represented as
√ single-tone sinusoids with a constant amplitude. The constant
while the corresponding envelope signal is given as 2|y(t)|.
Next, the instantaneous power of this envelope signal can be amplitude here is the magnitude of the modulating signal at
expressed as that time instant. Each such single-tone sinusoid on traversing
through the RF EH circuit results in a certain rectified dc rms
p(t) = 2|y(t)|2 (6) voltage. The overall average harvested dc power is then given
by the average of these rectified dc rms powers over the time
while the average input RF power of the received signal, using period T of the input RF signal, which requires specifying only
(4) and (3), is given as the amplitude distribution of the received signal. The above
idea is supported by the measurement observations, which
PinRF = E ŷ 2 (t)

n reveal that the rectified signal vout (t) follows the envelope of

2 o the baseband signal y(t). Accordingly, we propose that the
= E ĥ(t) ∗ x̂(t)
instantaneous output power at each time instant is obtained by
= E ĥ 2 (t) · E x̂ 2 (t)
 
feeding the rectifier a single-tone input, whose input power is
= E |h(t)|2 · E |x(t)|2
 
(7) determined by the magnitude of y(t) at that time instant.
1 The RF EH receiver employed in this study comprises a diode-based
where x(t) is any complex baseband waveform and h(t) is the
rectifier that is only utilized for RF-to-dc conversion—the rectifier under
complex baseband channel equivalent of ĥ(t). The third step in consideration does not serve the purpose of demodulating communication
(7) arrives from the assumption that the signal and channel are signals such as quadrature phase-shift keying (QPSK), QAM, and OFDM.
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

AYIR et al.: PRACTICAL WAVEFORM-TO-EH MODEL AND TRANSMIT WAVEFORM OPTIMIZATION 5

TABLE I
C OMPARISON OF N ONLINEAR R ECTIFIER M ODELS FOR RF WPT

For the mathematical representation of the model, it is measurement data points and estimated parameters with a tilde,
convenient to consider time T to be a random variable, which for them to be discernible.
is uniformly distributed within one time period interval of the At each of the k input power points (k ∈ [1, M]), a single-
modulating signal. This is because the model is based on only tone input is transmitted, for which the envelope signal is
the statistical distribution of the input signal’s instantaneous in [k] and f P ( p) = δ( p − pk ).
basically constant, so, pk = P̃ RF
power while the precise temporal evolution over time is irrele- Thus, for each input power level P̃ RF dc
in [k], we measure P̃ out [k]
vant. Consequently, y(T ) and p(T ) become random variables, and obtain the efficiency using (12) simply as
which we denote simply as Y and P, respectively, for brevity. P̃ dc
out [k]
The instantaneous output power is given as η(P) · P, where η̃k = . (13)
P̃ RF
in [k]
η(P) is the efficiency of the receiver for instantaneous input
power P that captures the behavior of the entire receiver Using this set of discrete η̃k , a smooth curve for η̃( p) can be
circuit, including the antenna port, matching network, and the achieved by using piecewise polynomial interpolation as
rectifier circuit for a given load resistor R L . R
X
In summary, as experiments corroborate in Section V, the η̃( p) = ai,l ( p − pi )l (14)
overall harvested dc power can be represented as l=0
dc
Pout = E{η(P) · P} for pi ≤ p ≤ pi+1 , where i ∈ [1, M − 1], and R is the
Z ∞ degree of the piecewise polynomial. To avoid discontinuities
= η( p) · p · f P ( p)dp (12) in η̃( p) at the internal points pi (i ∈ [2, M − 1]), we can opt
0 for higher-order spline interpolation such as the widely used
where f P ( p) denotes the probability density function of the cubic spline. The coefficients ai,l in (14) are obtained from
instantaneous power of the envelope signal. Equation (12) the constraints on the polynomial in (14) per [56].
represents the proposed rectifier model based on the power If (14) were to be represented with a cubic spline (R = 3),
distribution of the baseband/modulating signal’s envelope. The the coefficients ai,2 would be given as solution to the equations
model predicts the output dc power/voltage for any baseband
ai−1,2 ( pi − pi−1 ) + 2ai,2 ( pi+1 − pi−1 ) + ai+1,2 ( pi+1 − pi )
waveform y(t), given its amplitude/power distribution f P ( p),
and η( p) that contains the parameters of the model that η̃i+1 − η̃i η̃i − η̃i−1
=3 −3 . (15)
characterize the receiver. In Sections III-A and III-B, we will ( pi+1 − pi ) ( pi − pi−1 )
explain how to estimate them before using the model. With M input power points, we have M such equations which
would yield M unique ai,2 coefficients. Once the value of ai,2
A. Determination of η( p) is attained, the remaining coefficients can be computed as

The parametric model in (12) relies on having the knowl- ai,0 = η̃i
edge of the efficiency of the rectifier for each input power, η̃i+1 − η̃i ( pi+1 − pi )

ai,1 = − · 2ai,2 + ai+1,2
that is, the model parameters. To achieve this, one should ( pi+1 − pi ) 3
first determine the rectifier efficiency at M input power points ai+1,2 − ai,2
ai,3 = . (16)
either through measurements or simulations. Let us denote the 3( pi+1 − pi )
 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

Thus, by varying the input power levels of a single- The corresponding cumulative density function is given as
tone sinusoid at RF, recording the corresponding recti- FP ( p) =
fier efficiency, and then interpolating for the intermediate    
points, we attain a continuous curve for estimated η̃( p) 1 p
F cos−1
− 1 , when N = 2

T

2π 1 f PinRF


using (14). 



 s !

 1

3p 1
−1
FT cos −
B. Determination of f P ( p) 
 2π 1 f 4PinRF 2
dc

The other information needed for computing Pout for any 

 s !!
given baseband waveform using (12) is the probability density
 3 p 1
, when N = 3
 −1
− cos − −


4PinRF

function of the instantaneous power of the baseband wave-  2
form. It can be computed as follows considering a base- (21)
band waveform Y = y(T ), which is a function of random
variable T . where T ∼ ∪(0, (1/1 f )]. Thus, the exact probability density
In the case that it is not possible to express Y as an function is given as f P ( p) =
equation, its f P ( p) can be estimated numerically from its
1

normalized histogram, given that the number of samples of  , N =2
√

 q
T is very large and the number of histogram bins is also RF
2π p 2P − p



 in
huge. Now, if gi is the normalized histogram value for the

 

bin pi < p ≤ pi+1 (i ∈ [1, number of bins]), with midpoint




p̌i , then the discrete probability is given as gi δ( p− p̌i ), and the
 


 1  1
 4π √ p  sq
 
corresponding estimated probability density function is given
 q 
√ √
RF RF
/3

by  3P in − p Pin + p


X  
f˜P ( p) = gi δ p − p̌i .


(17) 



i  

 1 , N = 3

+

 s
This approach is useful when y(t) is available as

    
√ √
 q q
RF
Pin /3− p
RF
 
a complex-valued sample sequence from simulations or


 3Pin + p
measurements.
In the other case, when the equation Y = y(T ) is known (22)
and used, first the corresponding cumulative distribution func-
where 0 ≤ p ≤ N PinRF . Unfortunately, Y = y(T ) is not
tion using (6) is given as
invertible for multisines with higher N .
FP ( p) = Pr(P ≤ p)
 r r 
p p IV. WAVEFORM O PTIMIZATION
= Pr − ≤Y≤
2 2 In Section III, we presented a novel rectifier model that
 r    r 
−1 p −1 p dc
yields Pout for any given waveform. In this section, we deter-
= FT y − FT y − (18)
2 2 mine the receive and transmit waveforms that are optimal for
RF WPT based on the model.
where T ∼ ∪(0, T ]. Thus, the probability density function is
given as
 r   r  A. Optimization Problem Formulation
1 d −1 p d −1 p
f P ( p) = · y − y − . (19) The primary goal of any RF WPT system is to maximize
T dp 2 dp 2 dc
Pout with the available energy and spectrum resources at the
If the modulating signal equation Y = y(T ) is not invertible transmitter side. This can be broadly expressed in the form of
or monotonic as required in (19), then we can resort to an optimization problem as
numerical computation of f˜P ( p) as discussed in the previous
dc
case. max Pout
x(t)
Let us now present two special cases when f P ( p) is exactly
1 T
Z
known. These are the two-multisine and three-multisine modu- s.t. |x(t)|2 dt = PTx
lating signals, where the adjacent tones have equal amplitude T 0
1 W/2
Z
and the same phase, as defined in the experimental setup
|X ( f )|2 d f ≤ α PTx (23)
in [9]. The modulating signal in these cases is given by as T −W/2
q
 2PinRF cos π1 f T ,

where
 when N = 2
Y= q (20) Z ∞
 P RF 3 1 + 2 cos 2π1 f T , when N = 3.
  
 X( f ) = x(t) · e− j2π f t dt (24)
in −∞
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

AYIR et al.: PRACTICAL WAVEFORM-TO-EH MODEL AND TRANSMIT WAVEFORM OPTIMIZATION 7

represents the Fourier transform of x(t), having 0 < α < 1 or equivalently, when f P ( p) is implicitly constrained to be a
times the instantaneous bandwidth W .2 The optimization proper probability density function, as simply
problem in (23) is purposed to ascertain the optimal transmit
max E{η(P) · P}
waveform x ∗ (t). The first constraint ensures that the RF WPT f P ( p)
source transmits at the maximum available power level (PTx ), s.t. E{P} = PinRF (26b)
while the second constraint ensures that the signal resides
within the allotted channel of bandwidth W . which should yield the optimal power distribution f P∗ ( p) for
dc
If we now express the transmit waveform in a general form the envelope signal that maximizes Pout in RF WPT. It must
as defined in (1), the optimization problem narrows down to be noted that any number of waveforms could have the same
a waveform optimization problem where the multisine ampli- f P∗ ( p). Among those, we select the optimal waveform that
tude, frequency, and phase are optimized to yield maximum satisfies the constraints in (23) or (25) and other criteria.
dc
Pout . Moreover, given a wireless medium, there is an additional To maximize Poutdc
at a given PinRF , the receiver must operate
constraint on the spectral bandwidth of the waveform. The at maximum η at that PinRF . Now, it is evident from Fig. 1(c)
modified optimization problem can be expressed as that the EH receiver yields maximum η at only one unique
PinRF , say p̄, for a given R L . Consequently, the upper bound
dc
max Pout for (12) is given as
{An , f n ,φn }
( 
N E η( p̄) · P , PinRF ≤ p̄
E{η(P) · P} ≤
X
s.t. A2n = PTx η(E{ p}) · E{ p}, PinRF > p̄
n=1 (
W η( p̄) · PinRF , PinRF ≤ p̄
| fn | ≤ ∀n ∈ [1, N ] (25) = (27)
η PinRF · PinRF , PinRF > p̄


2
PN 2
where n=1 An denotes the transmit power as defined in where
(3). Again, the two constraints ensure the transmit power
p̄ = arg max η( p) (28)
and bandwidth limits are adhered, respectively. However, for p
yielding the globally optimum waveform with respect to (23), is the instantaneous input power of the envelope signal at
the above optimization problem requires also N to approach which maximum efficiency is attained. Furthermore, when
infinity. PinRF > p̄, we can assume that the product η(P) · P to be
We shall utilize the following approach for optimizing the concave and apply Jensen’s inequality, which is corroborated
waveform at the transmitter. First, we determine the opti- by actual experimental data, as shown in Fig. 1(b).
mal envelope signal at the harvester input which maximizes The upper bound implies that to maximize Pout dc
, the received
dc
Pout . Afterward, by incorporating the effect of a slow-fading envelope signal ought to have just one nonzero power level at
frequency-selective channel, we choose the transmit waveform p̄ or PinRF . Consequently, we propose a power density function
parameters that would yield the optimal waveform at the at the receiver, for an input signal with average RF input power
receiver input, while complying with the constraints. PinRF , given as
PinRF
 
f P ( p) = 1 − µ · δ( p) + µ · δ p − ∗
∗ ∗ ∗


B. Optimization Problem at the Receiver (29)
µ
The novel rectifier model presented in (12) relies on the where the parameter
power distribution of the envelope signal and thus is suitable 
RF
to be used for waveform optimization at the receiver. If we  Pin ,

PinRF ≤ p̄
dc
observe the model in (12), we realize that Pout is dependent µ =
∗
p̄ (30)
on η( p) and f P ( p), with the former pertaining to the energy 
 1, PinRF > p̄
harvester in use. Thus, only the latter parameter can be
controlled and optimized to maximize Poutdc
, which is the power represents the optimal relative duration of the power level.
distribution of the input RF waveform f P ( p). The waveform Employing the density function proposed in (29) and the
optimization problem at the receiver can thus be presented as subsequent µ∗ defined in (30), the average harvested dc power
is given as
∞ dc
Pout = E{η(P) · P}
Z
max η( p) · p · f P ( p)dp Z ∞
P RF
  
f P ( p)≥0
Z ∞0 η( p) · p · 1−µ∗ · δ( p)+µ∗ · δ p− in∗


= dp
µ
s.t. p · f P ( p)dp = PinRF 0
 RF 
Z0 ∞ P
= η in∗ · PinRF . (31)
f P ( p)dp = 1 (26a) µ
0 On substituting (30) in (31) and then comparing with (27),
2 The parameter 1 − α signifies the relative amount of spectral leakage
it is clear that (29) attains the upper bound and is thus an
dc
permitted in the adjacent frequency bands. The higher the value of α, the optimal solution for maximizing Pout . The proposed density
wider is the measured bandwidth of the transmitted signal. function represents a pulsed RF wave with duty cycle µ∗ .
 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. 1. Block diagram of the presented RF WPT system along with an example of measured harvested dc power (Pout dc ) and the corresponding receiver
efficiency (η) for a single sinusoid RF input at 879.5 MHz. The receiver is an off-the-shelf multiband rectifier with R L = 3.3 k and C L = 15 pF. The
maximum receiver efficiency η( p̄) ≈ 52% is achieved when PinRF = p̄ ≈ 11 dBm. When considering simultaneous power and data transfer, the RF WPT
system depicted in (a) is part of a more generic RF SWIPT system shown in [54, Fig. 1]. Such an RF SWIPT system constitutes separate signal paths for
information decoding and EH. (a) General block diagram. (b) Measured average harvested dc power. (c) Receiver efficiency.

This completes the optimization problem at the receiver. Next, represents a train of rectangular pulses with absolute pulse
we revert back to the original problem (25) of determining the duration τmin ≤ τ ≤ τmax and repetition interval T . The
optimal waveform at the transmitter. optimal duty cycle µ∗ in (30) can, thus, also be represented
as
τ
C. Optimal Waveforms at the Transmitter µ∗ = (35)
T
The optimal distribution at the receiver input as described
and, consequently, the optimal amplitude from the first con-
in (29) provides no information about the frequencies of the
straint in (23) is given as
optimal baseband signal. It only implies that the optimal enve- s
lope signal at the transmitter is also a rectangular pulse. At the PTx
transmitter side, the bandwidth constraint in (23) is critical, A= . (36)
µ∗
for which we assume that the spectrum regulations specify
α as a parameter for spectral containment and the maximum Now, to complete the waveform optimization problem
permissible absolute pulse duration τmax . The parameter τmax at the transmitter, we ought to determine the parameters
allows us to determine the minimum instantaneous bandwidth WSmin , τmin , T , and φ ∗ . We start by determining W
Smin and τmin
WSmin for the transmitted signal as follows. utilizing the second constraint in (23), based on the available
Now, with a pulsed RF signal, there is scope for frequency transmit power PTx and α. Consider the Fourier transform of
modulation at the baseband. However, to attain a high-PAPR x ∗ (t) given as
pulse at RF, we would need the absolute pulse duration to be X f ∗ + f = |Aτ sinc(τ f π)|.


(37)
as small as feasible, say τmin . This requires minimal frequency
variation at the baseband. Consequently, to maximize the Now, for any general bandwidth B, on substituting (37) in
power received through wireless propagation, we choose the the second constraint of (23) and using (36), we get
optimal baseband transmit frequency as 1 B/2
Z
α PTx ≥ |Aτ sinc(τ f π)|2 d f
f ∗ = arg max|H ( f )|2 T −B/2
f
A2 τ 2A2 τ 2 B/2
Z
Smin Smin  α sinc2 (τ f π)d f

W W ≥
s.t. f ∈ f L + , fU − (32) T T 0
2 2 ∞
X (−1)m (2π τ B)2m+1
where H ( f ) represents the Fourier transform of the baseband α≥ . (38)
π(2m + 1)(2m + 2)!
equivalent of channel ĥ(t), while f L and fU represent the m=0
lower and upper boundaries of the available channel bandwidth While it is not possible to invert (38) to obtain an expression
W = fU − f L , respectively. The total guard bandwidth W Smin for τ in terms of B, or vice versa, we can do so numerically
at the edges of W ensures that the spectral leakage to the and the results are shown in Fig. 2. Let us define a function
adjacent bands is restricted to (1 − α)PTx , as required in (23). q(τ, B) as
Now, the optimal baseband signal at the transmitter can be ∞
X (−1)m (2π τ B)2m+1
given by q(τ, B) = − α. (39)
m=0
π(2m + 1)(2m + 2)!
x ∗ (t) = exp j2π f ∗ t + jφ ∗ · r (t)
 
(33)
Smin in terms of τmax is the solution to
Then W
where
q τmax , W
Smin = 0


 (40)
 A,
 0<t <τ
r (t) = 0, τ <t <T (34) while τmin in terms of f is the solution to
∗

r (t + kT ), k ∈ Z q τmin , 2min f ∗ − f L , fU − f ∗ = 0.
 

(41)

 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

AYIR et al.: PRACTICAL WAVEFORM-TO-EH MODEL AND TRANSMIT WAVEFORM OPTIMIZATION 9

TABLE II
H ARDWARE C ONFIGURATION FOR THE T EST-B ED

that the envelope signal at the receiver varies slightly from

a rectangular wave due to delay spread. However, from an
RF WPT perspective, the distant multipath components are
already very insignificant due to the higher path loss they
suffer and thus can be neglected here.

Fig. 2. Absolute pulse duration corresponding to available instantaneous V. E XPERIMENTAL R ESULTS AND O BSERVATIONS
bandwidth for different values of parameter α as defined in (23).
In this section, we present the experimental results verifying
the theoretical results of Sections III and IV and discuss the
various observations. First, we present an overview of the
experimental test-bed setup developed for this work. Second,
we showcase the measurement results that corroborate the
efficacy of the proposed novel rectifier model for RF WPT
signals such as a CW, multisine signal, Gaussian noise, pulsed
RF, as well as RF SWIPT signals such as single-carrier (SC)
M-PSK, SC QAM, OFDM QPSK, and OFDM QAM. Finally,
we introduce experimental results that validate the superior
performance of the optimal pulsed RF waveform for RF WPT.

A. Measurement Setup
Fig. 3. Optimal signal at the transmitter: a pulsed RF signal with absolute A general block diagram of the test-bed used in this research
pulse duration τ , repetition interval T , and amplitude A, with the baseband work is presented in Fig. 1(a). The test-bed comprises a
frequency f ∗ ≪ f c . computer, a VST, an external PA, an RF energy harvester
accompanied by load resistance (R L ), load capacitance (C L ),
and an oscilloscope. The primary purpose of the test-bed is
WSmin for any τ is demonstrated in Fig. 2. Moreover, it is to determine the receiver efficiency η for the test waveforms.
evident in Fig. 2 that increasing α greatly impacts τmin . The receiver efficiency is determined as follows.
We observe that for instantaneous channel bandwidth up to A computer, equipped with MATLAB software, generates
10 MHz, τmin ≈ 0.1 µs for even 90% occupied bandwidth. the digital baseband waveforms and sends them to a VST,
This implies that transmitting multiple short pulses (τ < which generates an analog RF waveform for transmission.
0.1 µs) is not feasible, given the bandwidth and α constraints. The internal power amplification of the VST is not sufficient
In a practical RF WPT system, the transmitter always transmits for encompassing the entire power range of the RF energy
at the maximum permissible power PTx while utilizing the harvester, so we employ an external PA to generate the final
entire available bandwidth. Thus, the minimum pulse duration output RF signal with average power Pout RF
. The RF signal is
τ = τmin obtained from (41) is likely the best choice in transmitted to the RF energy harvester over a wired medium.3
practice. The received RF signal at the input of the RF EH has an
Next, the repetition interval T is obtained using (35) when τ average power PinRF . The RF energy harvester comprises a
is chosen. The determination of φ ∗ is insignificant from an RF matching network followed by a Schottky-diode-based rec-
WPT perspective and thus the task of waveform optimization tifier, which converts the incident RF energy to dc energy.
at the transmitter is complete. The optimal signal is a pulsed The dc rms value of the rectified waveform (Vout dc
) is measured
RF as shown in Fig. 3. The transmit waveform can carry
information through phase modulation, such as M-PSK, where 3 This research work primarily focuses on the receiver side of an RF

the amplitude of the signal stays constant. Thus, a pulsed WPT/SWIPT system. Moreover, we experimentally evaluate the receiver
efficiency for a multiband receiver with six RF bands [53], of which five
M-PSK signal, or more generally, a pulsed continuous phase bands are utilized for commercial wireless communications. Hence, we avoid
modulation signal is suitable for RF SWIPT. wireless measurements to avert any interference in the licensed RF bands.
Furthermore, it is essential to note that while the optimal Additionally, it was observed in [9] that for a few MHz bandwidth in the
863–873-MHz European ISM band, the wireless channel is essentially flat
envelope signal at the transmitter is a rectangular wave, and is well approximated with a path loss channel model, such as a wired
multipath propagation through a wireless medium ensures medium.
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

10 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES

TABLE III µ = 0.25 yields a pulsed RF waveform with a PAPR of 8,

O PERATIONAL PARAMETERS FOR THE E XPERIMENTS similar to that of an N = 4 multisine. Next, we choose
W ITH THE M ULTIBAND R ECTIFIER
a circularly symmetric complex Gaussian (CSCG) random
vector as the baseband signal for the Gaussian noise signal.
Consequently, the resultant envelope signal would have a
Rayleigh distribution. Furthermore, in the case of RF SWIPT
signals, we select the symbol rate to be 1 MS/s. Overall, the
operational parameters for the experiments are summarized in
Table III.
3) Rectifier Circuits: One of the novel contributions of this
research work is that the proposed rectifier model is invariant
of the receiver architecture, viz., the matching network and
the rectifier configuration. To validate this claim, we con-
duct the experiments for three different rectifier models: a
simple rectifier circuit that was fabricated for this research
work comprising a single Powercast PCC110 RF-to-dc con-
verter [61] with an LC matching network matched to −5 dBm;
an off-the-shelf multiband RF energy harvester Powercast
P21XXCSR [53] with support for six frequency bands, each
comprising two PCC110 chipsets with a T-matching network;
and a voltage multiplier rectifier circuit based on the HSMS-
282C diode, with an LC matching network matched to 0 dBm,
simulated in Keysight Advanced Design System (ADS) soft-
ware. The simple rectifier circuit developed on a PCB, along
with its component values is presented in Fig. 4(a), whereas
by an oscilloscope, which is regulated by the same computer the multiband rectifier along with the T-matching network
dc components is depicted in Fig. 4(b), while the schematic for
as before, through MATLAB. The computer registers the Vout
data for each test baseband digital waveform and computes the ADS simulation is presented in Fig. 4(c). In this way,
the receiver efficiency using (11). The hardware configuration we validate the proposed waveform-to-EH model for different
for the test bed is showcased in Table II. diode types, as well as different matching network and rectifier
1) Calibration of PinRF : We measure (and later calibrate) configurations.
RF
Pin , in our experiments, first by utilizing the real-time spec- In Section V-B, we shall present the experimental results
trum analyzer feature of the VST and later with the aid of an and compare them with simulation results based on the pro-
RF power meter. The measurements reveal that PinRF deviates posed rectifier model, in terms of the receiver efficiency. For
from its expected value owing to the inherent characteristics the rectifier model, we determine the instantaneous efficiency
of the VST and RF PA. This issue is resolved as follows. The η( p) from the measurements for a CW, as described in
spectrum analyzer and RF power meter readings reveal that Section III-A.
the VST used in this work could provide a linear output up to
0 dBm for a CW and that this linearity threshold declines with B. Results Validating the Proposed Rectifier Model
increasing the PAPR of the input signal. As a consequence, We measure the receiver efficiency for different RF WPT
RF
we opt for an RF PA to boost Pout to 20.5 dBm (so that and RF SWIPT waveforms, for varying R L , frequency band,
RF
Pin = 20 dBm after cable attenuation). However, the PA too and PinRF , and compare these outcomes with the predictions of
suffers from nonlinearity issues while transmitting high-PAPR the proposed rectifier model.
signals. To overcome the nonlinearity problem, we introduce 1) For Different Diode Types, Matching Networks, and
a lookup table-based power adjustment at the transmitter VST. Rectifier Configurations: In Fig. 5, we present the results
For each PinRF level, the calibrated lookup table entry is derived for frequency band 2, for different RF WPT waveforms. It is
from the corresponding spectrum analyzer and RF power meter evident from the three subfigures that the model predictions
measurements. The input power to the RF PA is adjusted agree closely with the measurement outcomes for all three
accordingly so that the true overall PinRF is linear and exactly as different RF EH circuits. Moreover, the measurements match
required, for all the waveforms and at all input power levels. very well even for multisine and pulsed RF waveforms, each
2) Operational Parameters: Let us now present the opera- with a different envelope signal distribution. Additionally,
tional parameters chosen for signals employed for the exper- we also validate the model for a CSCG noise signal, where
iments. For the baseband multisine signal, we choose f n = the envelope signal has a Rayleigh distribution, and the results
(n − ((N + 1)/2))1 f so that the center frequency of the in Fig. 5(a) and (b) confirm that the proposed model correctly
RF signal remains unchanged. Here, 1 f is the frequency predicts the receiver efficiency in this case as well. Overall,
spacing between adjacent tones. In the case of pulsed RF in each of the three rectifier models, we observe that a CW
waves, we vary the duty cycle µ to obtain a waveform and a pulsed RF wave (µ = 0.1) attain the highest η,
with a PAPR similar to a multisine signal. For example, which is about 37%, 52%, and5% for the fabricated rectifier,
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

AYIR et al.: PRACTICAL WAVEFORM-TO-EH MODEL AND TRANSMIT WAVEFORM OPTIMIZATION 11

multisine waveforms (which have been claimed to be suitable

for RF WPT in different research studies [7]) and we present
the results for frequency band 2 of the multiband rectifier,
in Fig. 6. It is apparent from Fig. 6 that the model predictions
are accurate, within the measurement threshold, for all the
multisines, from the low-PAPR two-multisine to the high-
PAPR 16-multisine. We observe that in the low-power region,
the high-PAPR multisines yield higher η: the higher the PAPR,
the higher the instantaneous power level of the multisine signal
which is essential to switch on the diodes at low PinRF levels.
Conversely, at higher PinRF levels, the higher instantaneous
power levels drive the diodes into saturation which may even
result in reverse current, thus reducing the rate of increase
dc
of Vout with PinRF , and consequently beginning the reduction
in η. This explains why the higher-PAPR signals reach the
maximum efficiency at lower PinRF , compared to lower-PAPR
signals and also why their receiver efficiency is significantly
worse thereafter.
Moreover, we also observe that the model predictions are
fairly precise across the entire range of PinRF , from the low
power region to the saturation region. An important observa-
tion is that while the 16-multisine is the suitable waveform
for lower PinRF , a CW still provides the maximum η, albeit at
a much higher PinRF .
While the high-PAPR multisine waveforms are a popular
choice for RF WPT, they are futile from a communication
scenario. Hence, from a more practical point of view, we inves-
tigate the performance of our model for commonly used infor-
mation signals appropriate for RF SWIPT, and the results are
showcased in Fig. 7. It is clear from the measurements that the
proposed model makes proper predictions for SWIPT signals
as well across the entire PinRF range, once their envelope signal
distribution is known. Since an M-PSK signal essentially has
the same envelope as a CW, the receiver efficiency for all
M-PSK signals is the same. This corroborates our hypothesis
of considering the instantaneous power of the envelope signal
as the basis for our proposed model. Furthermore, similar
to Fig. 6, we observe in Fig. 7 that the high-PAPR OFDM
modulations are more suitable for lower PinRF region than the
low-PAPR SC modulations and worse at higher PinRF region
Fig. 4. RF energy harvesters employed in the experiments. (a) Rectifier and vice versa. Even within the SC modulations, a 64-QAM
circuit developed for this research work, comprising an LC matching network, signal with a slightly higher PAPR than M-PSK signals yields
a Powercast PCC110 RF-to-dc converter [61] with fixed load capacitance marginally better receiver efficiency than M-PSK modulations
and resistance. The oscilloscope probe adds a series capacitance of 15 pF,
resulting in a total C L = 115 pF. (b) Off-the-shelf multiband RF energy at lower PinRF region.
harvester P21XXCSR [53], with each band comprising a T-matching network It must be noted that the two measured OFDM modulations
(as depicted separately) and two PCC110 RF-to-dc converters with fixed load seem to yield different receiver efficiency since these are each
capacitance and varying load resistance, which are connected to jumper pin
Z [see Fig. 1(a)] corresponding to each band. The 15-pF capacitance of the a single finite-length realization. However, on average, the
oscilloscope probe acts as C L . (c) Keysight ADS simulation of a rectifier PAPR of an OFDM single is independent of the underly-
circuit, comprising the SPICE model of an HSMS-282C series diode [62] ing modulation for high subcarrier points. Nonetheless, it is
connected in voltage multiplier configuration with a variable-load resistor and
capacitor.
inconvenient to perform a large number of measurements
to obtain an average result, and in such a case having a
model that can precisely predict the performance is preferred,
multiband rectifier, and simulated voltage-multiplier rectifier, since averaging over simulations through the model is much
respectively. Henceforth, we present the results only for the more convenient. Accordingly, we present the averaged OFDM
multiband rectifier. simulation result, based on the proposed model, in Fig. 7
2) For Multisine Waveforms and RF SWIPT Waveforms: and we observe that η is indeed independent of the baseband
Next, we evaluate the proposed model’s predictions for the modulation.
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

12 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES

Fig. 5. Efficacy of the proposed rectifier model evaluated for three different rectifier circuits, while varying PinRF . For each rectifier, we compare the measured
η for a CW, a multisine, a pulsed RF wave, and a noise signal with Rayleigh distributed amplitude. The measurement/simulation results are presented with
markers, whereas the model prediction is displayed as a solid line. (a) Receiver efficiency of RF WPT waveforms for the fabricated RF energy harvester
with R L = 287  and C L = 115 pF. (b) Receiver efficiency of RF WPT waveforms for an off-the-shelf multiband RF energy harvester with R L = 3.3 k
and C L = 15 pF in the frequency band 2. (c) Receiver efficiency of RF WPT waveforms for a voltage multiplier rectifier circuit simulated in ADS with
R L = 1 k and C L = 100 pF.

Fig. 6. Efficacy of the proposed rectifier model evaluated for multisine Fig. 7. Efficacy of the proposed rectifier model evaluated for RF SWIPT
waveforms, while varying PinRF , with R L = 3.3 k, and C L = 15 pF in waveforms, while varying PinRF , with R L = 3.3 k and C L = 15 pF in
the frequency band 2. The measurement results are presented with markers, the frequency band 2. Here, OFDM QPSK and OFDM 64-QAM represent
whereas the model prediction is displayed as a solid line. single high-PAPR realizations, while the others are general realizations of
these waveforms. In general, the receiver efficiency for OFDM waveform
is found to be independent of the baseband modulation. The measurement
results are presented with markers and the model predictions with solid lines.

It must be noted that in the case of an RF WPT system,

such as the one considered in our study, the received PinRF
varies between −30 and 20 dBm, since the best receiver
sensitivity of even commercial RF EH receivers varies between and pulsed RF), PAPR and matching networks, and the results
−30 and −15 dBm [50], [53], [54]. In contrast, the receiver thereof along with the model predictions are depicted in the
sensitivity of a typical communication node is much lower subplots of Fig. 8. These plots exhibit the efficacy of the
(≤ −80 dBm). Therefore, in the power range that is relevant proposed model in anticipating the receiver efficiency for a
to RF EH (even while employing the power-splitting scheme in wide range of load resistance values. In Fig. 8(a), we present
RF SWIPT), the bit error rate for the communication receiver the receiver efficiency for different multisines (thus for varying
is practically zero [50]. We confirmed the same through wire- PAPR) at certain PinRF . The PinRF levels are chosen to present
less measurements for varying transmitter–receiver separation a clear, unconstricted figure. It is evident from Fig. 8(a)
relevant to RF SWIPT in [54]. Hence, we omit to present that the proposed model can correctly estimate the receiver
the communication performance of the RF SWIPT waveforms efficiency for all the multisines and the entire range of R L .
in this study. Now, based on the observations of Figs. 5–7, The observations regarding the variation of receiver efficiency
we can deduce that, with the knowledge of f P ( p) and η( p), with R L and multisine PAPR are consistent with those in [26]:
the proposed model can rightly predict the receiver efficiency At lower PinRF values, a large R L is needed to generate higher
for any digital baseband waveform. dc
Vout and high-PAPR signals with their higher instantaneous
3) For Different Load Resistors: Next, we experimentally values are more suitable. Contrarily at high PinRF values, a low
evaluated the receiver efficiency for the various load resistors R L is needed to avoid saturating the receiver and thus low-
described in Table III for different input waveforms (multisines PAPR signals are more suitable.
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

AYIR et al.: PRACTICAL WAVEFORM-TO-EH MODEL AND TRANSMIT WAVEFORM OPTIMIZATION 13

Fig. 9. Contour plots for average input RF power versus load resistance,
showcasing the measurement results and model predictions of η for different
RF WPT waveforms. The measurement results are presented with markers,
whereas the model prediction is displayed as a solid line. (a) η = 25%.
(b) η = 50%.

Fig. 8(b). It is evident that the proposed model is capable of

anticipating the receiver efficiency for pulsed RF waveform,
within the measurement tolerance limits, for the entire range
of load resistance R L . The model outcomes are valid for
low as well as high PinRF regimes, including the saturation
region. While the results shown in Fig. 8(a) and (b) corre-
spond to the multiband RF energy harvester that comprises a
T-matching network, the effectiveness of the proposed model
in anticipating the receiver efficiency for an RF EH receiver
Fig. 8. Efficacy of the proposed rectifier model evaluated for different with a different matching network is shown in Fig. 8(c).
waveforms, PAPR, and matching networks, while varying R L . The multiband
harvester comprises a T-matching network, whereas the simulated rectifier We resort to ADS simulations for this purpose and utilize the
circuit comprises an LC matching network. The measurement results are voltage multiplier rectifier circuit shown in Fig. 4(c) which
presented with markers and the model prediction is displayed as a solid line. comprises an LC matching network. We present the results
(a) Receiver efficiency of different multisine waveforms for varying R L , with
C L = 15 pF, at different PinRF in the frequency band 2 of the multiband energy for an eight-multisine signal in Fig. 8(c), wherein we observe
harvester. (b) Receiver efficiency of a pulsed RF waveform with 50% duty that the efficacy of the proposed waveform-to-EH model in
cycle for varying R L , with C L = 15 pF, at different PinRF in the frequency anticipating the receiver efficiency is invariant of the change
band 2 of the multiband energy harvester. (c) Receiver efficiency of the
eight-multisine waveform for varying R L , for the voltage multiplier rectifier in the matching network, for the entire range of PinRF and all
circuit with the LC matching network, simulated in ADS with C L = 100 pF. values of load resistance R L .
Now, to further affirm the effectiveness of the proposed
rectifier model for a wide range of PinRF and R L simultaneously,
Furthermore, the receiver efficiency has been evaluated for we present the contour plots of constant η versus PinRF and
the optimal pulsed RF waveform and the results are shown in R L , similar to [63]. The contour plots for η = 25% and 50%
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

14 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES

Fig. 10. Efficacy of the proposed rectifier model evaluated for different Fig. 11. Evaluating the receiver efficiency of multisine waveforms and
frequency bands of the multiband rectifier, for a µ = 0.25 pulsed RF pulsed RF waveforms to determine the waveform more suitable for RF WPT.
waveform, while varying PinRF , with R L = 3.3 k and C L = 15 pF. The The experimental results and model prediction corroborate the theoretical
measurement results are presented with markers, whereas the model prediction outcomes of the waveform optimization in Section IV.
is displayed as a solid line.

performance of a pulsed RF waveform in RF WPT. Here as

are shown in Fig. 9(a) and (b), respectively, for multisines
well, we present the results for the frequency band 2 of the
as well as pulsed RF waveforms with varying PAPR. The
multiband rectifier, with the operational parameters being the
primary observation is that the model predictions match well
same as before.
with the measurements. The aberrations can be accredited to
1) Pulsed RF Versus Multisines: To verify the theoretical
measurement inaccuracies owing to the resolution limitations
result of Section IV, we experimentally determined the
of the oscilloscope at high R L for high PAPR signals, where
receiver efficiency η of a pulsed RF wave for varying duty
the instantaneous voltage is very high. Furthermore, although
cycle µ and compared these with the measurements with
the two-multisine and µ = 0.5 pulsed RF waveform have
N -multisines. We opt for a multisine, with the same PinRF ,
the same PAPR, we observe that the pulsed RF waveform
as the reference waveform since multisines are recommended
achieves the same η at a lower PinRF . Even more remarkable
as optimal waveforms for RF WPT [7]. An N -multisine signal
is the observation that µ = 0.25 pulsed RF waveform has a
will have the same PAPR as µ = (1/N )-pulsed RF signal
contour plot similar to eight-multisine, even though the latter
and is expected to yield similar η performance. However, our
has twice the PAPR. We shall explore this in further detail in
model reveals (and the measurements confirm) that this is
Section V-C1.
not the case. To have an unconstricted plot, we choose to
4) For Different Frequency Bands: Thus far, we presented
showcase the plots obtained from the proposed rectifier model
the results only for the frequency band 2 of the multiband
in Fig. 11. Before the onset of saturation, a µ = (1/N )-
rectifier. We now confirm whether the proposed model can
pulsed RF signal yields higher efficiency than an N -multisine
make correct predictions of η for other the frequency bands
due to the former having peak instantaneous power for a
described in Table III. In Fig. 10, we present the measurement
longer duration. However, once the diodes enter the saturation
and modeling results for different RF bands for a pulsed
region, so-clipped multisine signals have a longer duration of
RF wave with 25% duty cycle. Although the multiband
peak instantaneous power, thus yielding higher η. Additionally,
harvester comprises the same RF-to-dc rectifier chips in all
it also results in a µ = (1/N )-pulsed RF signal reaching
the frequency bands [53], the preceding T-matching network
peak η at a lower PinRF than an N -multisine. These effects
components vary in each frequency band. It is evident in
can be observed for the cases of N = 2 and 32 in Fig. 11.
Fig. 10 that the proposed model can correctly predict η for all
Consequently, we need a pulsed RF with a longer duty cycle
the frequency bands of the off-the-shelf multiband rectifier.
µ = (1/M)(M < N ) for that signal to have a peak η at
This reaffirms that the efficacy of the proposed model is
the same PinRF as an N -multisine, as seen in the other two
invariant of the underlying matching network. Among all the
cases in Fig. 11. Overall, it is evident that a pulsed RF signal
bands, frequency band 3 of the multiband rectifier yields the
outperforms an N -multisine signal at any given PinRF that is
highest η of about 72%. Meanwhile, the unlicensed band 6 has
viable in RF WPT.
a maximum η of about 33%.
2) Receiver Efficiency η Variation With Duty Cycle µ: Our
theoretical exercise in Section IV led us to conclude that a
C. Results Validating the Waveform Optimization pulsed RF waveform with duty cycle µ is optimal for RF
The waveform optimization problem based on the proposed WPT, with the impulse signal being ideal, and the choice
novel rectifier model was discussed in Section IV, wherein we of µ is influenced by the available bandwidth and permitted
concluded that a pulsed RF waveform is an optimal transmit signal duration T . In Fig. 12, we present the variation of
waveform. In this section, we focus specifically on the η η for varying µ. We observe that the proposed model can
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

AYIR et al.: PRACTICAL WAVEFORM-TO-EH MODEL AND TRANSMIT WAVEFORM OPTIMIZATION 15

Fig. 12. Variation of receiver efficiency with varying µ and PinRF for a pulsed Fig. 14. Maximum attainable receiver efficiency at each PinRF across all the
RF waveform, determined by the proposed rectifier model and verified with load resistors, and the corresponding duty cycle for the frequency band 2 of
measurements, with R L = 3.3 k and C L = 15 pF in the frequency band 2. the multiband rectifier. The 10-k load resistor is optimal for a major portion
The model allows us to ascertain the optimal µ that attains maximum η at of PinRF , notably in the low-power regime.
each PinRF . The measurement results are presented with markers, whereas the
model prediction is displayed as a solid line.

varies as we vary both the load resistance R L and µ, and the

results are illustrated in Fig. 13. Apparently, for a given µ,
the smallest R L = 57  yields ηmax ≈ 21%, while the highest
R L = 10 k yields ηmax ≈ 62%. The smaller load resistors
are suitable at higher PinRF , while the larger R L are suited
for lower PinRF , as explained in Section V-B3. In some cases,
multiple load resistors provide the same ηmax but at different
PinRF : for example, R L = 470, 2200, and 7800  each yield
ηmax = 50%. In such cases, R L providing the desired Vout dc

is suitable. Once again, in general, the desired ηmax can be

attained at a lower PinRF by reducing the duty cycle µ.

D. Maximum Receiver Efficiency at Each PinRF

Once we figure out the maximum receiver efficiency ηmax
attainable for each load resistance R L , the next apparent goal
Fig. 13. Variation of receiver efficiency with duty cycle and load resistance is to determine the ηmax achievable for each average input
for a pulsed RF wave. We measure the receiver efficiency with a CW for
different resistors and the model allows us to determine the receiver efficiency RF power level PinRF . Based on the measurements of receiver
at any duty cycle µ, which shifts the curve to the left by 10 log10 (µ). The efficiency for a CW for different R L , and the subsequent model
maximum attainable efficiency is ≈62% for R L = 10 k. predictions for varying µ, we present the maximum attainable
η for each PinRF in Fig. 14. For the multiband rectifier in use,
we observe in Fig. 13 that R L = 10 k yields the maximum
correctly predict the impact of varying µ. The mismatch at efficiency for a CW, among all R L , at PinRF = 5 dBm. It is
lower PinRF can be attributed to measurement inaccuracies of possible to attain the same maximum efficiency for any PinRF <
the oscilloscope and the high noise floor of the PA. For a given 5 dBm by reducing µ. The optimal duty cycle at each PinRF is
PinRF , reducing µ by (1/N ) increases the peak instantaneous also shown in Fig. 14. Moreover, the optimal R L which yields
power of the RF waveform by N , when compared to a CW the maximum η at each PinRF is also depicted in Fig. 14.
(µ = 1). The increased peak instantaneous power yields higher
η at lower PinRF similar to multisines, albeit better. In fact,
reducing µ by N shifts the η plot for a CW in Fig. 12 by VI. C ONCLUSION
−10log10 µ dB. The proposed model allows us to determine This article studies the receiver efficiency of an RF WPT
η for very low values of µ, which is not possible to evaluate system and presents a novel rectifier model that characterizes
experimentally due to the limited dynamic range of the PA. the average harvested dc power in terms of the baseband signal
In Fig. 12, we reduce µ progressively and track the maximum of the input RF waveform. The proposed parametric model
receiver efficiency ηmax , which is the peak efficiency attainable is independent of the underlying matching network or the
at each PinRF , while employing the frequency band 2 of the rectifier and thus is applicable and accurate for any EH receiver
multiband rectifier with R L = 3300 . for which model parameters are estimated from measurements
3) Maximum Receiver Efficiency ηmax Versus Load Resis- or simulations. The rectifier model allows us to design the
tance and Duty Cycle: Furthermore, we examine how ηmax optimal transmit baseband signal for RF WPT such that
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

16 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES

the average harvested dc power is maximized. Furthermore, [17] H. Rahmani and A. Babakhani, “A wireless power receiver with an
a test bed to experimentally evaluate the receiver efficiency on-chip antenna for millimeter-size biomedical implants in 180 nm
SOI CMOS,” in IEEE MTT-S Int. Microw. Symp. Dig., Jun. 2017,
for any digital waveform is introduced. The experimental pp. 300–303.
results and simulations, both validate the predictions of the [18] Y. Sun, B. Greet, D. Burkland, M. John, M. Razavi, and A. Babakhani,
proposed rectifier model for popular RF WPT and RF SWIPT “Wirelessly powered implantable pacemaker with on-chip antenna,” in
IEEE MTT-S Int. Microw. Symp. Dig., Jun. 2017, pp. 1242–1244.
waveforms as well as validate the theoretical result of pulsed
[19] J. F. Kuhling, M. Feenaghty, and R. Dahle, “A wideband cascaded skew
RF waveforms being superior to multisine waveforms for RF planar wheel antenna for RF energy harvesting,” in Proc. IEEE Wireless
WPT and suitable for RF SWIPT. The receiver efficiency of Power Transf. Conf. (WPTC), Jun. 2018, pp. 1–4.
SWIPT with PSK, QAM, and OFDM is also evaluated through [20] M. M. Fakharian, “A wideband Rectenna using high gain fractal planar
monopole antenna array for RF energy scavenging,” Int. J. Antennas
the model and verified experimentally. Propag., vol. 2020, pp. 1–10, Jun. 2020.
[21] K. Saito, E. Nishiyama, and I. Toyoda, “A 2.45- and 5.8-GHz dual-
R EFERENCES band stacked differential Rectenna with high conversion efficiency in low
power density environment,” IEEE Open J. Antennas Propag., vol. 3,
[1] B. Clerckx, A. Costanzo, A. Georgiadis, and N. B. Carvalho, “Toward pp. 627–636, 2022.
1G mobile power networks: RF, signal, and system designs to make [22] J. Kim, B. Clerckx, and P. D. Mitcheson, “Prototyping and experi-
smart objects autonomous,” IEEE Microw. Mag., vol. 19, no. 6, mentation of a closed-loop wireless power transmission with channel
pp. 69–82, Sep. 2018. acquisition and waveform optimization,” in Proc. IEEE Wireless Power
[2] T. D. Ponnimbaduge Perera, D. N. K. Jayakody, S. K. Sharma, Transf. Conf. (WPTC), May 2017, pp. 1–4.
S. Chatzinotas, and J. Li, “Simultaneous wireless information and power [23] A. Georgiadis, G. V. Andia, and A. Collado, “Rectenna design and
transfer (SWIPT): Recent advances and future challenges,” IEEE Com- optimization using reciprocity theory and harmonic balance analysis
mun. Surveys Tuts., vol. 20, no. 1, pp. 264–302, 1st Quart., 2018. for electromagnetic (EM) energy harvesting,” IEEE Antennas Wireless
[3] M. Rayes, G. Nagib, and W. Abdelaal, “A review on wireless power Propag. Lett., vol. 9, pp. 444–446, 2010.
transfer,” Int. J. Eng. Trends Technol., vol. 40, no. 5, pp. 272–280, [24] P. N. Duy, N. Ha-Van, and C. Seo, “A design of 5.8 GHz rectenna array
Oct. 2016. for wireless energy harvesting applications,” in Proc. IEEE Wireless
[4] Wireless Power Consortium. Magnetic Resonance and Magnetic Power Transf. Conf. (WPTC), Nov. 2020, pp. 87–90.
Induction. Accessed: Apr. 6, 2023. [Online]. Available: [25] N. Ayir, M. Heino, and T. Riihonen, “Impact of input impedance
https://www.wirelesspowerconsortium.com/data/downloadables mismatch on the receiver efficiency of RF wireless power transfer,” in
/1/2/4/6/magnetic-resonance-or-magnetic-induction.pdf Proc. Wireless Power Week (WPW), Jul. 2022, pp. 412–416.
[5] A. A. DigiKey. Inductive Versus Resonant Wireless Charging: [26] N. Ayir and T. Riihonen, “Joint impact of input power, PAPR, and
Truce May Be Designer’s Best Choice. Accessed: Apr. 6, 2023. load resistance on the receiver efficiency of multisine waveforms in RF
[Online]. Available: https://www.digikey.fi/en/articles/inductive-versus- energy harvesting,” in Proc. IEEE Wireless Power Transf. Conf. (WPTC),
resonant-wireless-charging Jun. 2021, pp. 1–4.
[6] Ansys. Wireless Charging Technologies: Magnetic Resonance vs. Mag- [27] E. Boshkovska, D. W. K. Ng, N. Zlatanov, and R. Schober, “Prac-
netic Induction vs. Rf Harvesting. Accessed: Apr. 6, 2023. [Online]. tical non-linear energy harvesting model and resource allocation for
Available: https://www.ansys.com/blog/ces-wireless-charging-magnetic- SWIPT systems,” IEEE Commun. Lett., vol. 19, no. 12, pp. 2082–2085,
resonance-induction-RF-harvesting/ Dec. 2015.
[7] B. Clerckx, R. Zhang, R. Schober, D. W. K. Ng, D. I. Kim, and [28] M. Babaei, L. Durak-Ata, and Ü. Aygölü, “New practical nonlinear
H. V. Poor, “Fundamentals of wireless information and power transfer: energy-harvesting models for wireless-powered communications,” Wire-
From RF energy harvester models to signal and system designs,” IEEE less Commun. Mobile Comput., vol. 2022, pp. 1–19, Jun. 2022.
J. Sel. Areas Commun., vol. 37, no. 1, pp. 4–33, Jan. 2019.
[29] X. Xu, A. Özçelikkale, T. McKelvey, and M. Viberg, “Simultaneous
[8] Y. Chen, K. T. Sabnis, and R. A. Abd-Alhameed, “New for- information and power transfer under a non-linear RF energy harvesting
mula for conversion efficiency of RF EH and its wireless applica- model,” in Proc. IEEE Int. Conf. Commun. Workshops (ICC Workshops),
tions,” IEEE Trans. Veh. Technol., vol. 65, no. 11, pp. 9410–9414, May 2017, pp. 179–184.
Nov. 2016.
[30] E. Boshkovska, D. W. K. Ng, N. Zlatanov, A. Koelpin, and R. Schober,
[9] N. Ayir, T. Riihonen, M. Allén, and M. F. T. Fierro, “Waveforms
“Robust resource allocation for MIMO wireless powered communication
and end-to-end efficiency in RF wireless power transfer using digital
networks based on a non-linear EH model,” IEEE Trans. Commun.,
radio transmitter,” IEEE Trans. Microw. Theory Techn., vol. 69, no. 3,
vol. 65, no. 5, pp. 1984–1999, May 2017.
pp. 1917–1931, Mar. 2021.
[10] J. Hu, K. Yang, G. Wen, and L. Hanzo, “Integrated data and energy [31] R. Jiang, K. Xiong, P. Fan, Y. Zhang, and Z. Zhong, “Optimal design
communication network: A comprehensive survey,” IEEE Commun. of SWIPT systems with multiple heterogeneous users under non-linear
Surveys Tuts., vol. 20, no. 4, pp. 3169–3219, 4th Quart., 2018. energy harvesting model,” IEEE Access, vol. 5, pp. 11479–11489, 2017.
[11] Y. Alsaba, S. K. A. Rahim, and C. Y. Leow, “Beamforming in wireless [32] A. J. S. Boaventura and N. B. Carvalho, “Maximizing DC power in
energy harvesting communications systems: A survey,” IEEE Commun. energy harvesting circuits using multisine excitation,” in IEEE MTT-S
Surveys Tuts., vol. 20, no. 2, pp. 1329–1360, 2nd Quart., 2018. Int. Microw. Symp. Dig., Jun. 2011, pp. 1–4.
[12] C. R. Valenta and G. D. Durgin, “Harvesting wireless power: Survey [33] B. Clerckx and E. Bayguzina, “Waveform design for wireless power
of energy-harvester conversion efficiency in far-field, wireless power transfer,” IEEE Trans. Signal Process., vol. 64, no. 23, pp. 6313–6328,
transfer systems,” IEEE Microw. Mag., vol. 15, no. 4, pp. 108–120, Dec. 2016.
Jun. 2014. [34] S. Abeywickrama, R. Zhang, and C. Yuen, “Refined nonlinear Rectenna
[13] X. Lu, P. Wang, D. Niyato, D. I. Kim, and Z. Han, “Wireless networks modeling and optimal waveform design for multi-user multi-antenna
with RF energy harvesting: A contemporary survey,” IEEE Commun. wireless power transfer,” IEEE J. Sel. Topics Signal Process., vol. 15,
Surveys Tuts., vol. 17, no. 2, pp. 757–789, 2nd Quart., 2015. no. 5, pp. 1198–1210, Aug. 2021.
[14] A. Boaventura, D. Belo, R. Fernandes, A. Collado, A. Georgiadis, [35] M. R. V. Moghadam, Y. Zeng, and R. Zhang, “Waveform optimization
and N. B. Carvalho, “Boosting the efficiency: Unconventional waveform for radio-frequency wireless power transfer: (Invited paper),” in Proc.
design for efficient wireless power transfer,” IEEE Microw. Mag., vol. 16, IEEE 18th Int. Workshop Signal Process. Adv. Wireless Commun.
no. 3, pp. 87–96, Apr. 2015. (SPAWC), Jul. 2017, pp. 1–6.
[15] N. Ayir, M. F. T. Fierro, T. Riihonen, and M. Allén, “Experimenting [36] Y. Huang and B. Clerckx, “Waveform design for wireless power transfer
waveforms and efficiency in RF power transfer,” in a, Jun. 2019, with limited feedback,” IEEE Trans. Wireless Commun., vol. 17, no. 1,
pp. 1140–1143. pp. 415–429, Jan. 2018.
[16] Md. A. Ullah, R. Keshavarz, M. Abolhasan, J. Lipman, K. P. Esselle, [37] L. Cantos, G. Sacarelo, Y. Jeong, and Y. H. Kim, “Performance of a
and N. Shariati, “A review on antenna technologies for ambient RF waveform design for wireless power transfer with imperfect channel
energy harvesting and wireless power transfer: Designs, challenges and state information,” in Proc. Int. Conf. Inf. Commun. Technol. Converg.
applications,” IEEE Access, vol. 10, pp. 17231–17267, 2022. (ICTC), Oct. 2017, pp. 1293–1295.
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

AYIR et al.: PRACTICAL WAVEFORM-TO-EH MODEL AND TRANSMIT WAVEFORM OPTIMIZATION 17

[38] E. Boshkovska, A. Koelpin, D. W. K. Ng, N. Zlatanov, and R. Schober, [60] Farnell. R&S NRP-Z11 Datasheet. Accessed: Apr. 6, 2023. [Online].
“Robust beamforming for SWIPT systems with non-linear energy har- Available: https://www.farnell.com/datasheets/2176138.pdf
vesting model,” in Proc. IEEE 17th Int. Workshop Signal Process. Adv. [61] Powercast. PCC110 Chipset Datasheet. Accessed: Apr. 6, 2023.
Wireless Commun. (SPAWC), Jul. 2016, pp. 1–5. [Online]. Available: https://www.powercastco.com/documentation/
[39] H. Sun, F. Zhou, and Z. Zhang, “Robust beamforming design in a pcc110-pcc210-chipset-overview/
NOMA cognitive radio network relying on SWIPT,” in Proc. IEEE Int. [62] Datasheetspdf. HSMS-282X Datasheet. Accessed: Apr. 6, 2023.
Conf. Commun. (ICC), May 2018, pp. 1–6. [Online]. Available: https://datasheetspdf.com/pdf-file/963462/AVAGO/
[40] B. Clerckx and J. Kim, “On the beneficial roles of fading and transmit HSMS-282C/1
diversity in wireless power transfer with nonlinear energy harvest- [63] F. Bolos, J. Blanco, A. Collado, and A. Georgiadis, “RF energy
ing,” IEEE Trans. Wireless Commun., vol. 17, no. 11, pp. 7731–7743, harvesting from multi-tone and digitally modulated signals,” IEEE Trans.
Nov. 2018. Microw. Theory Techn., vol. 64, no. 6, pp. 1918–1927, Jun. 2016.
[41] Y. Zeng, B. Clerckx, and R. Zhang, “Communications and signals design
for wireless power transmission,” IEEE Trans. Commun., vol. 65, no. 5,
pp. 2264–2290, May 2017.
[42] J. Kim, B. Clerckx, and P. D. Mitcheson, “Signal and system design for
wireless power transfer : Prototype, experiment and validation,” 2019,
arXiv:1901.01156. Nachiket Ayir (Member, IEEE) received the M.Sc.
[43] E. Davut, O. Kazanci, A. Caglar, D. Altinel, M. B. Yelten, and degree in electronics and communications engineer-
G. K. Kurt, “A test-bed based guideline for multi-source energy har- ing from the International Institute of Information
vesting,” in Proc. 10th Int. Conf. Elect. Electron. Eng., Nov. 2017, Technology, Hyderabad, India, in May 2018.
pp. 1267–1271. He is currently with the Faculty of Informa-
[44] M. Cansiz, D. Altinel, and G. K. Kurt, “Effects of different modulation tion Technology and Communication Sciences, Tam-
techniques on charging time in RF energy-harvesting system,” IEEE pere University, Tampere, Finland, as a Doctoral
Trans. Instrum. Meas., vol. 69, no. 9, pp. 6904–6911, Sep. 2020. Researcher. His current research interests include
[45] S. Nikoletseas, T. Raptis, A. Souroulagkas, and D. Tsolovos, “Wireless wireless energy harvesting, optimization techniques,
power transfer protocols in sensor networks: Experiments and simula- software-defined radios, and simultaneous wireless
tions,” J. Sensor Actuator Netw., vol. 6, no. 2, p. 4, Apr. 2017. information and power transfer.
[46] K. Li, C. Yuen, and S. Jha, “Poster: Fair scheduling for energy harvesting
WSN in smart city,” in Proc. 13th ACM Conf. Embedded Networked
Sensor Syst., Nov. 2015, pp. 419–420.
[47] H. Dai, X. Wang, A. X. Liu, F. Zhang, Y. Zhao, and G. Chen,
“Omnidirectional chargability with directional antennas,” in Proc. IEEE
24th Int. Conf. Netw. Protocols (ICNP), Nov. 2016, pp. 1–10. Taneli Riihonen (Senior Member, IEEE) received
[48] P. S. Yedavalli, T. Riihonen, X. Wang, and J. M. Rabaey, “Far-field RF the D.Sc. degree (Hons.) in electrical engineering
wireless power transfer with blind adaptive beamforming for Internet of from Aalto University, Espoo, Finland, in August
Things devices,” IEEE Access, vol. 5, pp. 1743–1752, 2017. 2014.
[49] X. Fan et al., “Energy-ball: Wireless power transfer for batteryless Inter- He held various research positions at the School
net of Things through distributed beamforming,” Proc. ACM Interact., of Electrical Engineering, Aalto University, from
Mobile, Wearable Ubiquitous Technol., vol. 2, no. 2, pp. 1–22, Jul. 2018. September 2005 to December 2017. He was a Vis-
[50] J. Kim, B. Clerckx, and P. D. Mitcheson, “Experimental analysis of iting Associate Research Scientist and an Adjunct
harvested energy and throughput trade-off in a realistic SWIPT system,” Assistant Professor with Columbia University, New
in Proc. IEEE Wireless Power Transf. Conf. (WPTC), Jun. 2019, pp. 1–5. York, NY, USA, from November 2014 to December
[51] S. Claessens, N. Pan, D. Schreurs, and S. Pollin, “Multitone FSK 2015. He is currently an Associate Professor (Tenure
modulation for SWIPT,” IEEE Trans. Microw. Theory Techn., vol. 67, Track) with the Faculty of Information Technology and Communication
no. 5, pp. 1665–1674, May 2019. Sciences, Tampere University, Tampere, Finland. His research interests include
[52] B. Clerckx, J. Kim, K. W. Choi, and D. I. Kim, “Foundations of wireless physical-layer OFDM(A), multiantennas, relaying, and full-duplex wireless
information and power transfer: Theory, prototypes, and experiments,” techniques with current interest in the evolution of 6G systems.
Proc. IEEE, vol. 110, no. 1, pp. 8–30, Jan. 2022. Mr. Riihonen received the Finnish Technical Sector’s Award for the best
[53] Powercast. P211XXCSR-EVB Datasheet. Accessed: Apr. 6, 2023. doctoral dissertation of the year and the EURASIP Best Ph.D. Thesis Award
[Online]. Available: https://www.powercastco.com/documentation/ 2017. He has been nominated 12 times as an Exemplary/Top Reviewer/Editor
p21xxcsr-evb-datasheet/ of various IEEE journals and is serving as an Editor for IEEE T RANSACTIONS
[54] N. Ayir and T. Riihonen, “Efficiency–throughput trade-off of pulsed RF ON W IRELESS C OMMUNICATIONS . He has previously served as an Editor
waveforms in simultaneous wireless information and power transfer,” for IEEE C OMMUNICATIONS L ETTERS and IEEE W IRELESS C OMMUNI -
in Proc. Wireless Power Technol. Conf. Expo (WPTCE), Jun. 2023, CATIONS L ETTERS .
pp. 1–15.
[55] R. Zhang and C. K. Ho, “MIMO broadcasting for simultaneous wire-
less information and power transfer,” IEEE Trans. Wireless Commun.,
vol. 12, no. 5, pp. 1989–2001, May 2013.
[56] R. Burden and J. Faires, Numerical Analysis. Pacific Grove, CA, USA:
Brooks Cole, 2004. Mikko Heino (Member, IEEE) received the D.Sc.
[57] Mini-Circuits. Mini-Circuits ZHL-4240+ Datasheet. Accessed: degree in radio engineering from Aalto University,
Apr. 6, 2023. [Online]. Available: https://www.minicircuits. Espoo, Finland, in 2020.
com/pdfs/ZHL-4240+.pdf He is currently a Post-Doctoral Research Fellow
[58] Apex Waves. NI PXIe-5645R Specifications. Accessed: Apr. 6, 2023. with the Unit of Electrical Engineering, Tampere
[Online]. Available: https://www.apexwaves.com/pdf/manuals/PXIe- University, Tampere, Finland. His research interests
5665/pxie-5645r-specifications.pdf include antenna isolation improvement methods,
joint communication and sensing, in-band full-
[59] Keysight. Keysight MSOX4104A Datasheet. Accessed: Apr. 6, 2023.
duplex antenna design, and the user effect of
[Online]. Available: https://www.keysight.com/us/en/assets/7018-
millimeter-wave handset antennas.
03631/data-sheets/5991-1103.pdf