This article has been accepted for publication in a future issue of this journal, but has not been

fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2019.2925638, IEEE Sensors
Journal
1

Energy Harvesting For Wearable Devices: A

Review
Yung-Wey Chong, Widad Ismail, Kwangman Ko, and Chen-Yi Lee

devices by 2021 [3]. Wearable devices are targeting personal

Abstract— In recent years, wearable devices have attracted
attention because of their ability to enhance the quality of life. This
disruptive technology has helped healthcare professionals with
intervening early in chronic diseases, especially amongst
independently living patients, and has facilitated real-time
monitoring of patients’ vital signs remotely. One of the major
bottlenecks that hampers the adoption of wearable device is
continuous power supply. Most wearable devices solely depend on
battery supply. When the energy stored in the battery is depleted,
the operation of wearable devices is affected. To overcome this
limitation, efficient energy harvesters for wearable devices are
crucial. The paper primarily aims to present a comprehensive
classification of different energy sources that can be capitalised to
power wearable devices. In addition, this research paper deals
with the key challenges that must be considered in the
development of autonomous wearable devices for telemedicine
applications with a proposed system design for wearable device
that uses energy harvesting technology.
Index Terms—Energy harvesting, wearable device, Internet of
Things, Telemedicine, Wireless Sensor Network, Energy Resource
I. INTRODUCTION
The ageing population has been growing at an unprecedented
rate worldwide owing to the improvements in quality of life.
United Nations estimated that by 2017, over 962 million people
worldwide would be over the age of 60 years. The number of
elderly people will further increase to 1.4 billion by 2030 as
fertility declines and life expectancy rises [1]. Many publicly
funded healthcare systems are facing massive increases in costs
and crowdedness in medical institutions owing to the need for
nursing care. Many of these patients are diagnosed with obesity
and diabetes, chronic diseases that can be prevented with
appropriate diet and sufficient physical activity [2]. Wearable
devices constitute a new technological paradigm for
encouraging physical activity. In addition, they can be used for
early detection of medical illnesses, emergency notification and
computer-assisted rehabilitation. The unprecedented growth in
wearable devices has attracted significant interest from the
commercial and research communities. ABI Research
forecasted that the market for wearable devices will increase
from 202 million devices in 2016 to more than 501 million
health sector [4] as well as the military sector for assessing the
physical and mental health of soldiers in combat [5].
Many researchers have worked towards developing wearable
devices that are small and unobtrusive and can improve users’
quality of life. Sonoda [6] and Poh [7] presented wearable
photoplethysmography (PPG) sensors that can be used at the
earlobe for heartbeat measurement. These optical techniques
contain sensor to measure the blood volume pulse that
propagates throughout the body to provide information about
the cardiovascular system. Vogt [8] acknowledged that
wearable devices must be developed for attracting fashionistas
and must be able to be used during magnetic resonance imaging
(MRI). As such, a Bluetooth Low Energy (BLE) wearable
device that can measure vital signs, e.g. heart rate, peripheral
capillary oxygen saturation and temperature, was conceived. In
addition to being used as standalone devices, sensors can be
embedded into garment to measure electrocardiogram (ECG),
e.g. MagIC [9], LifeShirt [10] and MAIN shirt [11]. Sensor
integration into garments helps prevent scenarios in which users
forget to put on wearable devices. This is crucial, especially if
caregivers/medical practitioners are monitoring elderly patients
or in the case of patients with critical illnesses. While garments
provide mobility for users, different factors, e.g. size of the
electronic component, power consumption and electrode
position, must be considered to ensure uninterrupted sensing.
Unlike wireless sensor networks, timely delivery of data
from a device to the cloud server is crucial because several
wearable devices are used in emergencies, such as detecting
falls and heart attack. Any delay may potentially cost a person
their life. Therefore, a continuous source of electrical power is
crucial to prevent interruption of the operation of wearable
devices [12]. However, energy limitations remain a critical
constraint because many wearable devices are powered solely
by battery energy. When battery energy is depleted, health
monitoring will be interrupted until the battery is charged or
changed. Wearable devices should be easy to use and require
low levels of maintenance. These devices should be able to
operate continuously with minimal human intervention [13].
The use of large-capacity batteries in wearable devices
increases the size and weight of these devices, making them
unattractive for use and causing discomfort to users [14]. An

The research project is funded by USM RUI Grant under grant number Kwangman Ko is with School of Computer Engineering, Sangji University,
1001/PNAV/8014078 (Corresponding author: Widad Ismail) Republic of South Korea (e-mail: kkman@sangji.ac.kr)
Yung-Wey Chong is with National Advanced IPv6 Centre, Universiti Sains Chen-Yi Lee is with the Department of Electronics Engineering, National
Malaysia, Penang 18000, Malaysia (e-mail: chong@usm.my). Chiao Tung University, Hsinchu, Taiwan (e-mail: cylee@si2lab.org).
Widad Ismail is with School of Electrical and Electronic Engineering,
Universiti Sains Malaysia, Nibong Tebal 14300, Malaysia (e-mail:
eewidad@usm.my)

1558-1748 (c) 2019 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2019.2925638, IEEE Sensors
Journal
2

Fig. 1. Typical architecture of a wireless body area network

ideal wearable device should be invisible or resemble a fashion
item.
To ensure user adoption, it is crucial that next-generation
power sources provide greater functional capabilities, as well as
improve device comfort and lifetime. Scavenging energy from
the environment is important from the viewpoint of wearable
devices. Doing so can reduce dependency on battery energy and
provide unique features that are not provided by conventional
batteries [15, 16]. The combination of wearable devices and
energy-harvesting technology has created fantastic
opportunities in wearable devices. IDTechEx reported that the
energy-harvesting market grew from $0.7 billion in 2012 to $
22.6 billion in 2014 [17]. The combination of low power
requirements, tiny devices that blend into the body and the need
to eliminate maintenance ensures that energy harvesting is
suitable for use in wearable devices.
II. ENERGY HARVESTING TECHNIQUES IN WEARABLE
DEVICES
Generally, wearable devices are composed of sensors for
monitoring vital signs; they have low-power processing
capabilities to process the vital sign signals so that they can be
visualised by healthcare providers. Such devices perform tasks
such as collecting signals from sensors, processing them and
storing and transmitting the processed information wirelessly.
They comprise sensors, a radio frequency (RF) transceiver and
a processing unit and are integrated into wearable structures
such as shoes, clothes, badges and bracelets, as shown in Fig. 1.
These devices can be categorised as invasive or non-invasive
depending on their placement in the human body. They are
usually small and thin, capable of wireless communication and
are characterised by low power consumption. Data from these
devices are sent to a data center through wireless
communication media so that the user can be monitored
regularly either by healthcare service providers or caregivers.
Energy harvesting can be performed sporadically to provide
continuous power to wearable devices [18, 19]. The most
promising sources of energy for wearable devices can be
classified into two broad categories: ambient sources and
external sources, as shown in Fig. 2. Ambient sources such as
solar power, thermoelectric power and radio frequency (RF) are
available in the surroundings at almost no cost. The
characteristics of ambient sources are unique in terms of
predictability, controllability and conversion efficiency [20].
This is because these sources may be affected by time, location
and weather conditions. In contrast, external sources are
predictable and controllable because they are deployed
explicitly in the environment.
A. Solar energy
For several decades, solar power has been used to power
consumer products such as calculators and wristwatches. It is
amongst the most widespread energy sources owing to the
magnitude of energy harvested. Photovoltaic (PV) cells, which
are nonlinear semiconductor devices, generate electricity when
the chemicals in them are exposed to solar radiation or light.
The efficiency of solar cells depends on the material used in
them. Three materials are typically used in PV cells:
amorphous, monocrystalline and polycrystalline silicon.
Amorphous silicon cells can be manufactured on flexible
surfaces, making them suitable for wearable devices. Although
the efficiency of the cells based on this material is the lowest
(around 10%), it is widely used to manufacture cells for use in
indoor applications because it is cheap and sensitive to low
Fig. 2. Energy sources for wearable devices

1558-1748 (c) 2019 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2019.2925638, IEEE Sensors
Journal
3

to 100 mW/cm2 of energy can be harvested with conversion

Fig. 3. Equivalent circuit of solar cell [22]
levels of natural light. By contrast, crystalline silicon is more
appropriate for outdoor applications because the cells based on
this material can achieve conversion efficiencies as high as
15%–20% under strong lighting [21].
Solar cells can be simplified into an equivalent circuitry, as
shown in Fig. 3 [22].
The current–voltage (I–V) characteristic of a solar panel can
be expressed as follows:
+,+ - .,+ /0
#,+ 67,+ 80
𝐼"# = 𝐼"% − 𝐼( )𝑒 12 +3 − 1 5 − 809
(1)
Vt, the junction thermal voltage can be calculated as follows:
<=>
𝑉; = ?
(2)
where IO is the dark saturation current of diode, IPH is the photo-
generated current, RSH is the panel parallel shunt resistance, RS
is the panel series resistance, A is the diode quality factor, k is
the Boltzmann constant, q is the electron charge, nS represents
the number of cells in the panel connected in series and T is the
temperature in Kelvin.
Equations (1) and (2) prove that the power generated by a
solar panel depends on environmental conditions. The
magnitude of electricity generated is directly proportional to the
size of PV cells and level of illumination. During daytime, up
efficiency between 15% and 20% [23]. The light density is
lower when PV cells are exposed to indoor lighting, and the
corresponding energy density is between 10 µW/cm2 and 100
µW/cm2, with a conversion efficiency of 8% [24].
The high conversion efficiency and long-term availability of
solar energy over extended duration makes it attractive for
wearable devices that need continuous monitoring. One
example is Soliband [25], a wearable device that can operate
uninterrupted so long as a light source is available. Soliband
consists of monocrystalline silicon solar cells, a lithium
polymer (LiPo) battery, supercapacitor, power management
unit and photoplethysmogram (PPG) sensor that monitors heart
rate and blood oxygenation, as shown in Fig. 4. When a light
source is not available, the device draws power from the
onboard 20 mAh LiPo battery and supercapacitors. The battery
can service the device’s power demand of 13.7 mW for 4 h in
the absence of a light source.
Flexible PV cells fabricated from amorphous silicon have
been employed to promote the use of solar energy in wearable
devices [26, 27, 29, 30]. Toh [26] developed an autonomous
wearable device powered by a flexible energy-harvesting
mechanism in conjunction with an ultra-low-power-
management circuit (PMC) on a flexible PCB to continuously
monitor body temperature and transmit the signal of this vital
sign to a base node. Excess solar energy is stored in the
supercapacitor. The device can operate for more than 15 h
without drawing power from the battery under indoor
conditions with light intensity of 320 lux. The use of flexible
PV cells has reduced the weight of wearable devices. Wu et al.
[30] further enhanced flexible wearable device by including a
heartbeat monitor and accelerometer to measure heartbeat and
detect falls. Unlike Toh [26], who used proprietary 2.4 GHz

Fig. 4. (a) Integrated PV cell in Soliband [25], (b) components of device for heartbeat monitoring using integrated PV cells [25], (c) use of flexible PV cells for
distributed biometric monitoring [26] and (d) flexible PV system [28]

1558-1748 (c) 2019 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2019.2925638, IEEE Sensors
Journal
4

Fig. 5. (a) Thermoelectric circuitry [33], (b) typical TEG [34] and (c) TEG arrangement on a wearable device [35]

frequency for wireless transmission, Wu [30] transmitted data
from the device to a mobile phone via standard Bluetooth low
energy (BLE).
Although solar/light energy sources are free, they are
uncontrollable and only partially predictable. The magnitude of
energy harvested is influenced by weather patterns and time of
the day. Because PV cells are non-linear devices, they have an
optimal operating point at which maximum power output is
generated. Hence, solar energy is usually employed in
conjunction with the maximum power point tracking (MPPT)
technology to maximise conversion efficiency. One of the
challenges associated with using MPPT in wearable devices is
ensuring that the power management unit consumes very little
power. Wu et al. [31] enhanced conventional MPPT by using
the output current to extract the maximum power to power pulse
sensors used in wearables. The proposed solar energy harvester
provided an efficiency of 66.5%, which is sufficient energy to
be supplied to pulse sensor and transmit data via BLE.
B. Thermoelectric energy
The concept of using thermoelectric to power wearables is
not new. Back in 1999, Seiko developed a thermoelectrically
powered wrist watch that used body heat [32]. Energy can be
harvested from thermal sources through the thermoelectric
effect by exploiting the Seebeck effect. According to the
Seebeck effect, thermoelectric energy generators (TEGs), a
property of thermoelectric materials, can be positioned on a
body to convert the temperature differential DT between the
body skin and the ambient environment into voltage.
The voltage characteristics of a TEG are as follows:
DV = −aDT (3)
where a is the Seebeck coefficient.
As shown in Fig. 5, a TEG consists of n-type material placed
in series with a p-type material. When temperature difference
across these materials occured, heat electrons move from the
hotter thermoelectric material towards the cooler material. In
the process, electric potential is formed, causing current to flow
in a closed circuit. A good thermoelectric material is usually
characterised by low thermal conductivity and high electrical
conductivity. Commonly used thermoelectric materials include
bismuth telluride, lead telluride and calcium manganese oxide.
To boost voltage output, multiple n-type and p-type
semiconductors can be connected in series to create a
thermopile by alternating the metal contact pads [36]. A DC-
DC converter with high conversion efficiency may be necessary
if the voltage output is too low [35], as shown in Fig. 6.
TEG is suitable for wearable device because the human body
converts energy into thermal energy that is dissipated to the
environment. A temperature difference of 5 °C to 10 °C
between the human body and the surrounding environment
makes TEGs suitable for wearable devices [37]. To convert
Fig. 6. System overview of a TEG-harvesting-driven wearable device [35]

1558-1748 (c) 2019 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2019.2925638, IEEE Sensors
Journal
Fig. 7. General RF energy-harvesting system
energy from body heat into electrical energy, thermopiles can
be integrated into clothing materials. Leonov [38] developed an
experimental shirt containing thermopiles by placing a round
hot plate measuring 3 cm in diameter in direct contact with skin.
A larger cold plate measuring 4 cm in diameter was glued to
carbon fabric. The system could generate approximately 5 mW
at ambient temperatures of 15 °C–27 °C. Because the thickness
of these materials was only 4 mm, the developed shirt was
comfortable for users. The authors highlighted that the
magnitude of energy harvested was unrelated to the user’s
metabolic rate but rather depended on the overall body heat
content and wind speed. Hoang [39] developed a thermal-
powered wearable device that included an accelerometer to
detect falls. Because the energy generated by the device was
520 µW at 15 °C, the produced voltage was used to charge a
capacitor and a power management unit was included to serve
as a bridge between the thermal source and sensor node.
C. Radio frequency energy
Harvesting energy from radio frequency (RF) is gaining
popularity because of the broad deployment of wireless
technologies. Unlike other energy sources in ambient
environment, RF energy can be harvested from both external
and ambient sources, as shown in Fig. 7. Ambient RF energy is
harvested from the electromagnetic waves generated by from
broadcasting and wireless communication infrastructure. The
power level the RF waves used for harvesting energy depends
on the distance between the transmitter and the receiver.
Because the distance between infrastructure units is large in
rural areas compared to that in urban areas, the power level in
rural area can be as low as –40 dBm.
As external RF energy sources, electromagnetic radiation
and magnetic resonance are used to generate electric power and
wirelessly transfer it from the source to the receiver within a
predefined wavelength margin. Resonant coupling using two
planar resonators can be performed in the case of implanted
devices. The primary resonator is placed outside the body,
1558-1748 (c) 2019 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
5
while the secondary resonator is connected to the implanted
device. Energy can be transferred via magnetic coupling
between the two resonators. The magnitude of energy harvested
is predictable and controllable, and the energy efficiency is 80%
[40]. Although the energy efficiency is higher compared to that
of ambient RF energy sources, external RF energy sources can
operate only at small distances because the output power of RF
sources is limited by government regulations [41].
To harvest RF energy, radiated electromagnetic waves are
collected using antennas and RF is converted into electric
energy by using rectifier circuits [42]. To achieve high
rectification efficiency, silicon Schottky barrier diodes are
widely used in RF energy harvesters. Although this diode has a
low built-in voltage, it is suitable for this application because a
greater number of harmonic signals is generated at higher
voltages owing to the nonlinear characteristics of the diode,
which decreases rectification efficiency [43]. RF energy-
harvesting performance is defined in terms of sensitivity and
conversion efficiency [44]. Conversion efficiency, ηRF, is the
ratio of output power that can be supplied to the sensor, Pout, to
the input power received by the RF energy harvester, Pin. A low
energy conversion efficiency is often caused by impedance
mismatch, parasitic effect and harmonic generation. In perfectly
matched antennas, ηRF can be expressed as
"BC3 G
#BC3EF / 8I
η8A = "D1
= "/J
, (4)
where VoutDC is the output power and RL is the resistive load
of the energy harvester. In non-linear circuitry, a RF energy
harvester will reflect a part of the power, which leads to
reduction of 𝑃LM . Sensitivity is the minimum power needed to
power the integrated circuit (IC). Sensitivity is a function of the
IC technology used in wearable devices.
In the past, many researchers have developed wearable
devices powered by RF energy. Mandal [45] designed a battery-
less wearable tag that could monitor users’ vital signs, such as
respiratory sounds, PPG, blood pressure and body temperature.
The wearable tag was powered by harvesting external RF
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2019.2925638, IEEE Sensors
Journal
Fig. 8. (a) Noninvasive tattoo-based biofuel cell that harvests biochemical
energy from lactate [54] and (b) implantable glucose fuel cell [92]
energy by using a two-stage complementary metal-oxide
semiconductor (CMOS) rectifier. An output power of 15 µW
was generated at a distance of 3.1 m from the transmitter
operating at 800 mW. Zhang [46] used a multistage rectifier to
convert 915 MHz RF energy to power 0.18-µm CMOS
technology. A similar implementation of the use of RF energy
harvesters in wearable devices can be found in Xia [47] and
Dini [48].
D. Human-powered energy
In addition to ambient energy or external RF energy source,
human-powered energy harvesting has attracted attention in
recent years [12, 49]. The idea was conceived because human
energy, which originates from food consumption, can be
transformed into chemical, thermoelectric and kinetic energy.
There is large potential for the use of human energy in wearable
devices considering that human energy can provide 35–100
times more energy than a battery [50]. In addition, the holy grail
of wearable devices is to develop biological implantable
devices that can monitor a user’s health without being intrusive.
This can be achieved using human-powered energy harvesters
or microgenerators that convert body motions, heat, or chemical
reactions into useful energy sources for wearable devices.
Energy harvesting from humans can be classified broadly into
two categories, namely, biochemical and biomechanical.
Biofluids inside the human body contain a variety of
substances and active enzymes [51]. Under certain conditions,
when chemical bonds are broken, energy is released.
Biochemical energy can be scavenged to provide the electrical
energy needed by ultra-low-power implantable sensors, as
shown in Fig. 8. The availability of harvestable chemical
sources depends on human age and human health and food
intake. A key biochemical energy source is biofuel cells. In
biofuel cells, power is generated via chemical reactions
between the anode and the cathode and a catalyst is used to
accelerate these chemical reactions. The anode oxidises a
chemical compound and reduce it to the cathode. An infinite
amount of energy can be harvested via this route so long as the
specific chemical substances exist in the body.
Glucose is one of the most common chemical substances that
can be extracted from body to generate electrical energy for
wearable devices. Glucose fuel cells can be classified based on
the catalytic schemes used in them, namely, enzymatic,
microbial and abiotic. Each of these catalytic schemes provides
different efficiency and robustness. An enzymatic glucose fuel
cell metabolises glucose into acid and releases electrons that
1558-1748 (c) 2019 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
6
Fig. 9. (a) Simple show-mounted rotary magnetic generator [61] and (b)
piezoelectric energy scavenging in shoes [62]
generate electrical energy. Up to 100 µW/cm2 of electrical
energy can be harvested using an enzymatic glucose fuel [52,
92]. While enzymatic glucose fuel cells can provide energy
continuously, changes in temperature and pH conditions may
disrupt or destruct the enzyme structure. A microbial glucose
fuel cell uses living microorganisms to oxidise glucose. It can
generate up to 1000 µW/cm2 of energy. One of the challenges
linked with the adoption of microbial glucose fuel cells is
stabilising the microbe population and preventing bacteria from
escaping the cell, which may cause infection in users [53].
Owing to safety issues, microbial glucose fuel cells have not
been adopted in implants. Abiotic glucose fuel cells oxidise
glucose by using a solid-state anode catalyst. Although this type
of catalyst is the least efficient amongst various catalysts,
producing less than 10 µW/cm2, it is reliable and biocompatible.
Lactate can be harvested from human bodily fluids, such as
saliva or sweat. Jia [54] designed a temporary transfer tattoo
that harvests biochemical energy from the lactate in the
perspiration generated during physical activity. The power
densities of this type of cell can be as high as 70 µW/cm2, which
shows the amount of energy released by the human body during
physical activity. Energy can be harvested from differences in
potential hydrogen (pH) levels using metal pH electrodes
(Au/IrO2) against reference electrodes (Ag/AgCl) [55]. Low
power voltage, which is proportional to the pH level, is input to
an electrostatic energy harvester to amplify the voltage.
Another energy source from the human body is endocochlear
potential (EP), an electrochemical gradient available in the
inner ear of mammals. Mercier [56] demonstrated that EP can
power an ultra-low-power implant inside the ear for up to 5 h.
However, owing to anatomical constraints, the amount of
energy that can be extracted ranges between 1.1 and 6.3 nW
only.
Another way in which the human body can be used as a
primary generator is through human movement. Biomechanical
activities such as heartbeats and foot strikes can be converted
into electrical energy to power low-energy wearable devices
Fig. 10. (a) Energy-harvesting generator installed on sheep’s heart [64] and
(b) electromagnetic energy from breathing [65]
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2019.2925638, IEEE Sensors
Journal
[57, 58]. Body muscles can generate motions and stop motions
[59]. When an energy-harvesting device is placed on the body,
energy lost to body motions can be absorbed and electrical
energy can be generated without interfering natural motions
[60]. Energy generation can be triggered by either voluntary or
involuntary. There are two types of biomechanical
microgenerators, namely, inertial (vibration-driven) generators
and force-driven generators.
In force-driven generators, electrical energy is generated by
converting mechanical energy by using direct force. At MIT
Media Lab, researchers developed running shoes that can
generate power by using a piezoelectric sole, heel generator and
electromagnetic generator as the shoes are pressed on the floor,
as shown in Fig. 9 [61, 62]. The results showed that the
piezoelectric sole can generate 2 mW of power, heel generators
can produce 8 mW and electromagnetic generator can produce
250 mW. Inertia-driven generators generate electrical energy
by using the inertia of a moving object within the generator
coupled with electromagnetic, electrostatic, or piezoelectric
mechanisms. The electromagnetic method was derived from
Faraday’s law of induction, which defines how changes in the
magnetic environment inside a wire coil induce voltage. The
electrostatic method results from the conservative Coulomb
force, according to which electrical energy is generated when
the distance between two charges in a capacitor changes,
leading to the generation of current in a circuit. Voltage is
produced by a piezoelectric generator when certain materials
are stressed.
Generator placement is an important factor for motion-
powered energy harvesters. Büren [63] compared the
performance of three different generators, namely, Coulomb-
damped resonant generator (CDRG), velocity-damped resonant
generator (VDRG) and Coulomb-force parametric generator
(CFPG), at nine positions on the body of a person walking on a
treadmill. The results of the experiment showed that the amount
of energy generated at lower body locations (hip, knee and
ankle) is four times higher than that generated at upper body
locations. Merrett [49] further investigated the effect of
harvestable energy from a human’s inertia by considering the
effects of additional parameters, such as generator orientation.
The results showed that the factors of rotation and misalignment
can significantly affect the amount of energy that can be
harvested.
Involuntary actions such as cardiac motion, blood pressure
and breathing can generate biomechanical energy, as shown in
Fig. 10. Unlike voluntary actions that are non-periodic,
involuntary actions can provide regular energy to wearable
devices, especially implants. Zurbuchen [64] used cardiac
contractions as a reliable and continuous source of energy to
power low-power pacemakers. Their energy harvester was
based on the mechanism of automatic watches, and it could
harvest 11.1 µJ of electrical energy when powered by constant
90 bpm heartbeat. Energy can be harvested from the breathing
process as well. The flow of wind from inhalation and
exhalation generates differences in air pressure, which in turn
can be scavenged to produce electrical energy by using small
wind turbines. This method is suitable for use in wearable
1558-1748 (c) 2019 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
7
Fig. 11. Typical block diagram of energy-harvesting-enabled wearable
device
electronic masks [65].
III. DESIGN CHALLENGES AND CONSIDERATIONS
The data rates of wearable devices are usually low [66]. For
example, the highest data rate required for heartbeat monitoring
is 100 samples/s. At a resolution of 10 bits, the data rate would
be as low as 1 kbps, consuming only 0.65 µW of transmitter
power [67]. With the unique requirements of wearables, one can
estimate their minimum power requirements based on the
power requirements of each module, namely, sensor,
transmitter and processing unit. This information can be used in
the design of energy-harvesting systems. Owing to the unstable
flow of harvested energy [20, 68], the converted energy may be
stored in an energy storage device such as a supercapacitor or a
battery, as shown in Fig. 11. By storing the energy in a storage
device, power can be supplied to the processing unit,
transceiver and other modules in a constant and stable manner.
Nevertheless, harvested energy may be used directly by the
microcontroller (MCU) in battery-less wearable devices.
In energy-harvesting-enabled wearables, the power
management unit (PMU) plays an important role in regulating
and maximising the power level. An efficient recharging
mechanism is required so that the lifetime of the energy storage
device can be extended. The PMU can opportunistically use the
period in which maximum energy is available and function
minimally when the ambient energy is the least. When
sufficient energy is available to drive the entire system, the
dependency on battery power will be reduced. Many studies
have been conducted to optimise the use of energy harvesting
in wearable devices. While this market vertical shares many
challenges with the more general market vertical of wireless
sensor networks (WSNs), there are many differences between
the two technologies. When designing and implementing
energy harvesting into wearable devices, several factors must
be considered, namely, the amount of energy that can be
harvested to make the system autonomous, efficiency of energy
storage device, choice of communication protocols and cost.
Fig. 12. Design challenges
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2019.2925638, IEEE Sensors
TABLE I
POWER DENSITY OF VARIOUS ENERGY SOURCES
Energy sources Power density
2
Ambient light 100 mW/cm (direct sun
100 µW/cm2 (indoor illumination)
Thermoelectric 60 µW/cm2
Radio frequency 1 µW/cm2 (ambient)
15 µW (external)
Human 1000 µW/cm2 (biochemical)
4 µW/cm3 (biomechanical - microgenerator)
200 µW/cm3 (biomechanical - piezoelectric)
While energy-harvesting techniques have been studied
extensively, open design challenges and considerations for
energy harvesting in wearable devices must be reflected upon.
These challenges include power density and management,
energy storage and network-based design, as shown in Fig. 12.
A. Power density and management
One of the major issues when designing energy-harvesting
systems is achieving adequate energy efficiency for minimum
power consumption overhead and for providing regular energy
flow. Although various energy sources are available, the
characteristics of each energy source are different, as
summarised in Table 1. Ambient sources may be affected by
time, location and weather conditions, leading to uncertainties
in terms of Quality of Service (QoS). To maximise the
advantages of energy harvesters, efficient power management
is crucial. During the design process, the weaknesses and
strengths of available energy sources must be considered.
Maximum power point tracking (MPPT) is a technique used
in solar-powered systems to maximise power extraction under
all conditions. The technique is often used for output power
higher than 100 mW. One of the challenges associated with
power management circuits for solar energy harvesters is
designing a MPPT circuit with minimum power consumption
overhead. Conventional MPPT methods cannot be applied
directly to solar energy harvesters in wearable devices because
the energy harvested from small PV cells is often lower than the
energy consumed by the power management unit. Instead, of
using MPPT, Brunelli [69] stated that an alternative simple and
low-cost method should be adopted. The authors proposed a
Fig. 13. Mechanically assisted start-up circuit to kickstart electrical energy
extraction from TEG [36]
1558-1748 (c) 2019 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Journal
8
Advantages Disadvantages
High power density; Non-continuous;
mature technology dependent on weather/lighting conditions
Widely available Low power density
Widely available Power output depends on distance between harvester and
RF energy source
Light weight Power output varies;
depends on activity
semi-MPPT design with two voltage comparators to control the
power usage of sensor nodes. One of the drawbacks of this
design is its dependency on a light source. When input power is
not available and capacitor voltage drops below the threshold
voltage, the MCU cannot operate. Fractional open circuit
voltage (FVOC)–based MPPT that uses small PV cells was
adopted by Jafer [70] and Toh [26]. In this method, fewer and
smaller components are used, thus reducing power
consumption and making its integrated with wearable devices
easier. This method can provide stable power supply with
MPPT efficiency of 81% at 0.5 mW, even under low light
intensity. Therefore, it is important to consider efficiencies
when MPPT is used in wearable devices.
Similar to solar energy, wearable devices powered by
thermoelectric energy require a certain threshold voltage for
start-up. This challenge can be overcome by deploying a
transformer in the system. Damschke [71] used a transformer to
provide start-up voltages of 20 mV to a system that uses thermal
energy. This is because temperature gradients may be
inadequate from the viewpoint of operating CMOS switches for
conventional boosting without battery. Nevertheless, the
transformer design was bulky and unsuitable for wearable
devices. Another method is the use of motion-activated
switches to assist the start-up circuit. Ramadass [36] proposed
a mechanically assisted start-up circuit by using a motion-
activated switch, as shown in Fig. 13. Small vibration from
human motion turns on the switch and powers the start-up
block.
In addition to circuit design, physical integration of energy-
harvesting wearable devices affects their efficiency. For
example, flexible PV cells that bend according to the contour
shape of the human body provide better efficiency than non-
flexible PV cells. The bending angle exerted on the PV cells
Fig. 14. Illustration of a PV panel bent by 30° [26]
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2019.2925638, IEEE Sensors
Journal
affects the electrical power harvested from a light source owing
to differences in light intensity, as shown in Fig. 14 [26]. For
TEG-powered wearable devices, researchers must consider
device structure, body heat, air speed surrounding the person,
human anatomy and device placement on the human body [38].
This is because the thermal interfaces between skin and the
TEG influence power density [35]. To obtain the maximum
power for wearable devices, TEG interfaces should be large
enough for better thermal matching without obstructing the
human’s movement. The use of heat sinks can help dissipate the
heat traversing the thermal harvester into ambient air, creating
a steeper temperature gradient. Nevertheless, the choice of heat
sink must match the ambient temperature of the wearable
device to avoid cooling of the skin [72]. In addition, the
materials used to fabricate TEGs must be able to withstand
continuous thermally induced stresses.
Although energy can be harvested from the human body, the
magnitude of this energy is very low. In addition, there are
tradeoffs between biomechanical energy and thermoelectric
energy sources. For a certain condition, thermoelectric devices
are effective in converting 70% of their energy to electrical
energy, whereas biomechanical devices can only convert 1% of
their energy to electrical energy [73]. The techniques for
exploiting biochemical energy are limited and immature. The
magnitude of energy harvested is very small, and in cases
wherein enzymatic compounds are used, the lifespan of the
energy harvester is limited [51]. Chemical compounds degrade
with age and health, which may have adverse effects on human
health.
Nevertheless, the challenges associated with biochemical
energy harvesting can be eliminated by using RF energy
harvesters. Wirelessly powered implants can safely provide
power to implants that help mitigate peripheral artery disease
(PAD) restenosis [74]. One of the factors that can influence the
power density of RF energy harvesters is the choice of antenna
and frequency band [75]. Wearable devices based on RF energy
harvesters can be designed with either one or two radios. In dual
radio designs, the RF harvester is separated from the module for
wireless communication with other nodes. Alternatively, single
radio designs can be used in wearable device to reduce device
footprint. Multi-band or wideband rectennas (rectifying
antenna) can be used to enhance the energy conversion
efficiency. Dini [48] developed a battery-less tri-band textile-
based wearable rectenna that harvests energy from the GSM
900 and 1800 bands, as well as the WiFi band. Because the
system is battery-less, a start-up circuit containing an external
inductor and an input capacitor is required. Kuhn [76]
introduced a wideband RF energy harvester for outdoor
wearable devices that operates between 900 MHz to 2.45 GHz.
The relationship between output voltage (VoutDC) and
conversion efficiency (ηRF) was investigated considering
several rectifier topologies. An interesting finding was that the
output voltage of double rectenna was 20% higher than that of
a single frequency system.
Energy harvested from ambient environment may be
discontinuous, thus impacting the operation of wearables. If the
power consumption of wearables is more than the power
1558-1748 (c) 2019 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
9
harvested from the environment, device operation will be
discontinuous and the time between periods of operation will
be determined by the energy stored in the device [77].
Therefore, device voltage level must be adjusted to replenish
the energy for ensuring continuous device operation.
B. Power consumption
Power consumption of the wearable device can influence the
design decision of the energy harvesters. Besides exploring
power density and management of energy harvesters, it is
necessary to minimise total power consumption in wearable
device in order to achieve energy-neutral operation [100]. This
is to ensure that the power output from energy source are
sufficient to support power consumption as shown Fig. 15. The
power output from energy source Ps must be greater than power
consumption by embedded device Pc to ensure the operation of
wearable device. One method to address power consumption
issue in wearable device is using software. Park et al. [98]
created a low-power ECG monitoring system that consume
maximum 30 mW while in operation. Marzencki et al. [99]
reduced the energy consumption to 16.6 mW by optimization
of data processing algorithm. To reduce the energy
consumption of the processing unit, ECG data is obtained after
applying low- and high-pass filters to eliminate noise, muscle
artifacts and interference.
Another method to conserve the energy in wearable device is
reducing radio transceiver’s power consumption since radio
transceiver is the most power-hungry block. Numerous
approaches have been proposed such as duty cycle optimisation
[95, 96, 97, 102] and ultra-low-power wake-up radios [94]. In
duty cycling approaches, energy consumed by data
transmission and processing circuitry are reduced by optimising
the active periods of wearable devices. Wearable device can
switch to sleep mode when it is not sensing health data to
overcome energy limitations [101]. Hsu [97] developed an
adaptive duty cycling algorithm that change the duty cycle
based on predicted availability of solar energy. The prediction
model is based on Exponentially Weighted Moving Average
(EWMA) that exploit diurnal cycle in solar energy for energy-
neutral operation. Shigeta et al. [102] implemented duty cycle
control through adaptive optimal operation point tracking. The
proposed approach considers risk of energy shortage and
capacitor leakage when calculating the optimal stored energy
level so that it can be used when energy source is unstable. The
Fig. 15. Power output from harvested energy versus power consumption for
wearable device (a) without storage device and (b) with storage device [68]
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2019.2925638, IEEE Sensors
TABLE II
COMPARISON AMONGST DIFFERENT ENERGY STORAGE MECHANISMS FOR ENERGY HARVESTERS IN WEARABLE DEVICES [87]
Battery-less
Storage mechanism Direct connection with energy
harvester without storage mechanism
Power limitation Limited by amount of energy
harvested
Energy storage None
Charge rate Not affected
Cycle lifetime limitations Not affected
proposed adaptive duty cycle control reduced the energy
leakage by 5.56%.
Pattern of active and sleep mode has to be designed carefully
so that delivery of health information will not be affected. In
addition, energy savings may not be optimal if wearable device
need to wake up regularly because transition from sleep to
active state requires energy as well. Jelicic [94] proposed a two-
tier architecture with wake-up radio to reduce response latency.
The wake-up radio consists of low-power transmitting and
receiving circuitry with event-driven wake-up rules. Since the
choice of microcontroller/microprocessor, sensors and radio
transceiver may affect the power consumption, wearable device
designers need to tune their application based on power
consumption information so that it can operate on limited power
reserves [93].
C. Energy storage
Harvested energy can either stored to energy storage devices
based on the magnitude of output power and impedance before
being supplied to the MCU or used directly to power the MCU,
as shown in Fig. 16. The choice of whether to use a storage
device in wearable devices is influenced by several factors,
which are as follows:
1) Location of wearable device: either outside body or
Fig. 16. Design of wearable devices with and without energy storage device
1558-1748 (c) 2019 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Journal
10
Battery Supercapacitor
Chemical Physical
Reaction kinetics and mass transport Electrolyte conductivity
High Limited
Kinetically limited High; same as discharge
Mechanical stability and chemical Side reactions
reversibility
implant
2) Type of energy source
3) Usage of wearable device: either it is a critical
application that requires continuous power supply or a
non-critical healthcare application that does not require
such a power supply, such as a fitness tracker.
Although battery-less wearable devices are ideal, any
disruption in the source from which energy is harvested will
interrupt the service provided by the wearable device. In
addition, energy conversion may not be optimised and high
energy reliability of the device cannot be ensured without
suitable power conditioning. As such, energy storage is
important to ensure smooth power supplied to the load and to
serve as a reserve during periods in which ambient/external
energy is unavailable. Unlike other electronic devices, energy
storage for wearable devices must meet several requirements.
First, the energy storage device much be rechargeable because
replacing batteries can be problematic. For example, medical
professionals are required to perform surgery every 8 years to
replace the lithium battery in pacemakers [78]. Second, the
energy storage mechanism must be able to support long-lasting
applications with minimal effect on battery parameters, such as
round-trip efficiency and self-discharge. Two major choices are
available for energy storage, namely, batteries and
supercapacitors. The criteria for selecting the type of energy
storage depend on device lifetime and recharge cycles.
Batteries modify energy from electrochemical cells into
electrical energy, whereas capacitors store energy in the form
of an electric field. Rechargeable nickel-cadmium batteries
were first used in pacemaker implants in 1958. The battery cell
was charged inductively and wirelessly. Since then, battery
technology has evolved, and the use of implanted pacemakers
has increased. In the mid 1970s, lithium ion batteries greatly
Fig. 17. Electric double-layer capacitors
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2019.2925638, IEEE Sensors
Journal
extended the pacemaker battery life owing to the high energy
and power density of these batteries. Typically, the maximum
number of recharge cycles of Li-ion batteries is 1000 over 7
years [79]. Unlike other compounds, the lithium metal surface
is passivated by reaction with an electrolyte. In lithium-ion
batteries, the anode and the cathode are stacked together with a
polymer separator and the entire assembly is rolled and
encapsulated within metal-laminated pouches or rigid plastic
containers under compression.
To cater to wearable devices, flexible batteries have been
developed for the storage of smaller amounts of energy. Single
layers of anode, separator and cathode are piled together and
sealed using flexible encapsulation. Such flexible batteries
coupled with energy harvesters can eliminate the need for
frequent charging using a bulky battery charger [80]. Owing to
the proliferation of wearable devices, reinforcement, especially
current collectors, are used to increase battery capacity and
mechanical flexibility. Materials such as carbon nanotubes [81],
conductive fabrics [82] and graphene foams [83] have been
used as current collectors to improve the flexibility of entire
cells. In addition, the mechanical integrity of active layers has
been improved by deploying supporting active particles within
porous membranes, carbon nanotubes and graphene networks
or by synthesising active materials directly on conductive
supports. While such designs provide the flexibility that is
much needed in wearable devices, the conductivity of flexible
current collectors is very low, which limits the maximum
operating discharge rate [27]. Another disadvantage of using
battery technology is the possibility of leakage, leading to
chemical poisoning, especially when used in implants.
Repeated overcharging and high temperatures (above 60 °C)
may damage the batteries, resulting in chemical leakage. The
leaked chemicals may cause chemical burns or other health
issues in the human body.
The limitations of batteries make supercapacitors an
attractive alternative as energy storage mechanism. A
supercapacitor is an energy storage device that employs thin
dielectric layers and electrodes with large surface area. It stores
electrical energy at an electrode-electrolyte interface, as shown
in Fig. 17. It has high pulse power capability, which means it
can handle short duration power surges. If a wearable device
must have a small form factor and function for extended
durations without failure, a battery may not be a suitable
economic choice. Supercapacitors are classified broadly into
three categories based on the energy storage mechanism:
electrical double layer capacitors (EDLCs), hybrid
supercapacitors and pseudo-supercapacitors [84]. In EDLC, the
energy storage and release mechanism is based on nanoscale
charge separation at the electrochemical interface between an
electrolyte and an electrode [85]. It contains nanoporous
materials as active electrode materials with huge capacitance
and relatively long cycle lifetimes. Pseudocapacitors are based
on Faradaic redox reactions involving high-energy electrode
materials such as metal oxides, metal-doped carbons and
conductive polymers (Stoller et al. 2008). These electrode
materials have much higher energy density at the price of
shorter cycle lifetime compared to EDLCs. Hybrid
1558-1748 (c) 2019 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
11
supercapacitors offer the features of EDLC and
pseudocapacitors by combining the electrodes of EDLC and
pseudocapacitors.
Similar to battery technology, supercapacitor technology has
evolved tremendously owing to the use of carbon materials such
as graphene to provide flexibility. Various electrode materials
and supercapacitor configurations can increase their energy
density to same level as that of a battery. For example, graphene
is a better conductor of electricity than copper and 200 times
stronger than steel. Flexible graphene-based supercapacitors
can be easily woven into textiles providing more promising
solutions for wearable devices [86]. Supercapacitors can
maintain a long cycle lifetime with minimal change in
performance. The lifetime of a supercapacitor is 10–20 years,
and its capacity might reduce from 100% to 80% after 10 years,
making it suitable for use in implants. The low equivalent series
resistance (ESR) allows supercapacitors to provide high power
density and high load currents, which means they can be
charged within a few seconds. Comparison amongst different
energy storage mechanisms for energy harvesters in wearable
devices is shown in Table 2.
D. Network design
The network design of wearable devices is important because
some designs can exploit the presence of harvestable energy
sources and improve performance parameters. Several aspects
can be considered in network design, namely, wireless
communication technology, network protocols and network
architecture. Because the mobility requirement is important for
wearables, wireless communication technology is crucial to
ensure energy efficiency at the device level. To ensure mobility,
transmitters such as Ultra-Wide Band (UWB), ZigBee, cellular
and BLE are frequently used in wearables. The choice of
wireless communication protocol depends on the
interoperability offered by the standards. Although these
transmitters operate at a low duty cycle, it is vital that the
wireless connection between wearable devices and the
gateway/smartphone is stable so that data can be transmitted to
cloud servers for further processing. In a few 5G wearables, the
mmW band is used to direct a narrow beam to ensure energy
efficiency at the device level. Massive Multiple Input/Multiple
Output (MIMO) has been used to preserve Quality of Service
(QoS). The human body has a relatively large area, enabling the
deployment of a large number of wearable antenna elements
[88]. The spectral efficiency of wearables is critical owing to
the limited power resources available in wearables.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2019.2925638, IEEE Sensors
Journal
12

At the network protocol level, specialised protocols for

transmitting the data inherent to medical sensing may help
reduce energy costs. This is because MAC is responsible for
improving the reliability and performance of network
communication, as well as maximising its energy efficiency. In
a few wearables, data may be sent to an intermediary node
before reaching the cloud servers. During the transit, multi-hop
routing occurs. The challenge is to design an energy-aware
routing that can compute an optimal routing based on the
available energy. The routing protocol must consider the
mobility challenge that causes channel fading and increases the
bit error rate (BER) owing to interference with other wearables.
Zeng [89] proposed the GREES protocols in which the routing
decisions are made locally at each node by considering wireless
link quality, residual battery energy level of the node, energy
consumption and energy-harvesting rate. Voigt [90] modified
directed diffusion to incorporate information about the energy
source for wireless sensor nodes, i.e. whether they are running
on ambient energy or battery. The results showed that solar-
aware sensor nodes can provide significant energy saving.
Time synchronisation plays an important role in ensuring that
there is no time discrepancy between sensors. To synchronise
data, the energy cost of time synchronisation must be Fig. 19. Voltage obtained from (a) RF harvester and (b) solar energy
considered. Because the heartbeat rhythm is available in every harvester
living human, the rhythm can be used to perform time • How to provide regular energy flow without affected
synchronisation [91]. Biosensors in a body area network (BAN) by environment condition
can extract a heartbeat rhythm by detecting waveform peaks.
• How to minimise total power consumption
The peak sequences of the heartbeat are naturally synchronised
• How to ensure user’s mobility especially in hospital
between sensors because they are attached to the same source,
condition.
which is the heartbeat. When using the rhythm of a heartbeat,
As shown in Fig. 18, eMeD architecture is separated into four
biosensors do not need additional energy to receive periodic
subsystems namely; low-power microcontroller, multi-sensor
timing information from a central controller. As such, the
subsystem with event-driven sensor management scheme,
lifetime of wearables can be prolonged.
ZigBee-based communication module and hybrid energy
harvester. To achieve perpetual operation, hybrid energy
IV. SYSTEM DESIGN
harvesting module is used to collect and store the maximum
Although wearable devices have been widely amount of energy from the environment without affected by
commercialised, research and development are still ongoing environment condition. When solar energy is sufficient, it will
due to power management issue. The design of wearable device be used to charge supercapacitor. However, when solar energy
needs to be optimised in order to motivate people to use supply is unavailable during the night, RF energy is used to
wearable device to regularly monitor their vital sign. Based on operate the wearable device. Such design can reduce the
the design challenges and consideration, eMeD is introduced as dependency on single energy source and the needs to manually
self-sustainable wearable device that sense, process and recharge the storage device. ADP5090, an ultra-low power
transmit vital sign data via ZigBee. eMeD is proposed to boost converter, is used as power management unit. The
respond to related questions when designing an energy internal cold-start circuit in ADP5090 allows the PMU to
harvesting wearable device such as: operate at an input voltage as low as 380 mV. Ambient RF
energy converted to DC using Powercast P2110. The RF energy
harvester has very high conversion efficiency and can generate
charging voltage of 3.3V. The RF energy harvester enable
wearable device to be charged even when user is inside the
building. When user is outside the building, solar energy will
be used. MPPT algorithm is implemented in ADP5090 to
optimise charging process by sensing open circuit voltage from
the solar panel, calculating optimized charging voltage and uses
boost converter to generate the necessary charging voltage. The
charging voltage for RF and solar energy harvester is shown in
Fig. 18. Hybrid energy harvester
Fig. 19.

1558-1748 (c) 2019 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2019.2925638, IEEE Sensors
Journal
Fig. 20. Sensor management scheme
ZigBee communication module is used to provide low power
wireless connectivity to an IoT gateway. Unlike power-hungry
Bluetooth communication module, ZigBee provides a low-
power low cost solution for wearable device, making it suitable
when power resources are limited. In addition, ZigBee is more
suitable to be used in Internet of Medical Things (IoMT)
environment because it supports star, tree and mesh
networking. In hospital, doctors and nurses can obtain patient’s
information easily because each wearable device can send data
to IoT gateway at the same time.
The execution of task is optimised using event-driven sensor
management scheme. It is used to ensure accurate monitoring
and maximise power efficiency in eMeD. eMeD utilises
accelerometer to automatically recognise user’s state before
capturing vital sign data. The benefit for such design is twofold.
Firstly, the sensor management scheme assigns the minimum
set of sensors at any particular time as required and invokes new
sensors when state transitions happen. Heart beat and
temperature sensor are activated when accelerometer detects
that user is in static mode as shown in Fig. 20. This can prolong
the lifetime of the energy storage. Secondly, human activity
recognition can fuse data from multiple sensors in managing the
sensor and provide accurate reading. The processor operates in
active mode, but it will not activate the temperature and heart
beat sensor because human movement may affect the reading
of heart beat. The combination of low power processor and
communication module, hybrid energy harvester and event-
driven sensor management scheme will be able to prolong the
lifetime of wearable device.
V. CONCLUSIONS
While energy harvesting provides a solution for autonomous
wearable devices, other factors must be considered to design
1558-1748 (c) 2019 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
13
optimal wearables with a long lifespan. Apart from selecting a
suitable energy source, power and communication
requirements, routing protocols and energy storage must be
considered when designing wearable devices. Wearable devices
that offers a seamless experience to users will drive the adoption
of wearable devices, hence improving wellness, especially in
ageing populations. In this paper, eMeD is proposed to provide
an autonomous wearable device that can be used in IoMT
application. The system is designed to minimise power
consumption and maximise energy that can be harvested from
ambient environment.
REFERENCES
[1] Department of Economic and Social Affairs Population Division. World
Population Prospects: Key findings & advance tables. United Nations,
New York, USA, 2017.
[2] Ghamari M., Janko B., Sherratt R. S., Harwin W., Piechockic R.,
Soltanpur, C. A Survey on Wireless Body Area Networks for eHealthcare
Systems in Residential Environments. Sensors 2016; 16(6), 1–33.
[3] ABI Research. Wearable Device Market Share and Forecasts. 2016.
Retrieved April 10, 2018, from
https://www.abiresearch.com/press/wearables-workforce-enterprise-
device-shipments-wi/
[4] Aileni R. M., Valderrama A. C., Strungaru R. Wearable Electronics for
Elderly Health Monitoring and Active Living. Ambient Assisted Living
and Enhanced Living Environments (1st ed.). Elsevier Inc; 2017.
[5] Hoyt R. W., Reifman J., Coster T. S., Buller M. J. Combat Medical
Informatics: Present and Future, In: AMIA Annual Symposium
Proceedings; 2002; p. 335-339.
[6] Sonoda K., Kishida Y., Tanaka T., Higuchi K. Wearable
Photoplethysmographic Sensor System with PSoC Microcontroller, In:
2012 Fifth International Conference on Emerging Trends in Engineering
and Technology; 2012; p. 61–65.
[7] Poh M.-Z., Kim H., Goessling A., Swenson N., Picard, R. Cardiovascular
Monitoring Using Earphones and a Mobile Device. IEEE Pervasive
Computing 2012; 11(4); 18–26.
[8] Vogt C., Reber J., Waltisberg D., Lars B., Marjanovic J., Niko M., Klaas
P.P., Troster, G. A wearable Bluetooth LE sensor for patient monitoring
during MRI scans, In: 2016 IEEE 38th Annual International Conference
of the Engineering in Medicine and Biology Society (EMBC); 2016; p.
4975–4978.
[9] Di Rienzo M., Meriggi P., Rizzo F., Castiglioni P., Lombardi C., Ferratini
M., Parati, G. Textile technology for the vital signs monitoring in
telemedicine and extreme environments. IEEE Transactions on
Information Technology in Biomedicine 2010; 14(3); 711–717.
[10] Wilhelm F. H., Roth W. T., Sackner M. A. The LifeShirt. Behaviour
Modification 2003; 27(5); 671–691.
[11] Teichmann D., Kuhn A., Leonhardt S., Walter M. The MAIN Shirt: a
textile-integrated magnetic induction sensor array. Sensors (Basel,
Switzerland) 2014; 14(1); 1039–1056.
[12] Shaikh F. K., Zeadally S. Energy harvesting in wireless sensor networks:
A comprehensive review. Renewable and Sustainable Energy Reviews
2016; 55; 1041–1054.
[13] Hanson M. A., Powell Jr H. C., Barth A. T., Ringgenberg K., Calhoun B.
H., Aylor J. H., Lach J. Body Area Sensor Networks: Challenges and
Opportunities. IEEE Computer 2009; 42(1); 58–65.
[14] Dionisi A., Marioli D., Sardini E., Serpelloni M. Autonomous Wearable
System for Vital Signs Measurement With Energy-Harvesting Module.
IEEE Transactions on Instrumentation and Measurement 2016; 65(6);
1423–1434.
[15] Wang Z. L., Wu W. Nanotechnology-enabled energy harvesting for self-
powered micro-/nanosystems. Angewandte Chemie - International
Edition 2012; 51(47); 11700–11721.
[16] Paradiso J. A., Starner T. Energy scavenging for mobile and wireless
electronics. IEEE Pervasive Computing 2005; 4(1); 18–27.
[17] Zervos H, Harrop P., Das R. Energy Harvesting and Storage for Electronic
Devices: 2014-2024: Forecasts, technologies, players, IDTechEx Report
2014.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2019.2925638, IEEE Sensors
Journal
[18] Harb A. Energy harvesting: State-of-the-art. Renewable Energy 2011;
36(10); 2641–2654.
[19] Zhou G., Huang L., Li W., Zhu, Z. Harvesting Ambient Environmental
Energy for Wireless Sensor Networks : A Survey. Journal of Sensors
2014; 2014; 1–20.
[20] Ku M., Li W., Chen Y., Ray Liu K. J. Advances in Energy Harvesting
Communications: Past , Present , and Future Challenges. IEEE
Communications Surveys & Tutorials 2016; 18(2); 1384–1412.
[21] Sim Z. W. Radio Frequency Energy Harvesting for Embedded Sensor
Networks in the Natural Environment. Manchester, UK: The University
of Manchester; 2012.
[22] Sera D., Teodorescu R., Rodriguez P. PV panel model based on datasheet
values. In: IEEE International Symposium on Industrial Electronics;
2007; p. 2392–2396.
[23] Roundy S., Steingart D., Frechette L., Wright P., Rabaey J. Power sources
for wireless sensor networks. In: European workshop on wireless sensor
networks; 2004; p. 1–17.
[24] Randall J. F. Designing Indoor Solar Products: Photovoltaic Technologies
for AES; 2006.
[25] Dieffenderfer J. P., Beppler E., Novak T., Whitmire E., Jayakumar R.,
Randall C., Qu W., Rajagopalan R., Bozkurt, A. Solar powered wrist worn
acquisition system for continuous photoplethysmogram monitoring. In:
36th Annual International Conference of the IEEE Engineering in
Medicine and Biology Society; 2014; p. 3142–3145.
[26] Toh W. Y., Tan Y. K., Koh W. S., Siek L. Autonomous wearable sensor
nodes with flexible energy harvesting. IEEE Sensors Journal 2014; 14(7);
2299–2306.
[27] Ostfeld A. E., Gaikwad A. M., Khan Y., Arias A. C. High-performance
flexible energy storage and harvesting system for wearable electronics.
Scientific Reports 2016; 6:26122.
[28] Ostfeld A. E., Arias A. C. Flexible photovoltaic power systems :
integration opportunities , challenges and advances. Flexible Print
Electronics 2017; 2.
[29] Kartsch V., Benatti S., Mancini M., Magno M., Benini L. Smart Wearable
Wristband for EMG based Gesture Recognition Powered by Solar Energy
Harvester. In: 2018 IEEE International Symposium on Circuits and
Systems; 2018; p. 1–5.
[30] Wu T., Wu F., Redoute J.-M., Yuce M. R. An Autonomous Wireless Body
Area Network Implementation Towards IoT Connected Healthcare
Applications. IEEE Access 2017; 5; 11413–11422.
[31] Wu T., Arefin S., Redout J.-M., Yuce M. R. A Solar Energy Harvester
with an Improved MPPT Circuit for Wearable IoT Applications Design
of the Pulse Sensor. In: BODYNETS; 2016; p. 166–170.
[32] Kawata M., Takakura A. Thermoelectrically powered wrist watch. US
Patent US5889735A; 1999.
[33] Siddique A. R. M., Mahmud S., Heyst B. Van. A review of the state of
the science on wearable thermoelectric power generators (TEGs) and their
existing challenges. Renewable and Sustainable Energy Reviews 2017;
73; 730–744.
[34] Misra V., Bozkurt A., Calhoun B., Jackson T., Jur J., Lach, J., Lee B.,
Muth J., Oralkan O., Ozturk M., Trolier-Mckinstry S., Vashaee D.,
Wentzloff D., Zhu Y. Flexible technologies for self-powered wearable
health and environmental sensing. Proceedings of the IEEE 2015; 103(4);
665–681.
[35] Thielen M., Sigrist L., Magno M., Hierold C., Benini, L. Human body
heat for powering wearable devices: From thermal energy to application.
Energy Conversion and Management 2017; 131; 44–54.
[36] Ramadass Y. K., Chandrakasan A. P. A Battery-Less Thermoelectric
Energy Harvesting Interface Circuit With 35 mV Startup Voltage. IEEE
Journal of Solid-State Circuits 2011; 46(1); 333–341.
[37] Mateu L., Codrea C., Nestor L., Pollak M., Spies P. Human Body Energy
Harvesting Thermogenerator for Sensing Applications. In: 2007
International Conference on Sensor Technologies and Applications; 2007;
p. 366–372.
[38] Leonov V. Thermoelectric Energy Harvesting of Human Body Heat for
Wearable Sensors. IEEE Sensors Journal 2013; 13(6); 2284–2291.
[39] Hoang D. C., Tan Y. K., Chng H. B., Panda S. K. Thermal energy
harvesting from human warmth for wireless body area network in medical
healthcare system. In: Proceedings of the International Conference on
Power Electronics and Drive Systems; 2009; p. 1277–1282.
[40] Kim S., Vyas R., Bito J., Niotaki K., Collado A., Georgiadis A., Tentzeris
M. M. Ambient RF energy-harvesting technologies for self-sustainable
standalone wireless sensor platforms. Proceedings of the IEEE 2014;
102(11); 1649–1666.
1558-1748 (c) 2019 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
14
[41] Lu X., Wang P., Niyato D., Kim D. I., Han Z. Wireless Networks with RF
Energy Harvesting: A Contemporary Survey. IEEE Communications
Surveys & Tutorials 2014; 17(2); 1–34.
[42] Krikidis I., Timotheou S., Nikolaou S., Zheng G., Ng D. W. K., Schober
R. Simultaneous Wireless Information and Power Transfer in modern
communication systems. IEEE Communications Magazine 2014; 52(11);
104–110.
[43] Akkermans J. A. G., Van Beurden M. C., Doodeman G. J. N., Visser H.
J. Analytical models for low-power rectenna design. IEEE Antennas and
Wireless Propagation Letters 2005; 4(1); 187–190.
[44] Valenta C. R., Durgin G. D. Harvesting wireless power: Survey of energy-
harvester conversion efficiency in far-field, wireless power transfer
systems. IEEE Microwave Magazine 2014; 15(4); 108–120.
[45] Mandal S., Turicchia L., Sarpeshkar R. A low-power, battery-free tag for
body sensor networks. IEEE Pervasive Computing 2010, 9(1), 71–77.
[46] Zhang X., Jiang H., Zhang L., Zhang C., Wang Z., Chen X. An Energy-
Efficient ASIC for Wireless Body Sensor Networks in Medical
Applications. IEEE Transactions on Biomedical Circuits and Systems
2010; 4(1); 11–18.
[47] Xia L., Cheng J., Glover N. E., Chiang P. 0.56 V, -20 dBm RF-powered,
multi-node wireless body area network system-on-a-chip with harvesting-
efficiency tracking loop. IEEE Journal of Solid-State Circuits 2014;
49(6); 1345–1355.
[48] Dini M., Filippi M., Costanzo A., Romani A., Tartagni M., Del Prete M.,
Masotti D. A fully-autonomous integrated RF energy harvesting system
for wearable applications. In: EuMC 2013: 43rd European Microwave
Conference; 2013; p. 987–990.
[49] Merrett G. V, Huang H., White N. M. Modeling the Effect of Orientation
on Human-Powered Inertial Energy Harvesters. IEEE Sensors Journal
2015; 15(1); 434–441.
[50] McArdle W. D., Katch F. I., Katch V. L. Exercise Physiology: Energy,
Nutrition, and Human Performance. 5th Edition Lippincott, Williams &
Wilkins; New York; 2001.
[51] Akhtar F., Rehmani M. H. Energy Harvesting for Self-Sustainable
Wireless Body Area Networks. Pervasive Computing 2017; 32-40.
[52] Rapoport B. I., Kedzierski J. T., Sarpeshkar R. A glucose fuel cell for
implantable brain-machine interfaces. PLoS ONE 2012; 7(6); 1–15.
[53] Nguyen Van H. C. Implantable, miniaturized microbial fuel cell.
Technical Report, University of California Berkeley; 2012.
[54] Jia W., Valdés-Ramírez G., Bandodkar A. J., Windmiller J. R., Wang, J.
Epidermal biofuel cells: Energy harvesting from human perspiration.
Angewandte Chemie - International Edition 2013; 52(28), 7233–7236.
[55] He C., Arora A., Kiziroglou M. E., Yates D. C., O’Hare D., Yeatman E.
M. MEMS energy harvesting powered wireless biometric sensor. In: 2009
Sixth International Workshop on Wearable and Implantable Body Sensor
Networks; 2009; p. 207–212.
[56] Mercier P. P., Lysaght A. C., Bandyopadhyay S., Chandrakasan A. P.,
Stankovic K. M. Energy extraction from the biologic battery in the inner
ear. Nature Biotechnology 2012; 12; 1240–1243.
[57] Longhan X., Cai M. Human Motion: Sustainable Power for Wearable
Electronics. Pervasive Computing 2014; 17; 42–49.
[58] Network E. H. Energy Harvesting from Human Power. Engineering and
Pysical Sciences Research Council; 2011.
[59] Winter D. A. Biomechanics and Motor Control of Human Movement.
John Wiley & Sons; 2009.
[60] Riemer R., Shapiro A. Biomechanical energy harvesting from human
motion: theory, state of the art, design guidelines, and future directions.
Journal of NeuroEngineering and Rehabilitation 2011; 8(22).
[61] Kymissis J., Kendall C., Paradiso J., Gershenfeld N. Parasitic power
harvesting in shoes. In: Second IEEE International Conference on
Wearable Computing; 1998; p. 2–9.
[62] Shenck N. S., Paradiso J. A. Energy scavenging with shoe-mounted
piezoelectrics. IEEE Micro 2001; 21(3); 30–42.
[63] Von Büren T., Mitcheson P. D., Green T. C., Yeatman E. M., Holme A.
S., Tröster G. Optimization of inertial micropower generators for human
walking motion. IEEE Sensors Journal 2006; 6(1); 28–37.
[64] Zurbuchen A., Pfenniger A., Stahel A., Stoeck C. T., Vandenberghe S.,
Koch V. M., Vogel R. Energy Harvesting from the Beating Heart by a
Mass Imbalance Oscillation Generator. Annals of Biomedical
Engineering 2013; 41(1); 131–141.
[65] Delnavaz A., Voix J. Electromagnetic Micro-power Generator for Energy
Harvesting from Breathing. In: IECON 2012 - 38th Annual Conference
on IEEE Industrial Electronics Society; 2012; p. 984–988.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2019.2925638, IEEE Sensors
Journal
[66] Yeatman E. M. Advances In Power Sources For Wireless Sensor Nodes.
In: Proc. Int’l Workshop Wearable and Implantable Body Sensor
Networks; 2004; p. 20–21.
[67] Mitcheson P. D., Yeatman E. M., Rao G. K., Holmes A. S., Green T. C.
Energy harvesting from human and machine motion for wireless
electronic devices. Proceedings of the IEEE 2008; 96(9); 1457–1486.
[68] Sudevalayam S., Kulkarni P. Energy harvesting sensor nodes: Survey and
implications. IEEE Communications Surveys and Tutorials 2011; 13(3);
443–461.
[69] Brunelli D., Benini L. Designing and managing sub-milliwatt energy
harvesting nodes: Opportunities and challenges. In: Proceedings of the
2009 1st International Conference on Wireless Communication,
Vehicular Technology, Information Theory and Aerospace and Electronic
Systems Technology; 2009; p. 11–15.
[70] Jafer I., Stack P., MacNamee K. Design of New Power Management
Circuit for Light Energy Harvesting System. Sensors 2016; 16(3); 1-20.
[71] Damaschke J. M. Design of a Low-Input-Voltage Converter for
Thermoelectric Generator. IEEE Transactions on Industry Applications
1997; 33(5); 1203–1207.
[72] Van Bavel M., Leonov V., Yazicioglu R. F., Torfs T., Van Hoof C.,
Posthuma N. E., Vullers, R. J. M. Wearable Battery-free Wireless 2-
Channel EEG Systems Powered by Energy Scavengers. Sensors &
Transducers Journal 2008; 94(7); 103–115.
[73] Mitcheson P. D. Energy Harvesting for Human Wearable and Implantable
Bio-Sensors. In: 32nd Annual International Conference of the IEEE
EMBS; 2010; p. 3432–3436.
[74] Rothfuss M. A., Franconi N. G., Bocan K. N., Unadkat J. V, Gimbel M.
L., Mickle M. H., Sejdi E. Implantable Energy Harvesting Stents for
Transcutaneous Wireless Monitoring of Peripheral Artery Disease. IEEE
Sensors Journal 2018; 18(5); 2077–2090.
[75] Borges L. M., Chavez-Santiago R., Barroca N., Velez F. J., Balasingham
I. Radio-frequency energy harvesting for wearable sensors. Healthcare
Technology Letters 2015; 2(1); 22–27.
[76] Kuhn V., Seguin F., Lahuec C., Person C. A multi-tone RF energy
harvester in body sensor area network context. In: 2013 Loughborough
Antennas and Propagation Conference; 2013; p. 238–241.
[77] Zahid Kausar A. S. M., Reza A. W., Saleh M. U., Ramiah H. Energizing
wireless sensor networks by energy harvesting systems: Scopes,
challenges and approaches. Renewable and Sustainable Energy Reviews
2014; 38; 973–989.
[78] Mallela V. S., Ilankumaran V., Rao S. N. Trends in cardiac pacemaker
batteries. Indian Pacing and Electrophysiology Journal 2004; 4(4); 201–
212.
[79] Wang W. Energy harvesting system design and optimization for wireless
sensor network. University College Cork; 2014.
[80] Pu X., Li L., Song H., Du C., Zhao Z., Jiang, C., Cao G., Hu W., Wang
Z. L. A self-charging power unit by integration of a textile triboelectric
nanogenerator and a flexible lithium-ion battery for wearable electronics.
Advanced Materials 2015; 27(15); 2472–2478.
[81] Gaikwad A. M., Khau B. V., Davies G., Hertzberg B., Steingart D. A.,
Arias, A. C. A high areal capacity flexible lithium-ion battery with a
strain-compliant design. Advanced Energy Materials 2015; 5(3); 1–11.
[82] Lee Y. H., Kim J. S., Noh J., Lee I., Kim H. J., Choi S., Seo J., Jeon S.,
Kim T.S., Lee J.Y., Choi J. W. Wearable textile battery rechargeable by
solar energy. Nano Letters 2013; 13(11); 5753–5761.
[83] Li N., Chen Z., Ren W., Li F., Cheng, H.-M. Flexible graphene-based
lithium ion batteries with ultrafast charge and discharge rates.
Proceedings of the National Academy of Sciences 2012; 109(43); 17360–
17365.
[84] González A., Goikolea E., Barrena J. A., Mysyk R. Review on
supercapacitors: Technologies and materials. Renewable and Sustainable
Energy Reviews 2016; 58; 1189–1206.
[85] Vangari M., Pryor T., Jiang L. Supercapacitors: Review of Materials and
Fabrication Methods. Journal of Energy Engineering 2012; 139(2); 72–
79.
[86] Chen K., Wang Q., Niu Z., Chen J. Graphene-based materials for flexible
energy storage devices. Journal of Energy Chemistry 2018; 27(1); 12–24.
[87] Miller J. R., Simon P. Materials science: Electrochemical capacitors for
energy management. Science 2008; 321(5889); 651–652.
[88] Farahiyah N., Aun M., Soh P. J., Vandenbosch G. A. E., Schreurs D.
Revolutionizing Wearables for 5G. IEEE Microwave Magazine 2017;
18(3); 108–124.
[89] Zeng K., Ren K., Lou W., Moran P. J. Energy aware efficient geographic
routing in lossy wireless sensor networks with environmental energy
supply. Wireless Networks 2009; 15(1); 39–51.
1558-1748 (c) 2019 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
15
[90] Voigt T., Ritter H., Schiller J. Utilizing Solar Power in Wireless Sensor
Networks. In: 28th Annual IEEE International Conference on Local
Computer Networks; 2003.
[91] Li H., Tan J. Heartbeat-Driven Medium-Access Control for Body Sensor
Networks. IEEE Transactions on Information Technology in Biomedicine
2010; 14 (1); 44-51.
[92] Kerzenmacher S., Ducree J., Zengerle R., von Stetten F. Energy
harvesting by implantable abiotically catalyzed glucose fuel cells. Journal
of Power Sources 2008; 182; 1-17.
[93] Shnayder V., Hempstead M., Chen B.R., Allen G.W., Welsh M.
Simulating the Power Consumption of Large-Scale Sensor Network
Applications. In: Proceedings of the 2nd International Conference on
Embedded Networked Sensor Systems; 2004; 188-200.
[94] Jelicic V., Magno M., Brunelli D., Bilas V., Benini L. Benefits of wake-
up radio in energy-efficient multimodal surveillance wireless sensor
network. IEEE Sensors Journal 2014; 14(9); 3210-3220.
[95] Karlen W., Floreano D. Adaptive Sleep-wake Discrimination for
Wearable Devices. IEEE Transactions on Biomedical Engineering 2011;
58(4); 920-926.
[96] Magno M., Brunelli D., Sigrist L., Andri R., Cavigelli L., Gomez A.,
Benini L. InfiniTime: Multi-sensor Wearable Bracelet with Human Body
Harvesting. Sustainable Computing: Informatics and Systems 2016; 11;
38-49.
[97] Hsu J., Zahedi S., Kansal A., Srivastava M., Raghunathan V. Adaptive
Duty Cycling for Energy Harvesting Systems. In: Proceedings of the 2006
International Symposium on Low Power Electronics and Design; 2006; p.
180-185.
[98] Park C., Chou P.H., Bai Y., Matthews R., Hibbs A. An ultra-wearable,
wireless, low power ECG monitoring system. In: Proceeding of the IEEE
Biomedical Circuits System Conference 2006; p. 241-244.
[99] Marzhencki M., Tavakolian K., Chuo Y., Hung B., Lin P., Kaminska B.
Miniature Wearable Wireless Real-time Health and Activity Monitoring
System with Optimized Power Consumption. Journal of Medical and
Biological Engineering; 30(4); 227-236.
[100] Kansal A., Hsu J., Zahedi S., Srivastave M. Power Management in
Energy Harvesting Sensor Networks. ACM Transactions on Embedded
Computing Systems 2007; 6(4); 1-35.
[101] Anastasi G., Conti M., Di Francesco M., Passarella A. Energy
Conservation in Wireless Sensor Networks: A survey. Ad Hoc Network
2009; 7(3); 537-568.
[102] Shigeta R., Sasaki T., Quan D.M., Kawahara Y., Vyas R., Tentzeris M.,
Asami T. Ambient-RF-energy-harvesting sensor node with capacitor-
leakage-aware duty cycle control. In: Proceedings of 2012 IEEE Sensors;
1-10.
Yung-Wey Chong received her M.Sc
degree in advanced computer networks in
2012 from Universiti Sains Malaysia,
Malaysia. She is currently a lecturer at
National Advanced IPv6 Centre. She joined
the centre in 2009 as a research staff
member, working in VoIP protocols over
wireless communication networks. She is a
committee member of SOI Asia
(www.soi.asia), a project that utilize satellite-based Internet to
support interactive multimedia communications between
partner universities. She was involved in CONNECT2SEA
project (www.connect2sea.eu), a project funded under FP7,
which supports European Union and South-East Asia strategic
partnership and policy dialog. Her research interest is mainly in
embedded systems, wireless communications, and software-
defined networking (SDN).
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2019.2925638, IEEE Sensors
Journal
16

Widad Ismail received the bachelor’s Award from Ministry of Economic Affairs and Outstanding
(Hons.) degree in electronics and Research Award from National Science Council in 2009.
communication engineering from the
University of Huddersfield, U.K., in 1999,
and Ph.D. degree (Active Integrated
Antenna with Image Rejection) in
electronics and communication
engineering from the University of
Birmingham, U.K., in 2004. She is currently a Professor and a
Project Coordinator with the Auto-ID Laboratory, Universiti
Sains Malaysia. Her main areas of research are wireless system
design, RFID, active integrated antennas, and RF and
microwave systems engineering. She is a member of the
Wireless World Research Forum.

Kwangman Ko is a Professor at the School

of Computer and Information Engineering,
Sangji University, South Korea. Currently
his research focuses on the retargetable tool
suite (low power/energy optimized compiler,
simulator, debugger) for the embedded
systems, Virtual Machines and the Energy-
oriented Architecture Description Language
(EoADL). He is a member of IEEE and committee member for
the various Korea Information and Multimedia Societies.

Chen-Yi Lee received the B.S. degree from

National Chiao Tung University (NCTU),
Hsinchu, Taiwan, in 1982, and the M.S. and
Ph.D. degrees from Katholieke University
Leuven (KUL), Leuven, Belgium, in 1986
and 1990, respectively, all in electrical
engineering. From 1986 to 1990, he was
with IMEC/VSDM, Belgium, working in
the area of architecture synthesis for DSP. In February 1991, he
joined the department of electronics engineering and from
2003-2006, he was the chairman. During 2000-2003, he was the
director of National CHIP Implementation Center (CIC) and the
coordinator of Microelectronics Program of Engineering
Division, NSC, Taiwan. He was the dean of office of research
and development in 2007-2010, National Chiao Tung
University (NCTU). Currently, he serves as Co-Program
Director of National Program of Intelligent Electronics (NPIE)
and professor of the department of electronics engineering,
NCTU. His research interests mainly include VLSI algorithms
and architectures for high-throughput DSP applications. He is
also active in various aspects of micro sensing, low-power
system-on-chip, and deep learning. Dr. Lee served as the
program committee member of IEEE ISSCC in 2004-2006,
DATE TPC member in 2006-2007, IEEE ASSCC TPC member
in 2006-2014, IEEE VLSI Symposium JFE Program
Committee member in 2010-2014, IEEE TCAS-II Associate
Editor in 2010-2011, and the Past-Chair of Taipei Chapter of
IEEE Circuits and Systems (CAS) Society. He received the
Outstanding Technology Licensing Award from National
Science Council in 2007-2008, and Industry Contribution

1558-1748 (c) 2019 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.