Energy Harvesting Techniques For Internet of Thing

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/ACCESS.2021.3064066, IEEE Access
Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000.
Digital Object Identifier 10.1109/ACCESS.2021.Doi Number
Energy Harvesting Techniques for Internet of

Things (IoT)
Teodora Sanislav1, George Dan Mois1, Sherali Zeadally2, Silviu Folea1
1
Automation Department, Technical University of Cluj-Napoca, Romania
2
College of Communication and Information, University of Kentucky, USA
Corresponding author: Teodora Sanislav (e-mail: teodora.sanislav@aut.utcluj.ro).
ABSTRACT The rapid growth of the Internet of Things (IoT) has accelerated strong interests in the
development of low-power wireless sensors. Today, wireless sensors are integrated within IoT systems to
gather information in a reliable and practical manner to monitor processes and control activities in areas
such as transportation, energy, civil infrastructure, smart buildings, environment monitoring, healthcare,
defense, manufacturing, and production. The long-term and self-sustainable operation of these IoT devices
must be considered early on when they are designed and implemented. Traditionally, wireless sensors have
often been powered by batteries, which, despite allowing low overall system costs, can negatively impact
the lifespan and the performance of the entire network they are used in. Energy Harvesting (EH) technology
is a promising environment-friendly solution that extends the lifetime of these sensors, and, in some cases
completely replaces the use of battery power. In addition, energy harvesting offers economic and practical
advantages through the optimal use of energy, and the provisioning of lower network maintenance costs.
We review recent advances in energy harvesting techniques for IoT. We demonstrate two energy harvesting
techniques using case studies. Finally, we discuss some future research challenges that must be addressed to
enable the large-scale deployment of energy harvesting solutions for IoT environments.
INDEX TERMS Energy efficiency, Energy harvesting, Internet of Things, IoT device, Wireless sensor
networks
I. INTRODUCTION • IoT connectivity (protocols, gateways) is responsible for

Fifteen years ago, the International Telecommunications transferring data in the online, cyber-physical world.
Union (ITU) published its first report on the Internet of • The IoT cloud stores data and it is also the place where
Things (IoT) [1]. The IoT paradigm was first defined as a decisions are made.
new dimension added to the world of Information and • IoT analytics and data management are responsible for
Communication Technologies (ICTs) that allows making processing the data.
connections for anyone and anything, anytime and anywhere • End-user devices and user interfaces help to control and
to create a new dynamic network of networks [1]. Today, configure the system.
IoT is no longer an emerging trend. It has become one of the
Each of these components must address significant
most important technologies of the current century with
scientific and technological challenges for achieving efficient
applicability in many industries such as transportation,
and scalable implementations. For example, the energy
energy, civil infrastructure, smart buildings, environment
monitoring, healthcare, defense, manufacturing, and efficiency paradigm associated with IoT devices must be
production. IoT continues to grow. Experts predict that, by considered early on during their design phase [4]. Enabling a
2025, about 22 billion IoT devices will be connected to the seamless flow of information throughout the IoT ecosystem
Internet and will communicate in this IoT environment [2]. is another important challenge because wireless connectivity
The IoT ecosystem has the following main components is highly complex, and the fast-evolving wireless standards
[3]: contribute to this. The development of new Artificial
• IoT devices (sensors and actuators) are responsible for Intelligence (AI) techniques to analyze huge amounts of data
collecting data or controlling a certain process. and make real-time decisions is another major challenge.
Also, security and privacy must be considered because
existing security protocols such as Data Encryption Standard
VOLUME XX, 2021 1
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 publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
(DES), Advanced Encryption Standard (AES) and Rivest, QoS (Quality of service) assurance, and the efficient
Shamir, and Adelman (RSA), are not suitable for IoT devices management of large amounts of data [10]. Therefore, the
due to their resource (memory, processing power) constraints long-term and self-sustainable operation of wireless sensors
and heterogeneity [5], [6], [7]. (IoT devices) is a critical issue and must be addressed
We describe the challenge associated with the energy properly. Energy Harvesting (EH) technology is an
efficiency paradigm for some IoT devices (such as wireless environment-friendly solution that has the potential to
sensors) below. The IoT vision benefits from the features of extend the lifetime of these sensors. Energy harvesting
Wireless Sensor Networks (WSNs) [8], [9] and relies on refers to the harnessing of energy from various external
these systems for gathering data about the environment and sources and its conversion into electricity.
for performing actions following the analysis of the
collected data. IoT uses Internet Protocol (IP) connectivity
for assuring that every one of its components, or “things”,
has a distinct address, while WSNs do not necessarily
require a connection to the Internet. However, due to the
continuously increasing power of microelectronics
combined with their decreasing costs, improved power
efficiency of hardware technologies along with better
wireless communication protocols, WSNs have become a
vital component of the IoT ecosystem. WSNs can extend the
Internet, or the cyber environment, into physical spaces [10],
FIGURE 2. Wireless sensor (IoT device) components.
[4]. Moreover, IoT and WSN have become almost
inseparable [11], wireless sensor networks being recognized
This work deals with the analysis of energy harvesting
as a key enabler of IoT [12]. Some of the domains, wherein
techniques suitable for IoT devices. We summarize our
IoT technologies have been integrated with WSNs include
research contributions as follows:
healthcare [13], agriculture [14], smart cities [15] and smart
• We motivate why energy harvesting is important for the
buildings [16], manufacturing [17], and transportation
IoT ecosystem.
systems [18]. Figure 1 shows a typical IoT scenario where
• Based on a comprehensive review of energy harvesting
data also is collected using WSNs.
solutions for IoT, we analyze different techniques to
harvest energy from various sources.
• We present two IoT energy harvesting devices
developed by the authors, along with considerations
regarding their design, implementation, and testing.
We organize the rest of the paper as follows. The second
section highlights why energy harvesting is important for
IoT. In this section, we highlight the limitations of using
batteries as the main power source for wireless sensors. The
third section presents an overview on energy harvesting
techniques that can be used in IoT. We present a
classification of these techniques according to the sources
FIGURE 1. Typical IoT scenario including Wireless Sensor Networks from which energy is harvested energy. We analyze recently
(WSNs). proposed energy harvesting techniques for IoT. In the fourth
section we describe the energy harvesting models and
The use of wireless sensor networks has proven to be consumption models that have been proposed for
beneficial in specific applications where access is difficult, implementing energy harvesting IoT devices. We then
such as remote locations or on moving parts of machinery present two case studies from our previous works on solar
[19], [20], [21], [22]. Generally, a wireless sensor is made energy harvesting and Radio Frequency (RF) energy
up of four main components, namely the sensorial part, the harvesting. We describe the architecture, operation, and
processor, the transceiver, and a power supply, as Figure 2 energy characteristics of a solar powered environmental IoT
shows [23]. In general, wireless sensors have limited device based on Bluetooth Low Energy (BLE)
capabilities in terms of processing power and storage and communication. We present a prototype for a BLE-enabled
must operate on small amounts of energy because the power environmental beacon, powered by a RF energy harvesting
source, in most of the cases is a battery which has limited element in the fifth section. The sixth section discusses a
capacity. Other challenges encountered in the deployment of few technical challenges that still need to be addressed on
solutions that rely on wireless sensors include reliable the design and development of energy harvesting devices
communication, coverage and deployment issues, security, even though progress has been made in recent years on self-
VOLUME XX, 2021 2
powered IoT devices. Finally, we make some concluding the achievement of sustainable and energy efficient systems
remarks in the final section. [29]. These are some of the reasons why energy harvesting
elements are included in the design of current wireless
II. MOTIVATION FOR ENERGY HARVESTING IN IOT sensors. These energy harvesting components can act as
It is desirable for the network lifetime to be as long as secondary power sources or can completely replace batteries.
possible. Therefore, energy efficiency and long battery The use of energy harvesting extends mote’s lifetime by
lifetime are of crucial importance in the design and replenishing its energy from an energy source, such as solar
development of wireless sensors (IoT devices). However, a cells, vibration or fuel cells, acoustic noise, or mobile
battery has a finite capacity and, even though there have been suppliers (robots) [10]. In this way, the long-term and self-
breakthrough developments in communication protocols, sustainable operation of wireless sensors, one of the most
operating systems, and in the implementation of powerful important issues in the widespread use of IoT, is assured
power management mechanisms, highly efficient MAC [29].
(medium access control) and routing protocols, it will Energy harvesting or energy scavenging is the process
eventually get depleted. To ensure the operation of the through which energy from external sources, such as
network, costly maintenance operations for battery mechanical load, vibrations, temperature gradients, or light,
replacement must be performed. The scale of wireless sensor is captured and converted to obtain relatively small levels of
networks coupled with sensor placement, often in power supplied to electronic devices [30], [4]. The intake of
inaccessible locations, make these maintenance operations energy from the surrounding environment leads to a green
difficult in many cases [24]. The use of batteries as power energy source that replaces primary batteries or charges
supply for wireless devices has other drawbacks besides their secondary cells and represents a cost-effective and
ability to provide power for a finite amount of time. These environmentally sound method for wireless devices [31]. The
include reduced energy densities and leakage, that discharges three components of a common energy harvesting system are
the battery even when it is not used. Moreover, the the source (external energy that is collected), the harvesting
application space of batteries is limited by the impact of architecture (mechanisms), and the load (the consumer) [32].
temperature on their proper operation, extreme conditions The energy can be used immediately at the time it is
leading to capacity and power losses [25]. Another problem, harvested or it can be stored for future use, resulting in two
that is common to the operation of wireless sensors is the main architectures, namely, Harvest-Use and Harvest-Store-
proper management of short high-current pulses that affect Use [32]. Figure 3 shows the two distinct architectures for a
the capacity and lifetime of batteries [26]. Battery weight and wireless sensor node with energy harvesting elements.
dimensions directly affect capacity, and their reduction for Depending on the configuration of the system, belonging to
achieving small form factor designs will lead to a shorter one of the two architectures, its power generation part
operation time for the devices they power. Another major includes specific energy harvesting circuits that convert
concern is related to the environment because batteries ambient energy to Direct Current (DC) energy, power
contain harmful chemicals and toxins that make their management units that increase the efficiency of power
disposals more complex [27], [25]. Although most batteries generation and its use, and storage elements, that can store
can be recycled and many governmental initiatives have been energy and power the electronics. All these components and
established for increasing the recycling rates, improvements their associated characteristics and actions are the subject of
in this direction are still necessary [28]. Currently, there is intense research efforts, that target the manufacturing of
also a generalized trend of reducing the environmental energy-autonomous wireless sensors when deployed in the
impact of information and communication technology, and IoT environment.
this also applies to wireless sensors, where the goal is toward
FIGURE 3(a). Wireless sensor (IoT device) with energy harvesting FIGURE 3(b). Wireless sensor (IoT device) with energy harvesting
architectures: Harvest-Use. Adapted from [32]. architectures: Harvest-Store-Use. Adapted from [32].
VOLUME XX, 2021 3
III. ENERGY HARVESTING TECHNIQUES FOR IOT Integrated Circuits (ICs) [34]. The main types of PV cells,
For energy harvesting several energy sources can be depending on the materials in their composition, are the
considered. These can be classified into the following mono-crystalline, polycrystalline, and amorphous silicon, or
categories according to the sources from which the harvested thin-film cells [34], [36]. The parameter used for comparing
energy comes: ambient, mechanical, human, organic and their performance is their efficiency which is defined as the
hybrid. The ambient energy sources are available in the ratio of the maximum output power to the incident light
environment and can be easily accessed without any costs. power under 100 mW/cm2 illumination [37]. The first
These ambient sources can in turn be divided into the category (mono-crystalline PV cells) has the highest
following categories: solar, Radio Frequency (RF), thermal, efficiency, between 15 and 24%, but the corresponding
wind, and hydro based energy sources [24]. Vibrations and mono-crystalline PV cells are the most expensive. The
pressure are mechanical energy sources that are deployed second group, made up of polycrystalline PV cells, has an
explicitly in the environment for harvesting purposes [24]. efficiency between 14 and 20%, while thin-film cells are the
Humans through their motion and physiology, organisms, cheapest, but with an efficiency lower than 13.2% [36].
and plants, represent other sources of energy that can be In the context of wireless sensors, the use of solar cells for
scavenged. In some cases, the use of only one energy source harvesting energy from the environment is a mature
is not enough, therefore several types of energy sources must technique and is the most common mechanism used for
be combined to reap sufficient power needed by electronic prolonging the lifetime of the power supply [38], [39], [40].
equipment. Solar cells offer the highest power density, of approximately
15 mW/cm2, as compared to various other energy harvesting
A. AMBIENT ENERGY HARVESTING techniques [41]. Even though solar power is uncontrollable,
Solar and RF are the main ambient energy sources that are and the conversion efficiency is affected by the day-night
naturally or artificially always present in the environment and cycle, seasonal changes, weather conditions, and ambient
can be considered in making self-powered devices for the temperature, it can be predicted and modeled so that
IoT ecosystem. This is why in this work we focus only on adequate strategies are adopted for assuring continuous
these ambient sources. power to electronic devices [42], [43], [44], [45], [46]. To
ensure uninterrupted operation (during the periods such as
Solar energy harvesting night and the presence of clouds when ambient light is not
Solar power is the most abundant energy source on earth, available) of the device powered by solar energy harvesting
with approximately 173x1012 kW of energy produced components, an architecture that includes storage elements,
continuously, a quantity that exceeds by far the world’s as Figure 4 shows, is the most common [47], [48]. The
demand and use [33]. The total energy that reaches the energy storage can be a rechargeable battery, a
earth’s surface annually is approximately 3.4 x 1024 J, an supercapacitor, or a combination of the two with each choice
amount that is 7000 to 8000 times higher than the yearly having its own advantages and limitations. The power is
primary energy consumption of the globe [34]. This energy is stored when the amount generated exceeds the consumption
harvested through helio-chemical (photosynthesis), helio- of the device, or when it is in sleep or inactive mode and this
electrical (photovoltaic converters), and helio-thermal stored power is used when the energy source is missing or
(thermal energy production, solar water heaters) processes when required by the system [48]. PV cells are a well-known
[35]. The helio-electrical process is based on the technology which have been used in a wide range of IoT
PhotoVoltaic (PV) effect, that can be observed when two devices. When PV technology is used in conjunction with
dissimilar materials convert solar rays into DC (direct adequate energy storage, power consumption optimization,
current) power when struck by light [24]. The device that is appropriate circuit design and dimensioning, we can develop
used for generating electricity from sunlight is the solar cell efficient energy autonomous devices [49], [50], [51], [52],
or the photovoltaic cell, that is a solid-state electrical junction [53].
device [24]. It is usually built from silicon, using a process
like the one used in the manufacturing of transistors and
FIGURE 4. Main components of a solar energy harvesting system. Adapted from [47].
VOLUME XX, 2021 4
Energy harvesting using PV technology was initially Mechanical vibrations energy harvesting
intended for outdoor use. PV cells have been designed Mechanical vibrations have a sufficiently high energy density
mainly for generating energy from sunlight with conversion and, in some cases, where IoT devices are deployed indoors
efficiencies in indoor environments having smaller values as or in overcast areas, their use can replace solar harvesting
compared to higher conversion efficiencies with outdoor systems. The energy from low frequency (< 100 Hz) or high
environments [54], [55]. However, advances in frequency (> 100 Hz) vibrations is usually harvested through
manufacturing processes and circuits along with improved piezoelectric, triboelectric, electromagnetic, and electrostatic
designs of IoT devices have enabled the deployment of energy harvesters.
indoor solar powered systems [56], [57], [58], [59]. The Piezoelectric Energy Harvesters (PEHs) work based
Furthermore, significant progress has also been made in the on the combination between the mechanical and the electrical
development of energy supplies for moderate power systems, behaviors of certain categories of materials such as crystals
that consume power in the range of 1 to 10 W, combined and ceramics [69]. These harvesters do not require external
with the current popular use of low-power IoT devices voltage sources, have a minimum of moving parts, and can
requiring between 10 and 100 mW [60]. This allows energy generate power with voltage levels that can be easily
autonomous embedded systems to run data-intensive tasks conditioned (i.e., converted to DC) [70]. This type of
such as computer vision and facilitates the implementation of harvesters benefits from high-power density, simplicity in
edge computing technologies. their design and fabrication, and they use a wide range of
frequencies. Based on the literature review in this area, we
RF energy harvesting found that the piezoelectric energy harvesting is the most
Wireless Internet, radio and satellite stations, and digital widely researched. The works in [70 – 74] analyze state-of-
multimedia broadcasts are sources of RF or electromagnetic the-art of energy harvesting using piezoelectric generators at
signals, between 3 kHz and 300 GHz of frequency spectrum, micro and nano scale and highlight that this method is one of
that can be converted into electrical energy with the help of the most promising solutions to power IoT devices. In [75],
an antenna and a rectifier circuit, as Figure 5 shows. This the authors used PEHs installed on an actual roadway for five
type of energy exists around us (indoors and outdoors) at months, with vehicles traveling at speeds of 10–50 km/h. The
different levels, and it is always available (day or night). This maximum generated power of 2080 mW and a power density
energy can be retrieved without limit, but it has several of 20.79 W/m2 were recorded with a vehicle speed of 30
disadvantages such as low density and low efficiency that is km/h, and 2381 mW and a power density of 23.81 W/m2 at a
inversely proportional to distance [61]. Based on a review of vehicular speed of 50 km/h. Another test presented in [75]
the past literature in this area [61 - 63], we found that RF involved the use of eight PEHs installed on the highway rest
waves energy harvesting is the best solution in many area. The results obtained prove the harvesters’ ability to
scenarios when it comes to low-energy IoT devices. The successfully operate 24 LED indicators that can ensure
works in [64, 65] and [66] highlight the potential of using RF drivers’ safety at night, to monitor in real time the conditions
Energy Harvesters (RFEHs) to power IoT devices for inside the harvesters through their sensors (temperature,
environmental and healthcare monitoring. Other approaches strain, and leakage), and to collect traffic data such as the
for harvesting energy from RF signals include opportunistic number of vehicles passing the harvester zone.
charging from nearby smartphones [67], or the sharing of The ElectroMagnetic Energy Harvesters (EMEHs) work
energy between different wireless systems close to each other based on the relative motion between a conductor (such as a
[68]. coil) and a magnetic field (created by a magnet) in response
to mechanical vibrations. This category of harvesters has
B. MECHANICAL ENERGY HARVESTING attracted considerable attention due to their ability to
Mechanical vibrations and pressure are energy sources that generate high output currents, robustness, and their low-cost
surround us, and these sources can be considered in making designs. However, they continue to be challenging in terms
self-powered IoT devices used in a wide range of of poor transduction properties of planar magnets and the
applications. limited number of induction loops when it comes to IoT
small-scale devices [69].
FIGURE 5. Main components of a RF energy harvesting system. Adapted from [61].
VOLUME XX, 2021 5
In [76], the authors proposed a battery-free solution to extent than vibrations. For example, the work in paper [81]
power a Bluetooth board with a DC voltage equal to 2 V by a demonstrates the feasibility of using this energy source.
low voltage vibration electromagnetic converter (with an
open circuit output voltage equal to 1.8 V peak to peak for a C. HUMAN ENERGY HARVESTING
frequency of 24 Hz) as energy source. In [77], the authors The human body is an energy warehouse which can ensure
present an EMEH working at low frequency ambient alterative power supply through the collection of energy from
vibrations (< 100Hz). To demonstrate the design, a macro heat and motion [82]. This type of energy source can be
level prototype was used, and the voltage generated for a exploited by wearable and implantable electronic devices that
frequency of 50 Hz and 20 turns of copper coil is 20 mV, i.e., are IoT devices used to monitor the activity of healthy people
1mV per turn. The approach described in [78] presents a or the patient’s condition. However, this energy harvesting
viable hybrid solution which includes an EMEH to collect approach encounters difficulties due to the following factors:
the energy from the bridge’s vibrations and ambient wind human motion has a relatively low frequency (typically
surges to implement IoT devices for bridges’ health under several tens of Hz) and also it is highly stochastic and
monitoring. Two prototypes with multi-resonant frequencies irregular; the human body temperature depends on the
have been designed. The first prototype, suitable for narrow daytime rhythm and on the instant disturbance of daily
band vibration environments, has a frequency band from 1 to activities performed [82]. Additionally, the devices must be
18 Hz and acceleration levels below 0.4 g and generates an worn by people and therefore they must have reasonable size
open circuit voltage of 810 mV and an optimum power of and weight and must interfere only minimally with the
354.51 μW. It also produces an adequate voltage and power natural functions of the body [83].
levels (up to 7.84 μW) from wind surges from 0.5 m/s to The research conducted in this field has focused on the
9 m/s. The second prototype, suitable for vibration extraction of energy from human body focusing on heat [84 -
surroundings, has a frequency band from 1 to 45 Hz and 91] and biomechanical energy [83, 92 - 96].
acceleration levels below 0.6 g and generates an open circuit
voltage of 618 mV and an optimum load power of Human heat energy harvesting
2214.32 μW. It can harvest the power (up to 9.14 μW) from Human heat energy harvesting is based on the changes of the
ambient wind with speed from 0.5 m/s to 6 m/s. human body temperature and uses two types of energy
The ElectroStatic Energy Harvesters (ESEHs) use the harvesters: the ThermoElectric Energy Harvester (TEEH)
mechanical vibrations to move the charged capacitor plates that utilizes a spatial-temperature gradient and the
of a variable capacitor structure against the electrostatic PyroElectric Energy Harvester (PEEH) which requires a
forces between the electrodes which are separated by air, temporal variation in temperature. The work [84] presents an
vacuum, or a dielectric material [69]. Unlike PEHs and ultra-low power batteryless energy harvesting Body Sensor
EMEHs, ESEHs require a DC voltage (bias voltage) supplied Node (BSN) capable of acquiring, processing, and
by a battery to oppositely charge the capacitor plates. ESEHs transmitting ElectroCardioGram (ECG), ElectroMyoGram
generate high output voltage and relatively larger output (EMG), and ElectroEncephaloGram (EEG) data. The BSN is
power density, provide a wider choice of frequencies at the totally powered by a commercially available TEEH with
low-frequency range, and offer the possibility to build low- about 60 μW output power with 30 mV output voltage. In the
cost devices. In [79] the authors present an ESEH (whose heart-rate extraction mode where the transmitter is duty-
footprint is as small as 1 cm2) that can reach an output power cycled, the sensor node, including regulation, consumes only
of 495 μW sinusoidal vibration. Used in real life conditions, 19 µW. The authors of [85] describes a hidden TEEH of
under impact vibration inside of a tire tread, the harvester human body heat, integrated into an office shirt. It generates
generates an output power of 60 μW on a traveling speed of power in the range of 5-0.5 mW at ambient temperatures of
60 km/h. The result of the research contributes to the 15 °C to 27 °C. The tests made highlight that the
evolution of intelligent automobiles in terms of tire sensors thermoelectric shirt produces more energy during nine
(IoT devices). The work presented in [80] proposes an ESEH months of use (if worn 10 h/day) than the energy stored in
design based on the electrostatic coupling methods. The alkaline batteries of the same thickness and weight. Its
results demonstrate that the ESEH can harvest more than 1 technical properties make it a reliable power supply for low-
μW from 59 to 148 Hz, and more than 0.5 μW from 14 to power IoT devices used in healthcare. More information on
152 Hz at an acceleration of 2 grms (Root Mean Square TEEHs for human heat energy harvesting is provided in [86
acceleration). It was successfully used to power an energy – 88]. The temperature variations of the human body are not
autonomous temperature sensor node with a data high during the day. In this condition, the available heat
transmission beyond a distance of 10 m at 868 MHz. energy for PEEHs is limited. This is why when it comes to
human body energy harvesting applications, PEEHs are
Mechanical pressure energy harvesting combined with other types of energy harvesters. Such an
Mechanical pressure is exploited to implement energy approach is demonstrated in [89] through a proof of concept
harvesters using the piezoelectric method, but to a lesser of a hybrid harvester combining piezoelectric and
VOLUME XX, 2021 6
pyroelectric properties for building self-powered healthy This energy is sufficient for transmitting an electric signal
monitoring and interactive sensing systems. The tests made with a single switch using low power bistatic scatter radio
to determine the output voltage and current from the principles. The use of soil energy for powering wireless
pyroelectric effect, in the absence of any strain, by varying sensors has been proposed in [98], where the temperature and
the temperature of the hot plate to which the sensor is air moisture are measured and transmitted using BLE
attached from 295 K to 303 K, yielded an output voltage and technology to terminal devices such as smartphones. The soil
current pulse peaks up to 0.1 V and 20 nA respectively. cell fabricated by the authors supplies an average power of
These correspond to a peak power density of 2 mW / cm-3. 60 to 100 μW which is sufficient to power the BLE sensor so
Also, the approaches [90, 91] present viable hybrid solutions that it can perform the aforementioned tasks. These
which include PEEHs to implement wearable IoT devices. approaches can help in the development of environmentally
friendly monitoring applications in IoT contexts.
Biomechanical energy harvesting
Biomechanical energy available from human motion can be E. HYBRID ENERGY HARVESTING
classified into kinetic energy and elastic energy [82]. Given There are cases wherein the scavenging of energy from a
the complexity of the physical mechanisms, several types of single source does not generate the amount of power required
energy harvesting devices are used: electromagnetic, for the operation of the IoT device at all times. This
electrostatic, piezoelectric, and triboelectric [82, 92]. Based challenge led to the design of systems that include multiple
on our review of the literature [82, 92 - 94] in this area, we energy harvesting units, combined with energy storage
found that a wide range of devices and applications have components [99]. An example is solar energy harvesting. As
been reported. These published studies show that in terms of we have mentioned earlier, generating DC from light can
power generation, the electromagnetic energy harvesters are assure the required energy for the operation of wireless
the best candidate to capture the kinetic energy. However, the sensor nodes. However, the main drawback of using solar
benefit that the IoT devices based on smart material-based energy is its reliability, that is affected by weather conditions
energy harvesters offer to those who wear them cannot be and spatial-temporal factors, such as the Sun’s position
ignored. The authors of [95] described an electromagnetic during the day [41]. Therefore, hybrid solutions, where solar
energy harvester prototype to efficiently scavenge the kinetic energy harvesting is complemented by other mechanisms or
energy of human limbs swing. In real walking conditions, the power supplies have been proposed [41], [48], [99]. Hybrid
maximum power achieved was 1.84 mW and 2.95 mW for energy harvesters combine circuits that generate power from
the device worn on the wrist and ankle respectively, while single energy sources, such as solar, radio frequency, and
the corresponding power densities are 573.21 μW / cm3 and vibrations and can also use multiple types of transduction
919.01 μW / cm3, respectively. The results confirm that the mechanisms for converting energy to electricity [100]. By
energy harvester can entirely power a pedometer at various generating a power output equal to, or larger than the overall
walking speeds. The work in [96] presents the design, consumption of an IoT device for a certain period, energy-
implementation, and evaluation of a fiber-based generator autonomy is achieved. In [101], the authors presented the
that converts biomechanical energy (motions/vibrations) into design, implementation, and evaluation of a self-powered
electricity using electrostatic mechanisms. The average WSN node with wind, solar, and thermal energy. The
output power density of ~0.1 µW / cm2 makes this generator average generated capacity by the harvesting mechanisms of
usable as an effective building element for a power shirt to 7805.09 J exceeds the energy consumption of the node,
trigger a wireless body temperature sensor system and as a measured at approximately 2972 J, demonstrating the
self-powered active sensor to quantitatively detect human practicality of the hybrid approach.
motion. Table 1 presents a summary of this section. It describes
each energy source along with the technology used to harvest
D. BIOENERGY HARVESTING it, the advantages and disadvantages of these technologies,
There are also more special types of energy harvesting. For the range of power density obtained, and the application
example, the use of plants for generating the energy required domains that can get benefit if they use them to develop self-
for powering wireless sensors has been investigated in [97]. powered IoT devices.
The plant-as-battery approach is meant to simplify the
deployment of wireless sensors in agricultural applications, IV. ENERGY HARVESTING MODELING
where the measurement of soil moisture, ambient humidity, The previous section presented the main energy sources and
or the monitoring of plants for detecting pests are some of energy harvesting technologies that can be used to develop
their common functions. The authors of [97] describe an IoT devices. Regardless of the technology used, from the
approach that exploits the ability of plants to produce electric design phase of IoT devices with energy harvesting
signals that are harvested by a power management unit, characteristics, we must maintain a balance between the
generating a power between 800 and 1400 nW during a day. generated and stored power and that which is consumed.
VOLUME XX, 2021 7
TABLE 1. Analysis of energy harvesting techniques.
Application
Energy source Technology Power density Advantages Disadvantages
domains
Solar PV cell 10 - 100 mW / cm2 High-output voltage Unavailable at night Environment
(outdoor) Low fabrication costs Non-controllable monitoring,
< 100 μW / cm2 (indoor) Predictable healthcare,
agriculture
RF Antenna 0.01 - 0.1 µW / cm2 Available anywhere, anytime Distance dependent Environment
1- 10 mW / cm2 Predictable Low-power density monitoring
Controllable
Mechanical Piezoelectric 4 - 250 µW / cm3 High-power density Highly variable output Infrastructure
vibrations and No external voltage source Unpredictable monitoring,
pressure Simplicity design and fabrication automotive
Controllable
Electromagnetic 300 - 800 µW / cm3 High-output currents Relatively large size
Robustness Unpredictable
Low-cost design
Controllable
Electrostatic 50 – 100 µW / cm3 High-output voltage Requires bias voltage
Relatively larger output power Unpredictable
density
Possibility to build low-cost devices
Controllable
Human heat Piezoelectric < 35 μW / cm2 Sustainable and reliable Low-power density Healthcare
Pyroelectric Available Unpredictable
Controllable
Biomechanical Electromagnetic < 4 μW / cm3 Available Low-power density Healthcare
Piezoelectric < 300 μW / cm3 Controllable Unpredictable
Triboelectric
Electrostatic
Bio [102] Metal electrodes Extremely low wattage Available Extremely low wattage Environmental
Controllable Suitable for nanoscale sensing in
electrical devices agricultural
applications
In this context, various energy harvesting models and

consumption models have been proposed for two main
approaches, namely, Harvest-Store-Use and Harvest-Use
(not very common and not suitable for energy sources that
are uncontrollable or unpredictable). The energy harvesting
models used in the Harvest-Store-Use approach take into
account the capacity of the storage element (the harvested
energy can always be stored, or the harvested energy can be
stored up to a limit and the rest is lost). Also, the harvested
energy can be modeled as a deterministic or a stochastic
process. If the incoming energy and its fluctuations are
known in advance, as in the case of predictable energy
sources, the deterministic model is suitable. Otherwise, the FIGURE 6. Voltage on the storage element (supercapacitor) during
stochastic approach is the right one. However, in each one of operation - T and RH Wi-Fi sensor [53].
the cases, the basic idea for assuring battery-free operation
for IoT devices is to harvest at least the amount of energy ge element with enough energy for assuring autonomous
required for the proper operation of the electronics. Usually, operation.
the power source must provide enough energy for data As we have previously mentioned, although energy
processing operations, transmission and/or receive actions, harvesting poses many challenges, the majority of these can
and sleep periods. Figure 6 shows the voltage on a be overcome by efficient, energy aware, system designs
supercapacitor that is charged by a solar cell which powers a [103]. Since the amount of scavenged energy is so low (as
temperature sensor and a relative humidity wireless sensor. Table 1 shows), in most cases the energy harvesting IoT
These sensors send data periodically to a cloud platform device must operate on amounts of power as low as possible.
through the Internet. The graph indicates a favorable Several strategies have been proposed for prolonging the
situation wherein the solar cell is capable to charge the stora- lifetime of wireless IoT devices, including duty-cycling
VOLUME XX, 2021 8
[104], sleep scheduling [105], the reduction of the required

transmission distance for IoT devices through efficient
clustering [106], optimized strategies for adaptively setting
the rates of sensor reading and data transmission depending
on available energy [107], or the development of scheduling
schemes that take into account power consumption when
waking up the wireless sensing systems [108]. Some
practical techniques for reducing the energy required by an
IoT device, also used in the development of low-power
embedded systems, include Dynamic Voltage Scaling (DVS)
[109], the reduction of the frequency of the processing unit
[110], or the appropriate selection of peripherals in the device
FIGURE 7. Profile of the current drawn by a Wi-Fi sensor during wakeup
[111] or of the type of memory involved in data processing (T, RH, P, CO2, and light intensity) [127].
and storage (i.e., Flash or RAM – Random access memory)
[112], the adaptation of transmission power depending on To ensure autonomy from an energy perspective, the
required communication range and environment [113, 114], energy harvesting elements in the IoT device’s structure must
or logic for deciding the moment and format for sending be capable of providing the energy required for its proper
slow varying data [115 - 117]. All these approaches for operation. Therefore, the prediction of the amount of energy
assuring low-power operation must be scheduled and further harvested and stored energy is computed using energy
modified depending also on the amount of energy that is harvesting and storage models. Next, we present a brief
generated or stored by the IoT device at each moment in time overview of energy harvesting models presented in the
[50]. Since in most cases of wireless IoT devices, literature.
communication is the most expensive operation from an In [128 - 131], the authors applied deterministic models
energy perspective [118], many of the research efforts focus for power management of the energy harvesting devices. The
on efficient communication protocols [118 - 122] and on the approach in [128] uses cooperative Automatic Repeat
optimization of the transmission, through careful planning Request (ARQ) protocols to balance the energy consumption
and simple data exchanges [115, 123, 124]. However, with of self-power devices to match their own battery recharge
the development of extremely low-power communication rate. The research has shown that cooperative ARQ enables
technologies, such as BLE [125], and the inclusion in the wireless sensors to efficiently transmit data using lower
design of IoT devices of many sensors, some of them with levels of energy than in non-cooperative protocols. In [129]
special requirements regarding timing and energy supply, and [130], the authors describe an optimal scheduling scheme
such as gas sensors, it is possible that sensing and data (power management algorithm) for IoT devices to maximize
processing consume the most energy in their design [4, 126]. the average data transmission rate and to maintain the energy
In these cases, algorithms for determining the optimal time neutrality of these devices over a day. The work in [131]
for sensor sampling, considering the amount of energy highlights an optimal scheduling scheme for an IoT device
scavenged and stored, must be developed for assuring energy that uses two energy-harvesting sources (solar and wind).
autonomy in the case of energy harvesting IoT devices [50]. The proposed algorithm optimally sets the overall IoT device
The calculation of the energy budget required by the IoT power consumption based on the utility and on the energy
device during operation is also of great importance in the required by tasks, and uses the weather forecast information
selection of a proper power source. The energy profile [104] available at the beginning of each day and the power level of
of the IoT device can be used for estimating the amount of the storage element.
power required for its proper operation. Figure 7 presents the Since the solutions developed based on the deterministic
power signature of an IoT device that measures the approach need accurate predictions of the system state, which
temperature, relative humidity, carbon dioxide level, absolute are not always applicable to the IoT ecosystem, attention has
pressure, and light intensity, and sends the data to a shifted to stochastic models capable to adapt the energy
predefined IP address using Wi-Fi connectivity and the User management to the uncertainty of the energy harvesting
Datagram Protocol (UDP) protocol [127]. By knowing the method. Many approaches [132 – 140] in this direction are
amount of energy consumed during all the activities found in the literature. Some of these approaches use the
performed by the device and their duration, the average Poisson process (which is a continuous-time Markov
energy required by the system can be computed. This process) [132] or the Markov model [133 - 136], but in
estimation can then be used for choosing strategies for recent years researchers have developed other techniques that
optimizations (i.e., modifying the sleep / wakeup ratio, would be suitable in the energy harvesting modeling process,
sampling sensors with different rates, etc.) and for properly such as the Gama process [137], the Gaussian model [138],
designing the energy harvesting elements. reinforcement learning methods [139], or the Kalman Filter
based model [140] as we discuss below.
VOLUME XX, 2021 9
The work in [136] introduces a stochastic model that uses enable the dimensioning of the storage unit depending on the
Markov Chain theory to estimate the performance of battery- energy requirements of the device in whose structure it is
less IoT devices (using a capacitor as storage element) for used. Furthermore, the design and performance evaluation of
uplink and downlink transmissions. The proposed model hybrid energy harvesting IoT devices is facilitated by the
takes as inputs the current consumption of different device predictability of energy harvesting systems parameters, such
states, the capacitance, the energy harvesting power, how as the amounts of harvested, combined, and stored energy,
often a transmission will take place, the probability of and the efficiency of the recharging process, that is achieved
receiving a downlink packet, the uplink and downlink packet in the proposed methodology. The authors of [139] proposed
size, and the Spreading Factor to be used. The Spreading a reinforcement learning algorithm for solving the access
Factor is the amount of spreading code applied to the original control and continuous power control problems in an energy
data signal and has values between 7 and 12 [141]. The harvesting IoT system having limited uplink access channels.
model determines the uplink Packet Delivery Ratio The energy harvesting IoT system analyzed is composed of
achievable and the probabilities of successfully receiving multiple user devices and one base station. Simulation results
downlink packets. The authors evaluated the performance of have shown that the method proposed achieves better
the proposed model in terms of the device configuration, performance than other existing approaches (i.e., quasi
application behavior, and environmental conditions. The binary power control, discrete power control, modified
results show that a capacitor of 47 mF can support a water-filling) in terms of throughput. The authors of [140]
Spreading Factor of 7 for uplink and downlink at an energy propose a platform suitable for BSNs. The BSN monitors and
harvesting rate of 1 mW, even when more than once per records the instantaneous usable power generated by
minute data transmissions are performed. A Spreading Factor wearable IoT devices with energy harvesting characteristics
of 11 can be supported by a supercapacitor of 1 F at an (solar and TEEH), while monitoring human activity and
energy harvesting rate of 10 mW. Also, the authors stated environmental data. To predict the amount of usable
that it is possible to achieve communication between battery- harvested energy based on environmental parameters (light
less IoT devices, but it must be pointed out that the turn-on intensity, temperature difference) and human behavior
voltage threshold of these devices significantly affects their (activity level), the authors developed and validated a
performance. In [137], the authors proposed a stochastic Kalman Filter based model. The Mean Absolute Percentage
model for a RF harvesting system that uses the gamma Error (MAPE) is used to compare the prediction
process to model the energy stored into the storage element performance. The value of this metric for the proposed model
(in this case a finite battery). They also used the renewal is better compared with other models (regression, moving
reward theory (a generalization of the Poisson process) to average, and exponential smoothing) used in the same testing
establish an optimal transmission policy to enhance the conditions.
operation of the RF harvesting system. In [138], the authors Table 2 summarizes the aforementioned modeling
presented single form accurate probabilistic energy models approaches proposed for different types of energy sources
for hybrid power harvesting IoT devices. These models and highlights the goals and the main results obtained.
TABLE 2. Summary of energy harvesting models.
Architecture of IoT device with

Energy source Approach Goals and results
energy harvesting characteristics
Solar [129, 130] Harvest-Store-Use Deterministic model - Maximization of data rate
(scheduling schemes) - Dimensioning of the energy storage element
Solar and wind [131] Harvest-Store-Use Deterministic model - Optimal scheduling scheme to establish the overall
(scheduling schemes) power consumption policy based on the utility and on
the energy required by tasks
Generally applied [136] Harvest-Store-Use Markov model - Performance evaluation of energy harvesting IoT
devices in terms of the device configuration
(capacitance and turn-on voltage threshold),
application behavior (data transmission interval, data
packet size) and environmental conditions (energy
harvesting rate)
RF [137] Harvest-Store-Use Gamma process - Mathematical model for storage energy policy
Renewal reward theory - Optimal transmission policy
Hybrid [138] Harvest-Store-Use, hybrid energy Gaussian Mixture model - Predictability of energy harvesting parameters
harvesting IoT device - Simple design, dimensioning and performance
evaluation of energy harvesting IoT devices
Generally applied [139] Energy harvesting user Reinforcement learning - Optimal control policy for access control and
equipment that includes batteries methods continuous power control
Solar and human heat [140] Harvest-Store-Use Kalman Filter based - Energy harvesting platform for human activity and
model environmental monitoring
- Prediction model of the amount of harvested energy
VOLUME XX, 2021 10
As Table 2 shows, energy harvesting systems adopt the

Harvest-Store-Use architecture in the case of unpredictable
or non-controllable energy sources. Therefore, energy
storage is of paramount importance in the design and
operation of energy harvesting IoT devices. As pointed
above, many energy harvesting designs assume that energy
generation is not constant, and can only be predicted, and
therefore the efficient management and proper use of the
energy stored in supercapacitors or batteries is required.
The two options for energy storage in energy harvesting
IoT devices include rechargeable batteries and
supercapacitors, each one of them having their own
advantages and limitations [142]. There are also systems that
include both batteries and supercapacitors, leading to energy
FIGURE 8. IoT scenario for use-case systems’ operation. Adapted from
supplies that respond better to high discharge pulses [143]. [127], [50].
Batteries are the main choice because of their high energy
density, but supercapacitors are more often considered due to A. SOLAR-POWERED ENVIRONMENTAL BLE BEACON
their high maximum recharging cycle lifetime, that can result
in longer service life of the device they are powering. CASE STUDY SYSTEM OVERVIEW
However, supercapacitors or electric double-layer capacitors, BLE is a wireless communication technology [146] that is
that are nothing more than extremely large value capacitors, being increasingly used in IoT applications because of its
cannot come close to the energy density offered by batteries, simplicity, ubiquity in most current electronic devices, and
and suffer from high self-discharge rates, of 30% as low power operation [147], [148]. Applications such as retail
compared to 5% encountered in batteries [103, 142]. [125], environmental monitoring [126], indoor localization
Modeling the behavior of supercapacitors has important [149], museum guiding [150], or inventory management
implications for the design of energy harvesting IoT devices [151] are using BLE beacons. The IoT sensor device
[144]. Efficient energy modeling of energy harvesting IoT presented in [127] measures light intensity, temperature, and
devices enables the design of optimal energy buffers for relative humidity of the environment. We have described the
achieving uninterrupted operation. This challenge of advantages and limitations posed by BLE devices advertising
designing proper energy buffers was previously solved by the acquired data to nearby monitoring applications in
over-dimensioning [145]. general and of this device in particular in our previous work
[127]. Here, we focus on the energy harvesting mechanism
V. CASE STUDY SYSTEMS that was used in the device’s architecture and the strategies
In this section, we present two IoT case study systems, in that were implemented for achieving energy autonomy. We
which the sensing devices are fitted with energy harvesting achieved energy independence even in the case when a
mechanisms helping them achieve energy autonomy. We modified variant of the IoT device was fitted with a CCS811
describe the implementation details of the hardware along gas sensor [50], [152]. It is a well-known fact that the use of
with considerations and lessons learned from the gas sensors in energy-constrained devices, such as wireless
corresponding design, implementation, and operation phases. sensors, is problematic due to their requirements in terms of
In the first case study, we used solar energy harvesting, while power and wake-up times. The work in [50] presents the use
the second case study is based on RF energy harvesting. Both of adaptive duty-cycling strategies that resulted in energy-
systems use environmental BLE beacons, that operate in an efficient autonomous operation of the IoT device.
IoT scenario (Figure 8), thus becoming IoT sensing devices
[127], [50]. Beacons are a class of wireless sensing systems HARDWARE ARCHITECTURE
that only transmit data to base stations which can be fixed or The architecture of the IoT device consists of a
mobile [116]. The IoT devices in the two case studies programmable BLE-enabled radio on chip component, the
(monitoring systems) broadcast the measured parameters, Cypress Semiconductor EZ-BLE PRoC Module [153], an
and BLE-enabled terminals (such as smartphones) receive SHT21 temperature (T) and relative humidity (RH) sensor
this data. The smartphones running monitoring applications [154], an MPL115A2, absolute pressure (P) sensor [155],
relay the data to the cloud through Internet links. Using cloud and a PT3001 digital ambient light sensor [156]. The power
computing resources, we can perform data processing, source includes a bq25504 ultra low-power boost converter
visualization, and prediction on the data stored in the cloud. with battery management for energy harvesting applications
[157], two IXYS KXOB22-04X3L amorphous silicon solar
cells connected in series [158], a single-cell coin
VOLUME XX, 2021 11
rechargeable battery (accumulator) [159], and two can be switched off between readings so that the overall
TPS78330DDCT low-dropout linear regulators (LDOs) power consumption of the system can be minimized.
[160]. Figure 9 shows theses various components and Figure
10 shows the integrated hardware. As the solar energy source OPTIMIZATIONS FOR ACHIEVING ENERGY AUTONOMY
cannot be controlled, the device is built using a generate- AND RESULTS
store-use architecture, where a rechargeable battery stores the Our initial work has shown that the BLE beacon can achieve
energy that is subsequently supplied to the electronics of the energy autonomy without requiring advanced power-
device. optimization techniques when the gas sensor is not
operational [127]. The setting of the processor’s clock
frequency to a value of 3 MHz, the readings of the attached
sensors (T, RH, P, light) once every minute, and the
transmission of an advertisement every 3 s resulted in a 37
μA average current drawn by the device. This value is
sufficiently low to allow stand-alone operation, with the solar
cells being capable of charging the accumulator within a
single day with good weather. In this scenario, the sensors
are powered only once a minute, when they are read, and the
resulting data is used to update the advertisement packet. The
control module spends almost all the time in sleep mode and
wakes up once every three seconds for 3 ms to transmit the
advertisement packet, and 140 ms every minute for
communicating with the attached sensors.
The addition of the CCS811 gas sensor in the design of the
IoT device for providing data regarding the eCO2 (equivalent
CO2) and Total Volatile Organic Compounds (TVOC) levels
in the air required the use of more complex mechanisms for
achieving autonomy from an energy perspective [50]. These
firmware optimization mechanisms were needed because of
the higher power consumption and requirement for longer
wake-up periods of the gas sensor. We implemented adaptive
FIGURE 9. Architecture of solar-powered BLE beacon [127], [50].
duty-cycling strategies in the design of the firmware running
on the IoT device considering the light intensity and the
amount of energy stored in the accumulator. We incorporate
these strategies to lower the overall power consumption of
the IoT device. Furthermore, the advertising interval also had
to be increased for reducing the overall energy consumption
of the IoT device. Thus, we used an interval of 5 s instead of
the previous 3 s interval. The IoT device consumes more
energy when the attached sensors are active and take
measurements than when the transmission of advertisements
of the collected data takes place. Therefore, the sensors in the
structure of the IoT device are sampled with different rates,
depending on their power consumption and on the voltage
value estimated for the rechargeable battery and as well on
FIGURE 10. Manufactured solar-powered BLE beacon (CCS811 not
populated). the light intensity. Thus, the T, RH, and the light intensity
sensors are powered up and read every minute, when the
To enable better control over the power consumption of battery voltage is also measured and computed using a
the device, two TPS78330DDCT Low-DropOut linear special circuit in the design as described in [50]. If the
regulators (LDOs) were used for providing power to the intensity of light and the battery level are above some
processing and sensing parts of the device. Therefore, the thresholds, the wake-up period is prolonged, and the gas
firmware can control the power source for both the central sensor performs the reading operation. Otherwise, the
part of the device and the sensing component. As the beacon reading of the CCS811 sensor is postponed until a future
periodically samples the sensors in its structure, processes the wake-up period with favorable conditions (battery level and
acquired data, and broadcasts it, the sensors’ power supply light intensity are above previously set thresholds) takes
place. By choosing appropriate values for the thresholds and
VOLUME XX, 2021 12
for the maximum number of delay periods for the reading of capacitor is connected for storing the generated energy. The
the gas sensor, efficient activity planning that enables energy P1110 evaluation board generates energy from signals
autonomy can be reached despite unfavorable weather having frequencies in the range of 850–950 MHz and
conditions (presence of clouds) and of the day/night cycle, as charges the on-board supercapacitor powering the BLE-
the results in [50] have shown. We achieve this result by enabled IoT device. Although a capacitor assuring many
carefully dimensioning (size of the solar cells in the structure charging-discharging cycles is used instead of a rechargeable
of the IoT device) the energy harvesting circuits and by battery, it suffers from having high leakage current, with
efficient hardware and software designs (separate power values (5–10 μA) comparable to the sleep current (~1–10
source for the sensors, battery level measurement circuit, μA) of wireless sensor nodes. In the setup presented here, the
adaptive duty-cycling strategy). 50 mF capacitor discharges itself from 3.3 V to 2.3 V in less
than 5 hours. A Global System for Mobile Communications
B. RF ENERGY HARVESTING ENVIRONMENTAL BLE (GSM) phone operating within a band close to the one
BEACON required by the P1110 module (2G technology) placed at 5
cm from the harvester antenna is able to charge the capacitor
CASE STUDY SYSTEM OVERVIEW to 3.3 V in less than three minutes. By knowing these values
In the second use case system, we replaced the power source and the energy profile of the powered IoT device, an
of the BLE beacon by an RF energy harvesting element as estimation of the rate at which the energy harvesting
Figure 11 shows. The IoT device in this case is also based on mechanism should charge the energy storage element can be
a harvest-store-use architecture. Figure 11 shows the block computed, and adjustments to the design of the entire system
diagram of the system [4]. can be performed. We have presented a detailed description
of the design and the experimental results obtained in [4].
OPTIMIZATIONS FOR ACHIEVING ENERGY AUTONOMY

AND RESULTS
The firmware on the IoT device implements a simple duty-
cycled system where the attached sensors are read every 30
seconds and the advertisement period is set to 3 seconds.
Since we wanted to test the feasibility of using RF energy
harvesting for such a system, no other optimizations were
performed on the application running on the central unit of
the IoT device. Our experimental results demonstrated that
the 30 cm2 antenna of the P1110 RF Powerharvester can
deliver the power required by the IoT device that samples the
attached sensors every half minute and sends an
advertisement packet every 3 seconds, when a 2G GSM
phone closer than 10 cm is in an active call. However, this
was made possible after finding the optimal distance between
FIGURE 11. Architecture of RF-powered BLE beacon [4].
the energy source and the energy harvesting component and
the proper alignment between the two. The supercapacitor in
The resulting IoT device operates in a similar way as the
the design is another challenge because of its high self-
previous system we have described in the previous section.
discharge current, and should be replaced by other, more
However, due to the smaller amounts of scavenged energy
efficient, energy storage elements. Another challenge when
using RF harvesting elements, it was not possible to power
designing such circuits is represented by the efficient
on the CCS811 gas sensor. Since the generated energy
measurement of the signals involved. For example, the
cannot sustain the operation of the CCS811 gas sensors, the
current drawn by the BLE beacon varies in a very short
RF energy harvesting BLE beacon presented can measure T,
amount of time between microamperes and tens or hundreds
RH, light intensity, and atmospheric pressure.
of milliamperes. This quick variation requires measurement
equipment with wide input ranges and large sampling rates.
HARDWARE ARCHITECTURE
However, probably the biggest challenge when designing RF
As Figure 11 shows, the power supply of the first use case
energy harvesting IoT devices is in the miniaturization of the
system was replaced by an RF energy harvesting circuit that
hardware. IoT-enabled sensors should have small form
charges a supercapacitor. Since the CCS811 gas sensor
factors and should be low cost, while better RF energy
requires energy levels that cannot be provided by the energy
systems require larger antennas and more complex circuitry.
harvesting circuit and longer wakeup periods, it was removed
Although the work presented in [4] demonstrated that RF
from the IoT device. The power supply consists of a P1110
energy harvesting could generate enough power for certain
Powerharvester charging module [161], to which a 50 mF
VOLUME XX, 2021 13
IoT devices, this is possible only when specific conditions significant savings in terms of energy and of materials and
(such as an active source in the immediate proximity of the efforts used for the manufacturing and disposal of batteries.
system, signal in the frequency range required by the
harvester) are met. SOFTWARE
Algorithms for optimizing the energy consumption can
C. DISCUSSIONS AND LESSONS LEARNT FROM THE overcome problems such as the non-controllability of the
CASE STUDY SYSTEMS ENERGY HARVESTING energy source. The firmware developed for the first use case
DESIGNS system, that generates energy from sunlight, manages to
The development and analysis of the operation of the IoT adapt the operation of the IoT device and reduce the energy
devices presented here identified several design issues that used during periods when the conditions are not favorable
could be useful for researchers focusing on energy harvesting (cloudy weather, nighttime, small amount of energy stored in
IoT devices’ implementations. the accumulator). Software plays a major role in the
prediction of harvested energy in the case of solar energy,
HARDWARE while in the case of RF energy harvesting, hardware seems to
The hardware design of the two case study systems be more important in optimizing the operation of the IoT
demonstrated that significant energy savings can be achieved device. The use of an algorithm for selectively sending data
by using simple methods. One is the use of a separate power depending on its variation (i.e., send more rarely slow
source for the sensors attached to the IoT device. This way, changing data) would increase energy efficiency in both
the software has control over the operation of the sensors and cases.
can efficiently schedule their active periods. This design
feature enabled the solar powered IoT device to adapt the COST
reading of the attached sensors depending on the amount of Our case study systems show that energy harvesting leads to
energy generated and on the level of the accumulator. an increase of the cost of an IoT device of 15 to 20 Euros.
Another useful energy saving feature is the lowering of the However, in many designs, energy harvesting could provide
operational frequency of the microcontroller. The frequency at least the energy required for sleep mode periods and for
value should be close to the one that makes the static leakage currents. If we assume a sleep current between 3 and
component of the current drawn by the IoT device the main 10 μA, this could lead to savings of 26.3 to 87.6 mAh for the
consumer, and not the dynamic one. Reducing the operating main energy source, that could be a battery. For an energy
frequency from 48 MHz to 12 MHz did not significantly efficient IoT device, this could mean doubling of its lifetime,
increase the data processing time, but considerably lowered from 5 to 10 years.
the power consumption.
VI. TECHNICAL CHALLENGES
ENERGY HARVESTER Even if have witnessed recent advances in the design and
There is a category of energy harvesting methods which can development of energy harvesting devices over the past
provide sufficient energy for prolonged periods of time and decade, several technological challenges still need to be
also offer the advantage of miniaturization. These include addressed before the manufacturing of self-sustainable IoT
solar cells and piezoelectric circuits. On the other hand, there devices becomes prevalent. Some of them are:
are energy harvesting mechanisms that generate small • Harvested energy modeling. A balance between the
amounts of energy at discrete moments in time, while also generated power and the consumed power must be
occupying more space. An example of such a method is RF maintained. This implies the efficient power profiling of
energy harvesting. This represents a major challenge in the IoT devices and the adaptation of their operation to the
development of energy harvesting IoT devices, since these amount of harvested energy. The availability of the
must be small and based on simple designs. Currently, these harvested energy varies mostly with time in a non-
requirements seem to be fulfilled in a more facile way by the deterministic manner. Therefore, the estimation of the
first category of energy harvesting mechanisms (solar cells amount of energy scavenged is computed using
and piezoelectric circuits). One option for using RF energy prediction techniques and conventional power
harvesting could be to embed the antenna into objects such as management approaches (i.e., Maximum Power Point
clothes in the case of wearable IoT devices. Tracking (MPPT) and software Phase Locked Loops
An energy budget of between 50 and 60 μAh generated by (PLL)) are applied to manage the power coming from
an energy harvesting circuit could assure autonomous the energy sources. Recently proposed energy
operation from an energy perspective, without the need of forecasting models should be improved to provide
using a battery as the main power supply, for a wide range of accurate results, while the power management choices
IoT devices. Given that we expect billions of IoT devices should be made to minimize the loss of energy.
deployed in coming years, energy harvesting could provide Additionally, the power source must provide enough
energy for the following tasks: data processing
VOLUME XX, 2021 14
operations, transmission and/or receive actions, and • Size and cost efficiency. There are situations where the
sleep periods. In most cases, it is not the data processing size and the weight of IoT devices are critical (i.e.,
that is the most energy-demanding task. It is the wearable and implantable IoT devices). But these
transmission and reception of data over the wireless devices produce a small amount of energy which is not
Internet. Therefore, future researchers should develop enough to be used to perform their main functions (i.e.,
optimized consumption models to minimize the energy powering the device and the attached sensors, data
cost during wireless data transmissions and investigate transmission). Small-scale harvesting solutions (micro,
novel techniques to adapt the wireless communication nano) that can power IoT devices and support the
protocols according to the energy harvesting process’ operation of other functions (i.e., monitoring the health
characteristics. status of patients, acting as stimulators for regenerating
• Harvested energy storage. This involves the tissues) must be developed considering the low cost of
development of suitable storage elements such as fabrication. The scientific literature emphasizes that
rechargeable batteries and supercapacitors because the PEHs can be effectively used for powering small and
technology used for storing the harvested energy affects very-small IoT devices. Therefore, future researchers
the cost, size, and the operating life of the IoT devices. should develop new eco-friendly materials to enable
Batteries have high energy densities, but they are not micro and nanofabrication of PEHs with improved
well suited for long-life IoT devices due to the cycling flexibility and output power density. Recent advances in
degradation phenomenon [162]. Moreover, both high the field of microelectronics are promising and can be
and low temperatures reduce their capabilities. used to develop robust, miniaturized, low power, and
Supercapacitors have lower energy density than low-cost energy harvesters.
batteries, but cyclic degradation does not affect them. • Environmental impact with renewable energy sources.
Furthermore, supercapacitors suffer from increased Renewable energy sources help to mitigate
current leakage that would consume a large part of the environmental pollution, and thus, they are used to
harvested energy. Therefore, future researchers should develop new IoT devices because this industry has
investigate new techniques that can find the best experienced a significant growth in the past few years.
candidate for harvested energy storage which must meet Batteries used in IoT devices without energy harvesting
the following criteria: low cycling degradation, low mechanisms eventually get depleted and, in some cases,
current leakage, high energy density, and continued they are thrown away in several weeks or months. If
operation even in harsh environments such as at very there are no battery recycling mechanisms, then the
low and very high temperatures. environment suffers. The challenge in this case is the
• Energy harvesting from multiple sources. There are development of energy harvesting IoT devices with a
cases wherein a single source of energy harvesting is lifetime significantly longer than the one provided by
insufficient to power IoT devices. By combining energy batteries. Also, it is worth noting that some energy
from multiple sources, the reliability of IoT devices can harvesting IoT devices employ toxic or rare materials
be increased. The work in [163] presents the recent (i.e., bismuth telluride for TEEHs, lead zirconate titanate
developed architectures and techniques of low power for PEHs, cadmium for PV). Therefore, the use of eco-
management circuits that use energy harvested from friendly materials, such as electroactive polymers,
multiple heterogeneous sources of energy. The analysis carbon nanowire semiconductors, to design the
highlights that the proposed architectures are suitable for electronics components of the energy harvesting IoT
specific applications. For example, the complementary devices is another challenge that must be addressed.
use of harvesters or the Power ORing topology are Biodegradable and biocompatible IoT devices must be
simple schemes appropriate in cases where it is not considered by developers of such devices for a
expected that all the input energy sources deliver a sustainable future.
significant amount of power at the same time. The
multiple input switched-inductor and switched-capacitor VII. CONCLUSION
converter architectures are also used to combine energy Energy harvesting has been receiving a lot of attention by
for heterogeneous sources. Regardless of the chosen various research communities and industry involved in the
architecture, designers must consider the development of design and implementation of self-powered IoT devices. In
configurable impedance matching schemes for the this paper, we analyzed the energy harvesting technologies
purpose of MPPT control. Researchers should also focus used primarily in the IoT environment. We conclude that
their attention on the development of intelligent some energy harvesting technologies can provide a
algorithms capable of selecting the input sources of significant amount of energy for a long time using small size
energy depending on their availability thereby PV cells or piezoelectric devices. In contrast, other energy
eliminating the need for the energy storage element. harvesting techniques provide a small amount of energy at
discrete moments of time and require large circuits for
VOLUME XX, 2021 15
capturing the energy, but they do not depend on certain TVOC Total Volatile Organic Compounds
cycles such as day/night, working days/weekends, nor are UDP User Datagram Protocol
they easy to shield (i.e., RF). The appropriate energy WSNs Wireless Sensor Networks
harvesting technique depends mainly on the parameter to be
measured, the use case scenario of the IoT device REFERENCES
(fixed/mobile, surface/built-in), but also on its location [1] International Telecommunication Union (ITU), “ITU Internet
Reports 2005: The Internet of Things, Executive Summary,”
(indoor/outdoor). To demonstrate the potential of solar and International Telecommunication Union (ITU), Geneva, Nov. 2005.
RF energy sources, we described two IoT case study systems [2] K. L. Lueth, “State of the IoT 2020,” IoT Analytics, 19 Nov. 2020.
from our previous work with harvesting mechanisms in terms [Online]. Available: https://iot-analytics.com/state-of-the-iot-2020-
12-billion-iot-connections-surpassing-non-iot-for-the-first-time/.
of their design, hardware implementation and operation. [Accessed 29 Dec. 2020].
Finally, we have discussed some future technical challenges, [3] Y. Khan, “5 Essential Components of an IoT Ecosystem,” 21 Apr.
whose addressing will facilitate the large-scale deployment 2020. [Online]. Available: https://learn.g2.com/iot-ecosystem.
[Accessed 11 Nov. 2020].
of energy harvesting solutions for IoT. [4] T. Sanislav, S. Zeadally, G. D. Mois, and S. C. Folea, “Wireless
energy harvesting: Empirical results and practical considerations for
LIST OF ACRONYMS Internet of Things,” Journal of Network and Computer Applications,
AES Advanced Encryption Standard vol. 121, pp. 149-158, 2018.
[5] S. Zeadally, A. Das, and N. Sklavos, “Cryptographic Technologies
AI Artificial Intelligence and Protocol Standards for Internet of Things,” Internet of Things:
ARQ Automatic Repeat Request Engineering Cyber Physical Human Systems, in press.
BLE Bluetooth Low Energy [6] A. Das, S. Zeadally, and D. He, “Taxonomy and Analysis of Security
Protocols for Internet of Things,” Future Generation Computer
BSN Body Sensor Node
Systems, vol. 89, 2018.
DC Direct Current [7] R. Yugha and S. Chithra, “A survey on technologies and security
DES Data Encryption Standard protocols: Reference for future generation IoT,” Journal of Network
DVS Dynamic Voltage Scaling and Computer Applications, vol. 169, p. 102763, 2020.
[8] J.-M. Dilhac and V. Boitier, “1 - Wireless Sensor Networks,” in
ECG ElectroCardioGram Energy Autonomy of Batteryless and Wireless Embedded Systems,
eCO2 Equivalent CO2 J. Dilhac and V. Boitier, Eds., Elsevier, 2016, pp. 1-11.
EEG ElectroEncephaloGram [9] E. Suganya, S. Sountharrajan, K. S. Shandilya, and M. Karthiga,
“Chapter 5 - IoT in Agriculture Investigation on Plant Diseases and
EH Energy Harvesting Nutrient Level Using Image Analysis Techniques,” in Internet of
EMEH ElectroMagnetic Energy Harvester Things in Biomedical Engineering, V. E. Balas, L. H. Son, S. Jha, M.
EMG ElectroMyoGram Khari and R. Kumar, Eds., Academic Press, 2019, pp. 117-130.
[10] T. Sanislav, G. Mois, S. Folea, and L. Miclea, “Integrating wireless
ESEH ElectroStatic Energy Harvester
sensor networks and cyber-physical systems: challenges and
grms Root Mean Square acceleration opportunities,” in Control, Robotics & Sensors, 2016, Cyber-
GSM Global System for Mobile Communications Physical system design with sensor networking technologies, IET
ICs Integrated Circuits Digital Library, 2016, pp. 47-76.
[11] O. Bello and S. Zeadally, “Toward Efficient Smartification of the
ICTs Information and Communication Technologies Internet of Things (IoT) Services,” Future Generation Computer
IoT Internet of Things Systems, vol. 92, 2019.
IP Internet Protocol [12] M.T. Lazarescu, “Wireless Sensor Networks for the Internet of
Things: Barriers and Synergies,” in Components and Services for IoT
ITU International Telecommunications Union
Platforms, G. Keramidas, N. Voros, and M. Hübner, Eds., Springer,
LDOs Low-DropOut linear regulators Cham, 2017, DOI: 10.1007/978-3-319-42304-3_9.
MAC Medium Access Control [13] P.S. Mathew, A.S. Pillai, and V. Palade, “Applications of IoT in
Healthcare,” in Cognitive Computing for Big Data Systems Over IoT.
MAPE Mean Absolute Percentage Error
Lecture Notes on Data Engineering and Communications
MCU MicroController Unit Technologies, A. Sangaiah, A. Thangavelu, and V. Meenakshi
MPPT Maximum Power Point Tracking Sundaram, Eds., vol 14. Springer, Cham, 2018, DOI: 10.1007/978-3-
P Pressure 319-70688-7_11.
[14] K. Haseeb, I. U. Din, A. Almogren, and N. Islam, “An Energy
PEEH PyroElectric Energy Harvester Efficient and Secure IoT-Based WSN Framework: An Application to
PEH Piezoelectric Energy Harvester Smart Agriculture,” Sensors, vol. 20(7), 2081, 2020 DOI:
PLL software Phase Locked Loops 10.3390/s20072081.
[15] B. Hammi, R. Khatoun, S. Zeadally, A. Fayad, and L. Khoukhi,
PV PhotoVoltaic
“Internet of Things (IoT) Technologies for Smart Cities,” IET
QoS Quality of Service Networks, vol. 7, no. 1, 2018.
RAM Random Access Memory [16] M. Jia, A. Komeily, Y. Wang, and R. S. Srinivasan, “Adopting
Internet of Things for the development of smart buildings: A review
RF Radio Frequency
of enabling technologies and applications”, Automation in
RFEH RF Energy Harvester Construction, vol. 101, pp. 111-126, 2019, DOI:
RH Relative Humidity 10.1016/j.autcon.2019.01.023.
RSA Rivest, Shamir, and Adelman [17] C. Yang, W. Shen, and X. Wang, “The Internet of Things in
Manufacturing: Key Issues and Potential Applications,” IEEE
T Temperature Systems, Man, and Cybernetics Magazine, vol. 4, no. 1, pp. 6-15,
TEEH ThermoElectric Energy Harvester Jan. 2018, DOI: 10.1109/MSMC.2017.2702391.
VOLUME XX, 2021 16
[18] J. Contreras-Castillo, S. Zeadally, and J. Guerrero, “Sensor Energy. Energy, Environment, and Sustainability, H. Tyagi, A.
Technologies for Intelligent Transportation,” Sensors, vol. 18, no. 4, Agarwal, P. Chakraborty, and S. Powar, Eds., Singapore, Springer,
2018. 2018, pp. 163-186.
[19] L. Muduli, D. P. Mishra, and P. K. Jana, “Application of wireless [36] M. Prauzek, J. Konecny, M. Borova, K. Janosova, J. Hlavica, and P.
sensor network for environmental monitoring in underground coal Musilek, “Energy Harvesting Sources, Storage Devices and System
mines: A systematic review,” Journal of Network and Computer Topologies for Environmental Wireless Sensor Networks: A
Applications, vol. 106, pp. 48-67, 2018. Review,” Sensors, vol. 18, no. 2446, pp. 1-22, 2018.
[20] R. Vera-Amaro, M. E. R. Angeles, and A. Luviano-Juarez, “Design [37] H. Wei, H. Zhang, H. Sun, and B. Yang, “Preparation of polymer–
and Analysis of Wireless Sensor Networks for Animal Tracking in nanocrystals hybrid solar cells through aqueous approaches,” Nano
Large Monitoring Polar Regions Using Phase-Type Distributions and Today, vol. 7, no. 4, pp. 316-326, 2012.
Single Sensor Model,” IEEE Access, vol. 7, pp. 45911-45929, 2019. [38] J. Yick, B. Mukherjee, and D. Ghosal, “Wireless sensor network
[21] M. Farsi, M. A. Elhosseini, M. Badawy, H. Arafat Ali, and H. Zain survey,” Computer Networks, vol. 52, no. 12, pp. 1389-1286, 2008.
Eldin, “Deployment Techniques in Wireless Sensor Networks, [39] Y. Wang, P. Luo, X. Zeng, D. Peng, Z. Li, and B. Zhang, “A Neural
Coverage and Connectivity: A Survey,” IEEE Access, vol. 7, pp. Network Assistance AMPPT Solar Energy Harvesting System With
28940-28954, 2019. 89.39% Efficiency and 0.01–0.5% Tracking Errors,” IEEE
[22] J. Gabay, “Wireless Links Inside Rotating Machinery,” Digi-Key Transactions on Circuits and Systems I: Regular Papers, vol. 67, no.
Electronics, 27 Mar. 2014. [Online]. Available: 9, pp. 2960-2971, 2020.
https://www.digikey.com/en/articles/wireless-links-inside-rotating- [40] S. Cheong, M. Kang, Y. Kim, M. Park, J. Park, and D. Noh, “Solar-
machinery. [Accessed 23 Dec. 2020]. CTP: An Enhanced CTP for Solar-Powered Wireless Sensor
[23] M. Hulea, G. Mois, S. Folea, L. Miclea and V. Biscu, "Wi-sensors: A Networks,” IEEE Access, vol. 8, pp. 127142-127155, 2020.
low power Wi-Fi solution for temperature and humidity [41] C. Wang, J. Li, Y. Yang, and F. Ye, “Combining Solar Energy
measurement," IECON 2013 - 39th Annual Conference of the IEEE Harvesting with Wireless Charging for Hybrid Wireless Sensor
Industrial Electronics Society, Vienna, Austria, 2013, pp. 4011-4015, Networks,” IEEE Transactions on Mobile Computing, vol. 17, no. 3,
DOI: 10.1109/IECON.2013.6699777. pp. 560-576, 2018.
[24] F. Shaikh and S. Zeadally, “Energy harvesting in wireless sensor [42] J. Mohtasham, “Review Article-Renewable Energies,” Energy
networks: A comprehensive survey,” Renewable and Sustainable Procedia, vol. 74, pp. 1289-1297, 2015.
Energy Reviews, vol. 55, pp. 1041-1054, 2016. [43] M. Kumar, “Enhanced Solar PV Power Generation Under PSCs
[25] S. Ma, M. Jiang, P. Tao, C. Song, J. Wu, J. Wang, T. Deng, and W. Using Shade Dispersion,” IEEE Transactions on Electron Devices,
Shang, “Temperature effect and thermal impact in lithium-ion vol. 67, no. 10, pp. 4313-4320, 2020.
batteries: A review,” Progress in Natural Science: Materials [44] Q. Liu and Q. Zhang, “Accuracy Improvement of Energy Prediction
International, vol. 28, no. 6, pp. 653-666, 2018. for Solar-Energy-Powered Embedded Systems,” IEEE Transactions
[26] Texas Instruments Incorporated, “Using Input Current Limiting to on Very Large Scale Integration (VLSI) Systems, vol. 24, no. 6, pp.
Extend Battery Life,” Texas Instruments Incorporated, Dallas, TX, 2062-2074, 2016.
USA, Apr. 2020. [Online]. Available: [45] M. Hassan and A. Bermak, "Solar Harvested energy prediction
https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web& algorithm for wireless sensors," 2012 4th Asia Symposium on Quality
cd=&cad=rja&uact=8&ved=2ahUKEwid_J7kkuTtAhWnAxAIHcn4 Electronic Design (ASQED), Penang, Malaysia, 2012, pp. 178-181,
CMMQFjAJegQIHBAC&url=http%3A%2F%2Fwww.ti.com%2Flit DOI: 10.1109/ACQED.2012.6320497.
%2Fpdf%2FSLVAES7&usg=AOvVaw1EWJ0- [46] F. A. Kraemer, D. Palma, A. E. Braten, and D. Ammar,
zYdgxGJUzt9CUhUC. [Accessed 23 Dec. 2020]. “Operationalizing Solar Energy Predictions for Sustainable,
[27] A. Dionisi, D. Marioli, E. Sardini, and M. Serpelloni, “Autonomous Autonomous IoT Device Management,” IEEE Internet of Things
Wearable System for Vital Signs Measurement With Energy- Journal, vol. 7, no. 12, pp. 11803-11814, 2020.
Harvesting Module,” IEEE Transactions on Instrumentation and [47] J. Gruetter, “Solar Energy Harvesting - Low Power in a Compact
Measurement, vol. 65, no. 6, pp. 1423-1434, 2016. Footprint,” WWW.POWERSYSTEMSDESIGN.COM, Oct. 2010.
[28] Perchards, “The Collection of Waste Portable Batteries in Europe in [Online]. Available:
View of the Achievability of the Collection Targets Set by Batteries https://www.powersystemsdesign.com/articles/solar-energy-
Directive 2006/66/EC,” PERCHARDS LIMITED, Hertfordshire, harvesting/22/5742. [Accessed 3 Jan. 2021].
United Kingdom, Dec 2016. [Online]. Available: [48] Y. Wu, C. Li, Z. Tian, and J. Sun, “Solar-driven integrated energy
https://www.epbaeurope.net/wp- systems: State of the art and challenges,” Journal of Power Sources,
content/uploads/2016/12/Reportontheportablebatterycollectionrates- vol. 478, p. 228762, 2020.
UpdateDec-15-ExerptofChanges.pdf. [Accessed 23 Dec. 2020]. [49] H. Akinaga, “Recent advances and future prospects in energy
[29] R. Atat, L. Liu, J. Wu, G. Li, C. Ye, and Y. Yang, “Big Data Meet harvesting technologies,” Japanese Journal of Applied Physics, vol.
Cyber-Physical Systems: A Panoramic Survey,” IEEE Access, vol. 6, 59, no. 11, p. 110201, 2020.
pp. 73603-73636, 2018. [50] G. Mois, T. Sanislav, S. Folea, and S. Zeadally, “Performance
[30] N. Soin, “Chapter 10 - Magnetic Nanoparticles—Piezoelectric Evaluation of Energy-Autonomous Sensors Using Power-Harvesting
Polymer Nanocomposites for Energy Harvesting,” in In Micro and Beacons for Environmental Monitoring in Internet of Things (IoT),”
Nano Technologies, Magnetic Nanostructured Materials, Elsevier, Sensors, vol. 18, no. 1709, pp. 1-21, 2018.
pp. 295-322, 2018. [51] X. Liu and E. Sánchez-Sinencio, “A Highly Efficient Ultralow
[31] D. Altinel and G. K. Kurt, “Modeling of Hybrid Energy Harvesting Photovoltaic Power Harvesting System With MPPT for Internet of
Communication Systems,” IEEE Transactions on Green Things Smart Nodes,” IEEE Transactions on Very Large Scale
Communications and Networking, vol. 3, no. 2, pp. 523-534, 2019. Integration (VLSI) Systems, vol. 23, no. 12, pp. 3065-3075, 2015.
[32] S. Sudevalayam and P. Kulkarni, “Energy Harvesting Sensor Nodes: [52] F. Wu, J. Redouté and M. R. Yuce, “WE-Safe: A Self-Powered
Survey and Implications,” IEEE Communications Surveys & Wearable IoT Sensor Network for Safety Applications Based on
Tutorials, vol. 13, no. 3, pp. 443-461, 2011. LoRa,” IEEE Access, vol. 6, pp. 40846-40853, 2018.
[33] C. S. Psomopoulos, “Solar Energy: Harvesting the Sun's Energy for [53] G. Mois, Z. Szilagyi, T. Sanislav and S. Folea, "An HTTP-based
Sustainable Future,” in Handbook of Sustainable Engineering, J. environmental monitoring system using power harvesting," 2017
Kauffman and K. Lee, Eds., Dordrecht, Springer Netherlands, 2013, 21st International Conference on System Theory, Control and
pp. 1065-1107. Computing (ICSTCC), Sinaia, 2017, pp. 845-848, DOI:
[34] P. Breeze, “Chapter 13 - Solar Power,” in Power Generation 10.1109/ICSTCC.2017.8107142.
Technologies (Third Edition), P. Breeze, Ed., Newnes, 2019, pp. [54] C. Yang, R. Xue, X. Li, X. Zhang, and Z. Wu, “Power performance
293-321. of solar energy harvesting system under typical indoor light sources,”
[35] P. Saini, D. V. Patil and S. Powar, “Review on Integration of Solar Renewable Energy, vol. 161, pp. 836-845, 2020.
Air Heaters with Thermal Energy Storage,” in Applications of Solar
VOLUME XX, 2021 17
[55] S. Biswas, Y.-J. You, Y. Lee, J. W. Shim, and H. Kim, “Efficiency [72] H. Elahi, M. Eugeni, P. Gaudenzi, “A Review on Mechanisms for
improvement of indoor organic solar cell by optimization of the Piezoelectric-Based Energy Harvesters,” Energies, vol. 11, no. 7:
doping level of the hole extraction layer,” Dyes and Pigments, vol. 1850, 2018.
183, p. 108719, 2020. [73] H. Elahi, K. Munir, M. Eugeni, S. Atek, and P. Gaudenzi, “Energy
[56] C. S. Abella, S. Bonina, A. Cucuccio, S. D’Angelo, G. Giustolisi, A. Harvesting towards Self-Powered IoT Devices,” Energies, vol. 13,
D. Grasso, A. Imbruglia, G. S. Mauro, G. A. M. Nastasi, G. no. 21: 5528, 2020.
Palumbo, S. Pennisi, G. Sorbello, and A. Scuderi, “Autonomous [74] C. Covaci and A. Gontean, “Piezoelectric Energy Harvesting
Energy-Efficient Wireless Sensor Network Platform for Solutions: A Review,” Sensors, vol. 20, no. 12: 3512, 2020.
Home/Office Automation,” IEEE Sensors Journal, vol. 19, no. 9, pp. [75] J. Y. Cho et al., “A multifunctional road-compatible piezoelectric
3501-3512, 2019. energy harvester for autonomous driver-assist LED indicators with a
[57] T. V. Tran and W. Chung, “High-Efficient Energy Harvester With self-monitoring system,” Applied Energy, vol. 242, pp. 294-301,
Flexible Solar Panel for a Wearable Sensor Device,” IEEE Sensors 2019, DOI: 10.1016/j.apenergy.2019.03.075.
Journal, vol. 16, no. 24, pp. 9021-9028, 2016. [76] S. Bradai, G. Bouattour, S. Naifar, and O. Kanoun, “Electromagnetic
[58] D. Newell and M. Duffy, “Demonstrating the Scope of a Switched Energy Harvester for Battery-Free IoT Solutions,” in 2020 IEEE 6th
Supercapacitor Circuit for Energy Harvesting Powered Wireless World Forum on Internet of Things (WF-IoT), New Orleans, LA,
Sensor Loads,” IEEE Transactions on Power Electronics, vol. 34, no. USA, 2020, pp. 1-5, DOI: 10.1109/WF-IoT48130.2020.9221051.
8, pp. 7532-7541, 2019. [77] A. Kumar, S.S. Balpande, and S.C. Anjankar, “Electromagnetic
[59] R. A. Kjellby, T. E. Johnsrud, S. E. Loetveit, L. R. Cenkeramaddi, Energy Harvester for Low Frequency Vibrations Using MEMS,”
M. Hamid and B. Beferull-Lozano, "Self-Powered IoT Device for Procedia Computer Science, vol. 79, 2016, pp. 785-792, DOI:
Indoor Applications," 2018 31st International Conference on VLSI 10.1016/j.procs.2016.03.104.
Design and 2018 17th International Conference on Embedded [78] F. U. Khan and M. Iqbal, “Electromagnetic Bridge Energy Harvester
Systems (VLSID), Pune, India, 2018, pp. 455-456, doi: Utilizing Bridge’s Vibrations and Ambient Wind for Wireless Sensor
10.1109/VLSID.2018.110. Node Application,” Journal of Sensors, vol. 2018, article ID
[60] M. Hassanalieragh, T. Soyata, A. Nadeau, and G. Sharma, “UR- 3849683, 2018, DOI: 10.1155/2018/3849683.
SolarCap: An Open Source Intelligent Auto-Wakeup Solar Energy [79] Y. Naito and K. Uenishi, “Electrostatic MEMS Vibration Energy
Harvesting System for Supercapacitor-Based Energy Buffering,” Harvesters inside of Tire Treads,” Sensors, vol. 19, no. 4: 890, 2019.
IEEE Access, vol. 4, pp. 542-557, 2016. [80] Y. Lu, F. Cottone, S. Boisseau, F. Marty, D. Galayko, and P. Basset,
[61] L. G. Tran, H. K. Cha, and W.T. Park, “RF power harvesting: a “Low-frequency and ultra-wideband MEMS electrostatic vibration
review on designing methodologies and applications,” Micro and energy harvester powering an autonomous wireless temperature
Nano Syst Lett, vol. 5, art. no. 14, 2017, DOI: 10.1186/s40486-017- sensor node,” in 2016 IEEE 29th International Conference on Micro
0051-0. Electro Mechanical Systems (MEMS), Shanghai, 2016, pp. 33-36,
[62] M. Cansiz, D. Altinel, and G. K. Kurt, “Efficiency in RF energy DOI: 10.1109/MEMSYS.2016.7421550.
harvesting systems: A comprehensive review,” Energy, Volume 174, [81] S. Lee, T. Kang, W. Lee, et al. “Multifunctional Device based on
pp. 292-309, 2019, DOI: 10.1016/j.energy.2019.02.100. phosphor-piezoelectric PZT: lighting, speaking, and mechanical
[63] O. Assogba, A. K. Mbodji, and A. K. Diallo, “Efficiency in RF energy harvesting,” Sci Rep, vol. 8, no. 301, 2018, DOI:
energy harvesting systems: A comprehensive review,” in 2020 IEEE 10.1038/s41598-017-18571-9.
International Conf on Natural and Engineering Sciences for Sahel's [82] M. Zhou, M. S. H. Al-Furjan, J. Zou, and W. Liu, “A review on heat
Sustainable Development - Impact of Big Data Application on and mechanical energy harvesting from human – Principles,
Society and Environment (IBASE-BF), Ouagadougou, Burkina Faso, prototypes and perspectives,” Renewable and Sustainable Energy
2020, pp. 1-10, DOI: 10.1109/IBASE-BF48578.2020.9069597. Reviews, vol. 82, part 3, pp. 3582-3609, 2018, DOI:
[64] N. A. Eltresy et al., “RF Energy Harvesting IoT System for Museum 10.1016/j.rser.2017.10.102.
Ambience Control with Deep Learning,” Sensors 19, no. 20: 4465, [83] R. Riemer and A. Shapiro, “Biomechanical energy harvesting from
2019. human motion: theory, state of the art, design guidelines, and future
[65] M. Caselli, M. Ronchi, and A. Boni, “Power Management Circuits directions,” J Neuro Engineering Rehabil, vol. 8, no. 22, 2011, DOI:
for Low-Power RF Energy Harvesters,” J. Low Power Electron. 10.1186/1743-0003-8-22.
Appl., vol. 10, no. 3: 29, 2020. [84] Y. Zhang et al., “A Batteryless 19 µW MICS/ISM-Band Energy
[66] C. Lin, C. Chiu, and J. Gong, “A Wearable Rectenna to Harvest Harvesting Body Sensor Node SoC for ExG Applications,” IEEE
Low-Power RF Energy for Wireless Healthcare Applications,” in Journal of Solid-State Circuits, vol. 48, no. 1, pp. 199-213, Jan. 2013,
2018 11th International Congress on Image and Signal Processing, DOI: 10.1109/JSSC.2012.2221217.
BioMedical Engineering and Informatics (CISP-BMEI), Beijing, [85] V. Leonov, “Thermoelectric Energy Harvesting of Human Body
China, 2018, pp. 1-5, DOI: 10.1109/CISP-BMEI.2018.8633222. Heat for Wearable Sensors,” IEEE Sensors Journal, vol. 13, no. 6,
[67] A. Dhungana and E. Bulut, “Opportunistic Wireless Crowd Charging pp. 2284-2291, Jun. 2013, DOI: 10.1109/JSEN.2013.2252526.
of IoT Devices from Smartphones,” in 2020 16th International [86] C. Watkins, B. Shen, and R. Venkatasubramanian, “Low-grade-heat
Conference on Distributed Computing in Sensor Systems (DCOSS), energy harvesting using superlattice thermoelectrics for applications
Marina del Rey, CA, USA, 2020, pp. 376-380, DOI: in implantable medical devices and sensors,” in ICT 2005. 24th
10.1109/DCOSS49796.2020.00066. International Conference on Thermoelectrics, 2005., Clemson, SC,
[68] A. Dhungana and E. Bulut, “Energy balancing in mobile USA, 2005, pp. 265-267, DOI: 10.1109/ICT.2005.1519934.
opportunistic networks with wireless charging: Single and multi-hop [87] M. K. Kim, M. S. Kim, S. E. Jo, H. L. Kim, S. M. Lee, and Y. J.
approaches,” Ad Hoc Networks, vol. 111, 102342, 2021. Kim, “Wearable thermoelectric generator for human clothing
[69] N. Elvin and A. Erturk, “Introduction and Methods of Mechanical applications,” in 2013 Transducers & Eurosensors XXVII: The 17th
Energy Harvesting” in Advances in Energy Harvesting Methods, N. International Conference on Solid-State Sensors, Actuators and
Elvin and A. Erturk, Eds., Springer, New York, NY, 2013, DOI: Microsystems (TRANSDUCERS & EUROSENSORS XXVII),
10.1007/978-1-4614-5705-3_1. Barcelona, Spain, 2013, pp. 1376-1379, DOI:
[70] A. Nechibvute, A. Chawanda, and P. Luhanga, “Piezoelectric Energy 10.1109/Transducers.2013.6627034.
Harvesting Devices: An Alternative Energy Source for Wireless [88] H. Liu, Y. Wang, D. Mei, Y. Shi Y, and Z. Chen, “Design of a
Sensors,” Smart Materials Research, vol. 2012, article ID 853481, wearable thermoelectric generator for harvesting human body
2012, DOI: 10.1155/2012/853481. energy,” in Wearable Sensors and Robots. Lecture Notes in
[71] H. Li, C. Tian, Z. D. Deng, “Energy harvesting from low frequency Electrical Engineering, C. Yang, G. Virk, and H. Yang, Eds., vol
applications using piezoelectric materials”, Applied Physics 399, Springer, Singapore, 2017, DOI: 10.1007/978-981-10-2404-
Reviews, vol. 1, no. 4, 2014, DOI: 10.1063/1.4900845. 7_5.
[89] Y. Chen, Y. Zhang, F. Yuan, F. Ding, and O. G. Schmidt, “A
Flexible PMN-PT Ribbon-Based Piezoelectric-PyroelectricHybrid
VOLUME XX, 2021 18
Generator for Human-Activity Energy Harvesting and Monitoring,” Journal, vol. 19, no. 9, pp. 3501-3512, 2019, DOI:
Adv. Electron. Mater, vol. 3, 1600540, 2017. 10.1109/JSEN.2019.2892604.
[90] J.‐H. Lee, et al., “Highly Stretchable Piezoelectric‐Pyroelectric [108] R. Du, L. Gkatzikis, C. Fischione, and M. Xiao, “Energy Efficient
Hybrid Nanogenerator,” Adv. Mater., vol. 26, issue 5, pp. 765-769, Sensor Activation for Water Distribution Networks Based on
2014, DOI: 10.1002/adma.201303570. Compressive Sensing,” IEEE Journal on Selected Areas in
[91] H. Xue, Q. Yang, D. Wang, W. Luo, W. Wang, M. Lin, D. Liang, Communications, vol. 33, no. 12, pp. 2997-3010, Dec. 2015, DOI:
and Q. Luo, “A wearable pyroelectric nanogenerator and self- 10.1109/JSAC.2015.2481199.
powered breathing sensor,” Nano Energy, vol. 38, pp. 147-154, [109] J. Lee et al., “A Self-Tuning IoT Processor Using Leakage-Ratio
2017, DOI: 10.1016/j.nanoen.2017.05.056. Measurement for Energy-Optimal Operation,” IEEE Journal of
[92] P. D. Mitcheson, E. M. Yeatman, G. K. Rao, A. S. Holmes, and T. C. Solid-State Circuits, vol. 55, no. 1, pp. 87-97, Jan. 2020, DOI:
Green, “Energy Harvesting From Human and Machine Motion for 10.1109/JSSC.2019.2939890.
Wireless Electronic Devices,” Proceedings of the IEEE, vol. 96, no. [110] C. Zhuo, S. Luo, H. Gan, J. Hu, and Z. Shi, “Noise-Aware DVFS for
9, pp. 1457-1486, Sept. 2008, DOI: 10.1109/JPROC.2008.927494. Efficient Transitions on Battery-Powered IoT Devices,” IEEE
[93] Y.-M. Choi, M. G. Lee, and Y. Jeon, “Wearable Biomechanical Transactions on Computer-Aided Design of Integrated Circuits and
Energy Harvesting Technologies.” Energies, vol. 10, 1483, 2017. Systems, vol. 39, no. 7, pp. 1498-1510, July 2020, DOI:
[94] M. Cai, Z. Yang, J. Cao, and W. Liao, “Recent Advances in Human 10.1109/TCAD.2019.2917844.
Motion Excited Energy Harvesting Systems for Wearables,” Energy [111] G. Berthou, K. Marquet, T. Risset, and G. Salagnac, “Accurate
Technol., vol. 8, issue 10, 2000533, 2020, 10.1002/ente.202000533. Power Consumption Evaluation for Peripherals in Ultra Low-Power
[95] M. Cai and W. -H. Liao, “High-Power Density Inertial Energy embedded systems,” in 2020 Global Internet of Things Summit
Harvester Without Additional Proof Mass for Wearables,” IEEE (GIoTS), Dublin, Ireland, 2020, pp. 1-6, DOI:
Internet of Things Journal, vol. 8, no. 1, pp. 297-308, Jan. 2021, 10.1109/GIOTS49054.2020.9119593.
DOI: 10.1109/JIOT.2020.3003262. [112] H. Choi, Y. Koo, and S. Park, “Modeling the Power Consumption of
[96] J. Zhong, Y. Zhang, Q. Zhong, Q. Hu, B. Hu, Z. L. Wang, and J. Function-Level Code Relocation for Low-Power Embedded
Zhou, “Fiber-Based Generator for Wearable Electronics and Mobile Systems,” Appl. Sci., vol. 9, no. 11:2354. 2019, DOI:
Medication,” ACS Nano, vol. 8, no. 6, pp. 6273-6280, 2014, DOI: 10.3390/app9112354.
10.1021/nn501732z. [113] P. P. Priyesh and S. K. Bharti, “Dynamic transmission power control
[97] C. Konstantopoulos, E. Koutroulis, N. Mitianoudis, and A. Bletsas, in wireless sensor networks using P-I-D feedback control technique,”
“Converting a Plant to a Battery and Wireless Sensor with Scatter in 2017 9th International Conference on Communication Systems
Radio and Ultra-Low Cost,” IEEE Transactions on Instrumentation and Networks (COMSNETS), Bangalore, 2017, pp. 306-313, DOI:
and Measurement, vol. 65, no. 2, pp. 388-398, 2016. 10.1109/COMSNETS.2017.7945391.
[98] F. Lin, Y. Kuo, J. Hsieh, H. Tsai, Y. Liao, and H. Lee, “A Self- [114] H. U. Yildiz, B. Tavli, and H. Yanikomeroglu, “Transmission Power
Powering Wireless Environment Monitoring System Using Soil Control for Link-Level Handshaking in Wireless Sensor Networks,”
Energy,” IEEE Sensors Journal, vol. 15, no. 7, pp. 3751 - 3758, IEEE Sensors Journal, vol. 16, no. 2, pp. 561-576, Jan.15, 2016,
2015. DOI: 10.1109/JSEN.2015.2486960.
[99] B. Dong, Q. Shi, Y. Yang, F. Wen, Z. Zhang, and C. Lee, [115] L. Tan and M. Wu, “Data Reduction in Wireless Sensor Networks: A
“Technology evolution from self-powered sensors to AIoT enabled,” Hierarchical LMS Prediction Approach,” IEEE Sensors Journal, vol.
Nano Energy, vol. 79, p. 105414, 2021. 16, no. 6, pp. 1708-1715, March 15, 2016, DOI:
[100] H. Liu, H. Fu, L. Sun, C. Lee, and E. M. Yeatman, “Hybrid energy 10.1109/JSEN.2015.2504106.
harvesting technology: From materials, structural design, system [116] T. Santejudean, S. Folea, and G. Mois, “Analysis of Low-Power
integration to applications,” Renewable and Sustainable Energy Operation for an Environmental Monitoring Beacon,” in 2020 IEEE
Reviews, vol. 137, p. 110473, 2021. International Conference on Automation, Quality and Testing,
[101] F. Deng, X. Yue, X. Fan, S. Guan, Y. Xu, and J. Chen, “Multisource Robotics (AQTR), Cluj-Napoca, Romania, 2020, pp. 1-5, DOI:
Energy Harvesting System for a Wireless Sensor Network Node in 10.1109/AQTR49680.2020.9129917.
the Field Environment,” IEEE Internet of Things Journal, vol. 6, no. [117] D. Goldsmith and J. Brusey, “The Spanish Inquisition Protocol—
1, pp. 918-927, 2019. Model based transmission reduction for wireless sensor networks,” in
[102] T. Yamaguchi and S. Hashimoto, “A green battery by pot‐plant SENSORS, 2010 IEEE, Waikoloa, HI, USA, 2010, pp. 2043-2048,
power,” IEEJ Trans Elec Electron Eng, vol. 7, pp. 441-442, 2012, DOI: 10.1109/ICSENS.2010.5690285.
DOI: 10.1002/tee.21754. [118] X. Shi and G. Stromberg, “SyncWUF: An Ultra Low-Power MAC
[103] Murata Manufacturing Co., Ltd., “Sizing Solar Energy Harvesters for Protocol for Wireless Sensor Networks,” in IEEE Transactions on
Wireless Sensor Networks – Application Note M1002”, Murata Mobile Computing, vol. 6, no. 1, pp. 115-125, Jan. 2007,
Manufacturing Co., Ltd., Nagaokakyo, Kyoto, Application Note 10.1109/TMC.2007.250675.
ANM1002, 14 Dec. 2016. [Online] Available: [119] A. Ikpehai et al., “Low-Power Wide Area Network Technologies for
https://wireless.murata.com/media/products/apnotes/anm1002.pdf. Internet-of-Things: A Comparative Review,” IEEE Internet of
[Accessed 06 Feb 2021] Things Journal, vol. 6, no. 2, pp. 2225-2240, April 2019, DOI:
[104] S. C. Folea and G. Mois, “A Low-Power Wireless Sensor for Online 10.1109/JIOT.2018.2883728.
Ambient Monitoring,” IEEE Sensors Journal, vol. 15, no. 2, pp. 742- [120] M. Stusek, K. Zeman, P. Masek, J. Sedova, and J. Hosek, “IoT
749, Feb. 2015, DOI: 10.1109/JSEN.2014.2351420. Protocols for Low-power Massive IoT: A Communication
[105] Z. Zhang, L. Shu, C. Zhu, and M. Mukherjee, “A Short Review on Perspective,” in 2019 11th International Congress on Ultra Modern
Sleep Scheduling Mechanism in Wireless Sensor Networks” in Telecommunications and Control Systems and Workshops (ICUMT),
Quality, Reliability, Security and Robustness in Heterogeneous Dublin, Ireland, 2019, pp. 1-7, DOI:
Systems, L. Wang, T. Qiu, and W. Zhao, Eds., QShine 2017. Lecture 10.1109/ICUMT48472.2019.8970868.
Notes of the Institute for Computer Sciences, Social Informatics and [121] M. S. Bahbahani and E. Alsusa, “A Cooperative Clustering Protocol
Telecommunications Engineering, vol 234, 2018, Springer, Cham, With Duty Cycling for Energy Harvesting Enabled Wireless Sensor
DOI: 10.1007/978-3-319-78078-8_7. Networks,” IEEE Transactions on Wireless Communications, vol.
[106] A. A. -H. Hassan, W. M. Shah, A. -H. H. Habeb, M. F. I. Othman, 17, no. 1, pp. 101-111, Jan. 2018, DOI:
and M. N. Al-Mhiqani, “An Improved Energy-Efficient Clustering 10.1109/TWC.2017.2762674.
Protocol to Prolong the Lifetime of the WSN-Based IoT,” IEEE [122] A. Kumar, M. Zhao, K. Wong, Y. L. Guan, and P. H. J. Chong, “A
Access, vol. 8, pp. 200500-200517, 2020, DOI: Comprehensive Study of IoT and WSN MAC Protocols: Research
10.1109/ACCESS.2020.3035624. Issues, Challenges and Opportunities,” IEEE Access, vol. 6, pp.
[107] C. S. Abella et al., “Autonomous Energy-Efficient Wireless Sensor 76228-76262, 2018, DOI: 10.1109/ACCESS.2018.2883391.
Network Platform for Home/Office Automation,” IEEE Sensors [123] A. Jarwan, A. Sabbah, and M. Ibnkahla, “Data Transmission
Reduction Schemes in WSNs for Efficient IoT Systems,” IEEE
VOLUME XX, 2021 19
Journal on Selected Areas in Communications, vol. 37, no. 6, pp. [141] Semtech Corporation, “LoRa and LoRaWAN: A Technical
1307-1324, June 2019, DOI: 10.1109/JSAC.2019.2904357. Overview,” Semtech Corporation, Camarillo, CA, USA, December
[124] Y. Fathy and P. Barnaghi, “Quality-Based and Energy-Efficient Data 2019. Available: https://lora-developers.semtech.com/library/tech-
Communication for the Internet of Things Networks,” IEEE Internet papers-and-guides/lora-and-lorawan/. [Accessed 20 December 2020]
of Things Journal, vol. 6, no. 6, pp. 10318-10331, Dec. 2019, DOI: [142] S. Kim and P.H. Chou, “Energy Harvesting with Supercapacitor-
10.1109/JIOT.2019.2938101. Based Energy Storage” in Smart Sensors and Systems, YL. Lin, CM.
[125] P. Spachos and A. Mackey, “Energy efficiency and accuracy of solar Kyung, H. Yasuura, and Y. Liu, Eds., 2015, Springer, Cham, DOI:
powered BLE beacons,” Computer Communications, vol. 119, pp. 10.1007/978-3-319-14711-6_10.
94-100, 2018. [143] B. Schweber, “Energy-Harvesting Storage Options: Rechargeable
[126] S. C. Folea and G. D. Mois, “Lessons Learned From the Battery, Supercapacitor, or Both?,” Digi-Key Electronics, Thief
Development of Wireless Environmental Sensors,” IEEE River Falls, Minnesota, USA, 20 Feb. 2013. [Online] Available:
Transactions on Instrumentation and Measurement, vol. 69, no. 6, pp. https://www.digikey.ro/ro/articles/energyharvesting-storage-options-
3470-3480, June 2020, DOI: 10.1109/TIM.2019.2938137. rechargeable-battery-supercapacitor-or-both. [Accessed 18 Feb.
[127] G. Mois, S. Folea, and T. Sanislav, “Analysis of Three IoT-Based 2021].
Wireless Sensors for Environmental Monitoring,” IEEE Transactions [144] A. S. Weddell, G. V. Merrett, T. J. Kazmierski, and B. M. Al-
on Instrumentation and Measurement, vol. 66, issue 8, pp. 2056 - Hashimi, “Accurate Supercapacitor Modeling for Energy Harvesting
2064, 2017. Wireless Sensor Nodes,” IEEE Transactions on Circuits and Systems
[128] M. Tacca, P. Monti, and A. Fumagalli, “Cooperative and Reliable II: Express Briefs, vol. 58, no. 12, pp. 911-915, Dec. 2011, DOI:
ARQ Protocols for Energy Harvesting Wireless Sensor Nodes,” 10.1109/TCSII.2011.2172712.
IEEE Transactions on Wireless Communications, vol. 6, no. 7, pp. [145] R. G. Cid-Fuentes, A. Cabellos-Aparicio, and E. Alarcón, “Energy
2519-2529, July 2007, DOI: 10.1109/TWC.2007.05878. Buffer Dimensioning Through Energy-Erlangs in Spatio-Temporal-
[129] S. Reddy and C. R. Murthy, “Profile-Based Load Scheduling in Correlated Energy-Harvesting-Enabled Wireless Sensor Networks,”
Wireless Energy Harvesting Sensors for Data Rate Maximization,” in IEEE Journal on Emerging and Selected Topics in Circuits and
2010 IEEE International Conference on Communications, Cape Systems, vol. 4, no. 3, pp. 301-312, Sept. 2014, DOI:
Town, 2010, pp. 1-5, DOI: 10.1109/ICC.2010.5502464. 10.1109/JETCAS.2014.2337194.
[130] S. Reddy and C. R. Murthy, “Dual-Stage Power Management [146] S. Zeadally, F. Siddiqui, and Z. Baig, “25 years of Bluetooth
Algorithms for Energy Harvesting Sensors,” IEEE Transactions on Technology,” Future Internet, vol. 11, no. 9, 2019.
Wireless Communications, vol. 11, no. 4, pp. 1434-1445, April 2012, [147] K. E. Jeon, J. She, P. Soonsawad, and P. C. Ng, “BLE Beacons for
DOI: 10.1109/TWC.2012.032812.110623. Internet of Things Applications: Survey, Challenges, and
[131] S. Escolar, A. Caruso, S. Chessa, X. d. Toro, F. J. Villanueva, and J. Opportunities,” IEEE Internet of Things Journal, vol. 5, no. 2, pp.
C. López, “Statistical Energy Neutrality in IoT Hybrid Energy- 811-828, April 2018, DOI: 10.1109/JIOT.2017.2788449.
Harvesting Networks,” in 2018 IEEE Symposium on Computers and [148] Q. Liu, W. IJntema, A. Drif, P. Pawełczak, M. Zuniga, and K. S.
Communications (ISCC), Natal, 2018, pp. 00444-00449, DOI: Yıldırım, “Perpetual Bluetooth Communications for the IoT,” IEEE
10.1109/ISCC.2018.8538532. Sensors Journal, vol. 21, no. 1, pp. 829-837, 1 Jan.1, 2021, DOI:
[132] J. Yang, X. Wu, and J. Wu, “Optimal Online Sensing Scheduling for 10.1109/JSEN.2020.3012814.
Energy Harvesting Sensors With Infinite and Finite Batteries,” IEEE [149] H. Yao, H. Shu, X. Liang, H. Yan, and H. Sun, “Integrity Monitoring
Journal on Selected Areas in Communications, vol. 34, no. 5, pp. for Bluetooth Low Energy Beacons RSSI Based Indoor Positioning,”
1578-1589, May 2016, DOI: 10.1109/JSAC.2016.2551561. IEEE Access, vol. 8, pp. 215173-215191, 2020, DOI:
[133] O. H. Abdelrahman, “A MARKOV-MODULATED DIFFUSION 10.1109/ACCESS.2020.3038894.
MODEL FOR ENERGY HARVESTING SENSOR NODES,” [150] P. Spachos and K. N. Plataniotis, “BLE Beacons for Indoor
Probability in the Engineering and Informational Sciences, vol. 31, Positioning at an Interactive IoT-Based Smart Museum,” IEEE
no. 4, pp. 505–515, 2017, DOI:10.1017/s0269964817000158. Systems Journal, vol. 14, no. 3, pp. 3483-3493, Sept. 2020, DOI:
[134] M. Ku, Y. Chen, and K. J. R. Liu, “Data-Driven Stochastic Models 10.1109/JSYST.2020.2969088.
and Policies for Energy Harvesting Sensor Communications,” IEEE [151] P. Octaviani and W. Ce, “Inventory Placement Mapping using
Journal on Selected Areas in Communications, vol. 33, no. 8, pp. Bluetooth Low Energy Beacon Technology for Warehouses,” in
1505-1520, Aug. 2015, DOI: 10.1109/JSAC.2015.2391651. 2020 International Conference on Information Management and
[135] N. Dang, R. Valentini, E. Bozorgzadeh, M. Levorato, and N. Technology (ICIMTech), Bandung, Indonesia, 2020, pp. 354-359,
Venkatasubramanian, “A unified stochastic model for energy DOI: 10.1109/ICIMTech50083.2020.9211206.
management in solar-powered embedded systems,” in 2015 [152] ams AG. CCS811 Gas Sensor Solution Ultra-Low Power Digital Gas
IEEE/ACM International Conference on Computer-Aided Design Sensor for Monitoring Indoor Air Quality; ams AG: Premstaetten,
(ICCAD), Austin, TX, 2015, pp. 621-626, DOI: Austria, 2018.
10.1109/ICCAD.2015.7372627. [153] Cypress Semicond. Corporation, CYBLE-022001-00 EZBLE
[136] C. Delgado, J. M. Sanz, C. Blondia, and J. Famaey, “Batteryless PRoCModule—Cypress, Cypress Semicond. Corporation, San Jose,
LoRaWAN Communications Using Energy Harvesting: Modeling CA, USA, Feb. 2016.
and Characterization,” IEEE Internet of Things Journal, vol. 8, no. 4, [154] SENSIRION AG, Datasheet SHT21—Humidity and Temperature
pp. 2694-2711, 15 Feb.15, 2021, DOI: 10.1109/JIOT.2020.3019140. Sensor IC, Version, 4th ed., SENSIRION, Staefa, Switzerland, May
[137] I. T. Castro, L. Landesa, and A. Serna, “Modeling the Energy 2014. [Online]. Available: https://www.sensirion.com
Harvested by an RF Energy Harvesting System Using Gamma [155] NXP Semiconductors, Miniature I2C Digital Barometer, Review,
Processes,” Mathematical Problems in Engineering, vol. 2019, article NXP Semicond., Eindhoven, The Netherlands, Feb. 2013.
ID 8763580, 2019, DOI: 10.1155/2019/8763580. [156] OPT3001 Ambient Light Sensor (ALS), Texas Instrum., Dallas, TX,
[138] D. Altinel and G. K. Kurt, “Modeling of Multiple Energy Sources for USA, Jul. 2014.
Hybrid Energy Harvesting IoT Systems,” IEEE Internet of Things [157] BQ25504 Ultra Low-Power Boost Converter With Battery
Journal, vol. 6, no. 6, pp. 10846-10854, Dec. 2019, DOI: Management for Energy HArvester Applications, Texas Instrum.,
10.1109/JIOT.2019.2942071. Dallas, TX, USA, Oct. 2011.
[139] M. Chu, X. Liao, H. Li, and S. Cui, “Power Control in Energy [158] IXYS UK Westcode Ltd, Product Brief—IXYS Solar Products, 3rd
Harvesting Multiple Access System With Reinforcement Learning,” ed., IXYS UK Westcode, Chippenham, U.K., Nov. 2013. [Online].
IEEE Internet of Things Journal, vol. 6, no. 5, pp. 9175-9186, Oct. Available: http://ixapps.ixys.com/DataSheet/ixys_solar.pdf
2019, DOI: 10.1109/JIOT.2019.2928837. [159] MULTICOMP, Lithium-ion Rechargeable Battery (Coin Cell,
[140] D. Fan, L. Lopez Ruiz, J. Gong, and J. Lach, “EHDC: An Energy Lithium, 3V, 120mAh - LIR2450), MULTICOMP, 13 Oct 2020.
Harvesting Modeling and Profiling Platform for Body Sensor [Online]. Available: https://ro.farnell.com/b/multicomp. [Accessed
Networks,” IEEE Journal of Biomedical and Health Informatics, vol. 06 Feb 2021].
22, no. 1, pp. 33-39, Jan. 2018, DOI: 10.1109/JBHI.2017.2733549.
VOLUME XX, 2021 20
[160] Texas Instruments, TPS783xx 500-nA IQ, 150-mA, Ultralow

Quiescent Current Low-Dropout Linear Regulator, Texas SHERALI ZEADALLY earned his bachelor’s
Instruments, Dallas, USA, Nov 2014. [Online]. Available: degree in computer science from the University
http://www.ti.com/lit/ds/symlink/tps783.pdf. [Accessed 06 Feb of Cambridge, England. He also received a
2021]. doctoral degree in computer science from the
[161] Powercast Corporation. P1110-EVB – Evaluation Board for P1110 University of Buckingham, England.
Powerharvester® Receiver. Technical Manual. Powercast He is currently an Associate Professor in the
Corporation. Revision c. College of Communication and Information,
[162] “BU-808: How to Prolong Lithium-based Batteries,” 17 September University of Kentucky. His research interests
2020. [Online]. Available: include Cybersecurity, privacy, Internet of
https://batteryuniversity.com/learn/article/how_to_prolong_lithium_b Things, computer networks, and energy-
ased_batteries, 2020-09-17 [Accessed 20 December 2020]. efficient networking. He is a Fellow of the
[163] J. J. Estrada-López, A. Abuellil, Z. Zeng, and E. Sánchez-Sinencio, British Computer Society and the Institution of
“Multiple Input Energy Harvesting Systems for Autonomous IoT Engineering Technology, England.
End-Nodes,” J. Low Power Electron. Appl. 2018, 8, 6, 2018, DOI:
10.3390/jlpea8010006.
SILVIU CORNELIU FOLEA (M’07 –
TEODORA SANISLAV (M’12) received the SM’18) received the Ph.D. degree in control
B.S. degree in computer science and the Ph.D. systems and the Habilitation degree from the
degree in systems engineering from the Technical University of Cluj-Napoca, Cluj-
Technical University of Cluj-Napoca, Napoca, Romania, in 2005 and 2017,
Romania, in 2003 and 2013, respectively. respectively.
From 2004 to 2011, she was a Scientific He is currently a Professor with the
Researcher in a research institute in the field of Department of Automation, Technical
automation. She is currently a Lecturer with the University of Cluj-Napoca, Cluj-Napoca,
Automation Department, Technical University Romania. His current research interests include
of Cluj-Napoca. Her current research interests development of wireless environmental sensors, energy harvesting,
include cyber-physical systems, dependability embedded and reconfigurable systems, data acquisition systems and
analysis, Internet of Things and intelligent graphical programming.
systems.
GEORGE DAN MOIS (M’08) received the

Ph.D. degree in systems engineering from the
Technical University of Cluj-Napoca, Cluj-
Napoca, Romania, in 2011.
He is currently a Lecturer with the
Automation Department, Technical
University of Cluj-Napoca. Since 2017, he
has been a part of the Innovation and
Universities Group, Bosch Engineering
Center Cluj. His current research interests
include embedded system design and wireless
sensor networks.
VOLUME XX, 2021 21

Energy Harvesting Techniques For Internet of Thing

Uploaded by

Copyright:

Available Formats

Energy Harvesting Techniques For Internet of Thing

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Energy Harvesting Techniques For Internet of Thing

Uploaded by

Copyright:

Available Formats

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

Energy Harvesting Techniques for Internet of

Corresponding author: Teodora Sanislav (e-mail: teodora.sanislav@aut.utcluj.ro).

I. INTRODUCTION • IoT connectivity (protocols, gateways) is responsible for

VOLUME XX, 2021 1

VOLUME XX, 2021 2

VOLUME XX, 2021 3

VOLUME XX, 2021 4

FIGURE 5. Main components of a RF energy harvesting system. Adapted from [61].

VOLUME XX, 2021 5

VOLUME XX, 2021 6

VOLUME XX, 2021 7

TABLE 1. Analysis of energy harvesting techniques.

In this context, various energy harvesting models and

VOLUME XX, 2021 8

[104], sleep scheduling [105], the reduction of the required

VOLUME XX, 2021 9

Architecture of IoT device with

VOLUME XX, 2021 10

As Table 2 shows, energy harvesting systems adopt the

VOLUME XX, 2021 11

VOLUME XX, 2021 12

OPTIMIZATIONS FOR ACHIEVING ENERGY AUTONOMY

VOLUME XX, 2021 13

VOLUME XX, 2021 14

VOLUME XX, 2021 15

VOLUME XX, 2021 16

VOLUME XX, 2021 17

VOLUME XX, 2021 18

VOLUME XX, 2021 19

VOLUME XX, 2021 20

[160] Texas Instruments, TPS783xx 500-nA IQ, 150-mA, Ultralow

GEORGE DAN MOIS (M’08) received the

VOLUME XX, 2021 21

You might also like