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.2020.2986329, IEEE Access

Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000.
Digital Object Identifier 10.1109/ACCESS.2017.Doi Number

Wearables and the Internet of Things (IoT),

Applications, Opportunities, and Challenges: A
Survey
F. John Dian1, Senior Member IEEE, R. Vahidnia2, Member IEEE and A. Rahmati3
1
Department of Electrical and Computer Engineering, British Columbia Institute of Technology, Burnaby, BC V5G 3H2, Canada
2
Department of Electrical and Computer Engineering, British Columbia Institute of Technology, Burnaby, BC V5G 3H2, Canada
3
Department of Electrical and Computer Engineering, University of Waterloo, ON N2L 3G1, Canada
Corresponding author: R. Vahidnia (e-mail: [email protected])
This work was supported by the British Columbia Institute of Technology research grant.

ABSTRACT Smart wearables collect and analyze data, and in some scenarios make a smart decision and
provide a response to the user and are finding more and more applications in our daily life. In this paper, we
comprehensively survey the most recent and important research works conducted in the area of wearable
Internet of Things (IoT) and classify the wearables into four major clusters: (i) health, (ii) sports and daily
activity, (iii) tracking and localization, and (iv) safety. The fundamental differences of the algorithms
associated within each cluster are grouped and analyzed and the research challenges and open issues in each
cluster are discussed. This survey reveals that although Cellular IoT (CIoT) has many advantages and can
bring enormous applications to IoT wearables, it has been rarely studied by the researchers. This article also
addresses the opportunities and challenges related to implementing CIoT-enabled wearables.

INDEX TERMS Smart wearables, Internet of Things, cellular IoT.

I. INTRODUCTION advantages that the smart wearables can provide have

IoT-enabled wearables are smart devices that can be worn enabled the development of wearable technology in a fast
as external accessories, embedded in clothing and garments, pace. We have witnessed a rapid development of smart
implanted in the body, or even adhered to or tattooed on the wearable products adapted for various applications during
skin. These devices are able to connect to the Internet in the past few years. Smart watches, wrist bands, eye wears,
order to collect, send data and receive the information that headsets, ear-buds, body straps, foot and hand worn
can be used for smart decision making. These wearables are devices, and smart jewelries are some of the wearables that
becoming an increasingly important part of IoT technology have been developed for different applications (Figure 1).
and their development is moving from being simple
accessories to more specialized and practical applications.
Smart wearables can interact with an array of other devices,
such as smartphones, for the purpose of computing and
communication.
Due to mobility of the human and animals, smart wearable
devices are becoming increasingly important since they can
collect and send the data on the move and accordingly
receive information from the Internet which helps in
making smarter decisions. The use of smart wearables can
bring efficiency and optimization to the applications,
enhance the quality of life, and increase productivity or
safety.
The advancements in low power mobile networks, decrease
in the size of electronic devices and sensors, as well as the Figure 1. Different wearables developed for various applications

VOLUME XX, 2017 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
10.1109/ACCESS.2020.2986329, IEEE Access

Fitness activity trackers were the first big wave of wearable for online processing and consequently the IoT device will
devices in the market followed by Bluetooth headsets, receive some information that helps the IoT device to
smartwatches, and web-enabled glasses. The gaming operate. The prevalence of wearable IoT can be fully
industry added more wearables, with virtual reality and recognized when we have integrated IoT platforms and
augmented reality headsets. However, the important life- when many issues regarding data ownership, policies on
altering applications in wearable technology is found sharing data, privacy and safety issues are resolved.
mostly in health monitoring and medical use cases. The majority of the research works in the area of wearable
There exist several other works that have looked into IoT use unlicensed short range communication technologies
different aspects of wearable technologies. For example, [1] such as WiFi and Bluetooth mainly to monitor the user’s
provides a survey of available products in the market as health, activity, location and ensure his or her safety.
well as the research prototypes. The authors in [2] surveyed However, the application of such solutions is limited to the
all the existing electronic components used in the design of use of a gateway or mobile device to connect to the
garments used in wearable devices. The article in [3] Internet. By the emergence of cellular IoT technologies
presents a survey of the sensors used in wearable devices. (e.g., LTE-M and NB-IoT) introduced in 3rd Generation
However, to the best of our knowledge, this paper is the Partnership Project (3GPP) Release 13, new methods and
first survey about the IoT-enabled wearables applications, applications are expected to be proposed for wearable IoT
opportunities and challenges where we focus on various devices. In this paper, we reviewed the opportunities and
clusters of smart wearables and classify the existing challenges associated with cellular IoT-based wearables.
applications of these devices by addressing the fundamental The remainder of this paper is organized as follows. In
differences of the algorithms associated within a specific Section II, we survey and classify the applications of smart
application in each cluster. We also study the applications, wearables. Section III discusses the use of cellular IoT
opportunities and challenges of cellular IoT in more details (CIoT) in smart wearables. In Section IV, the open research
as the authors believe that cellular IoT can be a game challenges associated with wearable IoT is discussed.
changer in the area of IoT-based wearable technology. Future directions and conclusions are presented in Section
Since wearable IoT devices are battery operated, power V.
consumption is considered as an important design factor. II. WEARABLE IOT CLASSIFICATION
However, this does not mean that a wearable device cannot To classify the applications of wearable IoT, and complete
perform complex processing or it is limited in the amount this survey, first, the state-of-the-art research works, papers
of data that transmits due to extensive amount of power that and articles in this area were collected. This work is done by
in these situations will be consumed. In general, if the searching databases such as IEEE Xplore and Association for
wearable IoT device performs a large amount of processing Computing Machinery (ACM) digital libraries. Then the
or transmits large amount of data, then its battery needs to related works are categorized in clusters according to their
be charged more frequently. This is quite possible in many application and each research paper is placed in one of the
applications and use cases. For instance, in a situation that clusters as shown in Figure 2. It is seen that for some of the
the wearable IoT device is used to detect fall, the device clusters, there are a broader range of applications mostly due
might be worn during the time that the wearer is not asleep to their importance. Besides the main clusters, there are other
and can be charged during his/her sleep time. By the same wearable IoT devices that are used in other applications such
token, the wearable IoT devices that are used during a sport as virtual games in order to enhance gaming experience,
activity, can drain the battery during the length of the payment applications, and education. Investigating these
activity which is usually a few hours and can easily get other use cases are out of the scope of this survey.
charged afterwards. It is clear that if a wearable IoT device
is used to find the trajectory of a bird within a year, we not
only need to limit the amount of processing and data
transmission, but also try to find energy harvesting methods
to increase the duration of battery life cycle.
Wearable IoT technology can bring endless opportunities
for many applications. However, the true power of
wearable IoT will be realized when there will be an
integrated IoT system available. For this reason, most of the
existing research papers connect to the Internet in one of
the following two forms. First, the wearable IoT sends its
data to the cloud or a server on the Internet to be saved for
offline processing in future. The second form is to offload
Figure 2. Most researched wearable IoT application clusters
some of the computing from the wearable device. In this
form, the wearable IoT device sends its data to the Internet

VOLUME XX, 2020 9

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
10.1109/ACCESS.2020.2986329, IEEE Access

In the following sections, each category is presented by 4) Biochemical sensors: transdermal glucose.
listing the most significant published work. Analyzing the aforementioned signals gives extensive
insight about the health status of the patient/user.
A. Health The figure below shows the taxonomy of the use cases for
The health wearable IoT device is mainly used for remote non-invasive wearable IoT sensors excluding implantable
patient monitoring, treatment and in some cases for sensors [9].
rehabilitation purposes. The sensors collect the health-
related data and the device may perform limited
computation prior to transmitting the user/patient’s health
information to the the Internet for further analysis. The
device may also receive data to enable the user make
further decisions. In many applications, the wearable
devices are connected to smartphones to analyze the
collected data and then transmit it to a cloud computing-
based framework such as Microsoft Azure or Amazon Web
Figure 3. Taxonomy of wearable health care IoT sensors
Services (AWS) in order to store, process, and analyze the
data. Mobile health applications can be used to visualize the
analyzed data and provide insight about the user/patient’s Pulse rate can be read using wearables on the chest, wrist,
health. Moreover, in treatment applications, the analyzed earlobe, fingertip, and more via PPG, pressure, and radio
data can be used to send special commands to the wearable frequency (RF). In [10], a monitoring wearable device is
such as heating up the body or applying a shock. designed to detect the unambiguous changes in the
heartrate. The wearable alerts the user by a vibration to take
1) HEALTH TREATMENT & REHABILITATION the necessary medicine. The researchers in [11] proposed a
WEARABLE SYSTEMS:
multisensory wearable system to collect the necessary data
The IoT rehabilitation devices help disabled patients to
in order to warn the user of an impending cardiac arrest in
maintain and improve their physical or mental abilities. In
the early stages. A low power communication module use a
[4], a walker-based system for physiotherapy purposes is
smartphone to collect the body temperature and heart rate
proposed which continuously monitors and evaluates the
and then signal processing techniques and machine learning
movement metrics of the walker and sends them to the
algorithms are used to accurately diagnose and warn sudden
cloud where data is analyzed, and the results are presented
cardiac arrests. The work done in [12], presents a system
on a mobile application. An IoT-based armband for stroke
dedicated to monitoring the heart activity parameters using
rehabilitation is introduced by [5]. The wearable system
ECG mobile devices and sensors inserted in the fabric of
measures the bio-potential signals and the analyzed data is
the clothes. The assessed parameters are heart rate and
transmitted wirelessly to a robot hand. Then, the received
respiration.
signals are interpreted using a machine learning algorithm
The wearables equipped with thermistor nasal sensor can
to give users insight and feedback about their muscle
measure the respiratory rate. The sensor counts the number
movements. The robot hand assists the user in adjusting
of breaths taken by sensing the rise and fall of the
his/her posture and walking pattern in real time.
temperature if the air exhaled [13]. In [14], a smart IoT
In [6], the researchers have designed a smart wheelchair
solution is proposed for patients with asthma to measure
which allows the disabled person to easily interact with the
their respiration rate. The measured data is securely sent to
wheelchair via a mobile application that analyzes the data
the cloud where the patient’s information is analyzed. This
received from the sensors, and then visualizes the results
system uses watermarking and signal enhancement
for caregivers to remotely monitor the patient. The
techniques to secure the data transmission. In [15], using
publication in [7], introduces a rehabilitation system which
adhesive hydrogel, electrodes are placed on human chest to
is designed and implemented to help stoke patients
detect the pulsatile vibration created as a result of
suffering from upper limb disability. This rehabilitation
respiration. The proposed system is capable to be worn or
system exploits IoT through integrating gaming and
attached to the user’s shirt or chest belt. Then, collected
wearable technology.
data is digitized and wirelessly sent using an impulse radio
2) HEALTH MONITORING WEARABLE SYSTEMS: ultra-wideband transmitter which operates in 3.1 to 5 GHz
Depending on the type of sensors, the health monitoring frequency range. Due to its small size, light weight, low
wearables are classified into four major categories [8]: cost and low power signal processing circuitry the proposed
1) Bio-potential sensors: electroencephalography (EEG), respiration monitoring system has gained attention by other
electrocardiography (ECG), electromyography (EMG), and researchers. In [16], the researchers have designed a
photoplesmography (PPG) and etc. portable wearable IoT device to monitor the respiratory rate
2) Motion sensors: accelerometer and gyroscope. where the sensor is implemented in a smart vest to
3) Environmental sensors: ultrasound, pressure, temperature constantly monitor Chronic Obstructive Pulmonary Disease
and etc. (COPD) patients at home during the rest period between

VOLUME XX, 2020 9

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
10.1109/ACCESS.2020.2986329, IEEE Access

respiratory rehabilitation exercises. This total solution A summary of the aforementioned articles about healthcare
system also provides an e-health platform based on the IoT sensors in given in Table 1.
Internet of Medical Things (IoMT) paradigm.
The body temperature is typically measured by thermistor- TABLE 1. SUMMARY OF HEALTHCARE IOT SENSORS IN THE LITERATURE
Sensed Mobile Node
type sensors to detect the conditions such as hypothermia, parameter
Sensor Connectivity
app process
Wearable Ref

heat stroke and fevers. In [17], a wearable, IoT cloud-based RFID,

Orientation, IMU, load Wi-Fi &
Yes No Walker [4]
system is presented for real-time personal health force, distance cell, Bluetooth
ultrasound
monitoring. The collected data from the sensors can be

Rehabilitation
surface
viewed both on a cloud dashboard as well as an embedded electromyogra
phy
sEMG BLE No Yes Armband [5]

display on the wearable. In [18], a tablet-shaped ingestible Face image,

camera Wi-Fi Yes No Face [6]
eye blinks
sensor is developed to measure the core-body temperature Accelerome
Deflections,
based on gastric acid power generation. A custom acceleration,
ter,
Wi-Fi No No Leg, hand [7]
gyroscope,
integrated circuit (IC) is prototyped which could wirelessly orientation
flex
SFH 7051 Wi-Fi No No Wristband [10]
receive the temperature data transferred by a tablet-shape ECG and
Bluetooth Yes No Wristband [11]
device of diameter 10mm and height 8mm. temperature
Heartrate ECG &
Blood pressure (BP) is not considered a vital health sign inductive Leg, hand,
2G GPRS No Yes [12]
and is frequently measured alongside the other three vital sensor
WHMIS
chest

signs (i.e., pulse, respiratory and temperature). While most Passive

breathing
BP wearable devices are non-invasive, they are still airflow
Back-
No No headband [13]
scattering
obstructive. In many cases a chest wearable ECG is temperature
Respiratory change
connected to the other sensors with some wires. As far as Vibration impulse-radio
(piezoelectr ultra-wideband No Yes Chest [15]
we are concerned there is no comfortable wearable device
Monitoring

ic) transmitter
developed to continuously measure the BP with high Capacitive Bluetooth Yes Yes Smart vest [16]
LM35 Wi-Fi No No Finger [17]
accuracy. In [19], a survey on commercial wearable IoT IC mounted
devices to monitor the BP from a metrological point of Temperature on tablet- magnetic-field
No No Ingestible [18]
shaped coupling
view is presented. In this survey the lack of the traceability ingestible
piezoelectri
and reliability of the BP measurements is addressed. Blood Pressure
c
Wi-Fi Yes Yes Cuff [19]
Pulse oximeter sensors measure the blood oxygen using Blood oxygen
pulse-
GSM/GPRS No Yes Bracelet [20]
oximetry
PPG signals. The most widely used form of such sensors Near
are the wrist-wearable pulse oximetry wearables. In [20], a Blood glucose Infrared
radiation
Wi-Fi Yes No Finger [21]

non-invasive wearable cardiac monitoring and alert system Audio,

Mental well- acceleromet
is proposed which can continuously measure the cardiac being er and
Bluetooth Yes Yes Wristband [22]

values and using pulse oximetry. This solution detects the gyroscope

saturation level of oxygen as well as blood volume

variations in the tissues. The heart rate is obtained through B. Activity Recognition and sports
filtering and processing the sensed signals. The mobile This category is related to applications where the wearables
communication and GPS system enable emergency alerts are worn during sport activities to record different metrics
when the measured cardiac values are out of the threshold of the user/athlete activity in order to improve his/her
values. Moreover, the system gives this opportunity to the performance Also, applications of this cluster consider
user to self-trigger the alert system through Google voice gathering data regarding the recognition of daily activities
assistant. of humans and animals. Although activity recognition can
Some other wearable sensors with applications in have some applications in medical diagnostics and out-of-
healthcare could measure the blood glucose. For example, hospital health-care, the applications belonging in this
in [21], an ingestible smart drug is designed which cluster cover use cases beyond health category.
circulates in the body and senses the glucose level by In this section, we discuss the use of wearable IoT for two
passing IR radiation. The measured blood glucose level is related activities. The first set of activities belong to
then transmitted to the smartphone wirelessly. Another recognition of daily physical activities [23] and the second
application of IoT wearable devices is in mental wellbeing one is related to activities that are related to specific sports
monitoring where physiological signs are collected through [24]. The recognition of daily physical activities is usually
behavioral traits. In [22], an IoT-based wearable social dedicated to tracking routines and body movement related
sensing platform is developed through integrating behavior to skeletal muscles such as walking forward, walking
monitoring, privacy audio feature, and environment sensing backward, jogging, sleeping, running, going up or down the
in a naturalistic environment. In particular, privacy stairs, bending waist, frontal elevation of arms, bending
protected audio-wellbeing features are embedded into this knees, and jumping front or back. It also includes
platform to automatically evaluate speech information recognition of static postures such as sitting, sitting and
without the need to preserve raw audio data. relaxing, standing, laying down or the recognition of going
from one posture to another such as going from standing
posture to sitting one.

VOLUME XX, 2020 9

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
10.1109/ACCESS.2020.2986329, IEEE Access

Generally speaking, IoT can bring an abundance of For sport activities, measuring and estimating the
opportunities for sport players, organizations and fans to performance or efficiency of actions related to specific
increase their efficiency and offerings by creating an sport and providing feedback on parameters such as timing,
environment in which athletes can receive better training or angle or amount of applying or releasing a force can help
have access to data that helps to keep them healthier, improve the accuracy and performance. Using these
coaches are able to analyze injuries or find metrics on techniques, players can get real-time feedback of their
player performance, and organizations can offer fan performance and improve their performance to make it
engagement strategies or allow fans to receive personalized more consistent. There are many papers that discuss the use
offering from their favorite team. Wearable IoT is a game of wearable devices for improving the quality of activities
changer in sport activities. However, the use of wearable of a specific sport or even making the wearer do something
IoT is currently limited due to some leagues regulations and that would not be possible without wearing the wearable
some challenges that need to be resolved. Yet, as more and device. The wearer usually receives a message to help
more organizations recognize the advantages that IoT him/her to understand the quality of performance or provide
wearables can provide to players, coaches and fans, we will help on how to proceed. These messages might be seen on
see that wearable IoT devices become more and more the screen of the wearable or on displays in some other
prevalent in sports. Even though, wearable IoT can provide places. The wearer also may receive the messages or an
endless opportunities to sports as explained above , in this indication of a message using audio or haptics. The
section, we discuss those papers that are related to performance of activities related to many different sports
recognition of activities for specific sports such as skiing, have been researched such as badminton [42, 43],
climbing, martial arts, tennis, swimming, badminton, basketball [44, 45], rowing [46, 47] , swimming [48, 49],
weightlifting, or baseball. It should be mentioned that the hockey [50], skiing [51, 52], martial art [53] [35],
other opportunities mentioned above have not received weightlifting [54], tennis [55], baseball [56] and golf [57,
much attention. 58].
The human physical activity detection has found lots of
attention [25, 26, 27, 28, 29, 30, 23] .To detect human C. Tracking and localization
physical activity one or more sensors are installed on one or This category, is used mainly for tracking human and
several wearable devices attached to the body. These animal, to determine their location online. Finding the
sensors generate signals that can be analyzed to detect the position of a person or animal who is wearing a wearable
type of activity and find the information that differentiates device is important in many applications. Studying the
various activities. The existing methods usually achieve this trajectory trip of a bird, finding the location of a senior
goal in four stages. The first stage is pre-processing, noise citizen in a care-home facility, analyzing the movement of
cancelation and signal range adjustment which prepares the the people who are visiting an exhibition, or pet tracking
signal for the next stages. The second stage is feature are some examples of these applications. A comprehensive
extraction in which the features or parameters specific to study on localization using IoT technology is published in
the signal are extracted. The extracted features can usually [59], but the study does not concentrate on wearable IoT. In
be categorized as structural features or statistical ones. The general, the localization methods discussed in literature can
structural features find the correlation among different be divided into two main categories. The ones that use
signals. The statistical features can be found through time- offline training and the ones that without using offline
or transformed-domain analysis of a signal. Example of training can determine the location. The training dependent
statistical features are mean value, max value, peak, and methods can be classified into three clusters: 1)
size of signal. The transformation typically used for finding Fingerprinting 2) Stochastic-oriented models and 3)
the statistical features are Fourier Transform (FT) [31] [32], Machine learning based schemes. Fingerprinting has been
Discrete Cosine Transform (DCT) [33], or Wavelet studied extensively using various types of signals including
Transform (WT). The third stage is feature reduction which general signal pattern, audio signal, video signal and
tries to find most important features that can be used in motion [60, 61]. General signal pattern usually depends on
final stage by removing the redundant and irrelevant the Received Signal Strength Indicator (RSSI) value of the
information. Classification is the last stage in which the wireless signal in which the location fingerprint is found
reduced set of features are clustered to find patterns among through site survey and is recorded in a fingerprint data
various activities using a classification algorithm. base. This fingerprint information is used later by a
Classification algorithms can be done using the probability localization algorithm online to estimate the location.
of an activity according to a set of features and their Fingerprinting can be used on visual information captured
probabilities [34], the simulates between features according by a camera or audio signals captured by a microphone.
to a dataset, transformation of features to a space that can The IoT devices can also use collaborative localization
help in identifying the type of activity [35, 36], hierarchy methods that are based on mutual position measurement of
based model to map the features to activities based on a each IoT device as well as the relative distance of devices.
decision tree [37, 38, 39], or artificial neural network and The non-training dependent category of localization method
machine learning algorithm [40, 41]. can also be divided into various clusters. The main cluster

VOLUME XX, 2020 9

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
10.1109/ACCESS.2020.2986329, IEEE Access

includes the methods that use geometric information to major problem and there are many applications that
estimate the location based on multi-lateration, wearable devices are used to detect or prevent the falls.
triangulation, region overlapping and other geometric The wearable IoT sensors intended for safety mostly
related methods [62, 63, 64, 65, 66]. The other clusters of focus on three main applications: 1) fall detection and
non-training category are mobility, path planning and prevention 2) drowsiness fatigue detection 3)
statistical [59]. environmental condition monitoring.
A method used in cellular IoT 3rd Generation Partnership 1) FALL DETECTION AND PREVENTION
Project 3GPP Release 14 for location estimation based on Accidental fall is a major issue for elder people. Even if a
geometric cluster is Observed Time Difference of Arrival fall is not fatal, the effect of a fall degrades the quality of
(OTDOA). 3GPP Release 13 does not provide any life substantially. To constantly monitor for fall events, the
positioning information other than Cell Identity (CID), use of wearable devices are more applicable as compared to
which indicates the cell that the IoT device is located at. In the other methods which use a fixed infrastructure such as
Release 14 not only development support for the feature of fixed camera [67] [68] [69] or smart tile [70] due to lack of
OTDOA is presented, but also the standard introduces limitation in location determination or even scene blocking
enhancements to complete the User Equipment (UE) for those using camera. We hereunder discuss the literature
measurement requirements for enhanced CID. Note that that use wearable IoT for the purpose of fall detection or
OTDOA is a downlink-based positioning technique where prediction.
an IoT device measures the times of arrival of the To be able to detect fall usually the inertial sensors such as
Positioning Reference Signals (PRSs) received from several gyroscope or accelerometer are used. The fall detection
nodes relative to a reference node’s PRS transmission to system must be fast enough to detect fall fast to be
form the reference signal time difference (RSTD) beneficial. However, in order to detect fall events
measurements. The RSTD measurements are then accurately and minimize false positive, the fall detection
translated into a geographical hyperbola where the system must differentiate between a fall and other daily
positioned UE is considered to reside. By considering activities [71]. This can be done by finding parameters such
multiple RSTD measurements, the UE position can be as body posture, falling speed or angular velocity [72] [73].
estimated to be at the crossover point of the corresponding It should be noted that some of these parameters not only
hyperbolas. By measuring horizontal position error of the vary from person to person, but also change with aging
LTE-M network as shown in Figure 4, we can show the [74]. Therefore, to be able to detect falls correctly, the fall
cumulative distribution function (CDF) of horizontal error detection algorithms need substantial computation. The
position. It is clear that by using LTE-M network a fundamental idea behind most fall detection and prevention
resolution of approximately 50m is achievable horizontally algorithms is to use the motion information to differentiate
which is suitable for many wearable applications. between a fall and other types of daily activities. A fall can
be characterized as an involuntary move which is not
performed in a controlled manner and causes abrupt motion
with fast speed. The most popular methods for fall
prevention are threshold-based schemes in which the
motion information such as vertical speed profile are
compared with a set of threshold values to detect whether
the person is stable, has fallen, or is about to fall. The
threshold values can be fixed or adaptive values. In general,
the fix value thresholds result in lower computational
complexity, and lower performance as compared to the
adaptive threshold schemes.
Adaptive thresholds can be dynamically calculated based
on history and processing of individual data movement and
motion information [75] [76] [77] or based on clustering of
Figure 4. Simulation results showing the accuracy of positioning in LTE- people according to their gender, height, weight and age
M networks using Release 13 and Release 14.
[78]. In [79] pretested reference templates for each type of
fall using comparison between the angles and angular
D. Safety
velocities of the thigh segment between falls and normal
This category belongs to the wearables that are used to activities are used for fall detection. Some other papers
provide safe environment for the users. For instance, a suggest the use of machine learning for this purpose [80]
fatigue monitoring system can notify the alert the drivers [81] [82] such as using [83, 84, 85] the Support Vector
who fall asleep at the wheel and notify the employers. Or as Machine (SVM) via feature selection procedures among a
another example, the wearable devices can collect the air number of raw signals from the gyroscope and
quality data in the mines to assure the worker’s health and accelerometer sensors.
reduce risks for miners and costs for employers. Fall
prevention and detection especially in elderly people are a 2) DROWSINESS-FATIGUE-DETECTION

VOLUME XX, 2020 9

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
10.1109/ACCESS.2020.2986329, IEEE Access

In general, the drowsiness detection techniques and In [91], a wearable environmental sensor network to
alerting systems can be classified into five categories as 1) monitor the urban environment is presented. The system
image processing-based, 2) EEG-based, 3) artificial neural has seven environmental sensors including infrared
network-based, 4) vehicular-based and 5) subjective temperature sensor, atmospheric pressure, accelerometer,
measures [86]. In all these techniques, powerful processors temperature, relative humidity, ambient light, and IMU and
are required to process the complex computations and the wireless communication is through Wi-Fi. The power
detect drowsiness or fatigue. Another alternative is to send consumption of the system is so low that the rechargeable
the raw data into the cloud and put the burden of processing battery may last up to seven days. The authors in [92],
on the shoulders of the power servers. present a low-power wearable sensor network which
In [87], a drowsiness-fatigue-detection solution to increase measures CO2 concentration, the Earth's magnetic field,
road safety is proposed. The system is based on wearable temperature and relative humidity. The communication
smart glasses and is composed of a pair of wearable smart scheme of this sensor network is based on Bluetooth
glasses, an in-vehicle infotainment telematics platform, an technology.
onboard diagnostics-II-based automotive diagnostic bridge,
an active vehicle rear light alert mechanism, and a cloud- E. Others
based management platform. The detection system is Besides the four main clusters discussed so far, there are
capable to detect whether the driver is drowsy or tired in some papers that have used wearable IoT technology for
real time and once such a condition is detected, the other other applications such as education [93, 94], virtual and
drivers are automatically alerted through the active vehicle augmented reality [95, 96], law enforcement [97], or spying
real light alert mechanism. The collected data is also [98]. Since there have not been lots of attention by the
concurrently sent to a cloud-based management platform. research community for these applications, we do not
In [88], to detect the sleepy workers a low-cost EEG-based further discuss these use cases.
system is proposed. The modified safety smart helmet is
worn by the worker and transmits the data to a local server III. CIOT-ENABLED WEARABLE
which runs a random forest classifier algorithm to verify if We found that use of CIoT wearable has not received
the sleep condition is sever to alert the shift supervisor. If enough attention in the literature. In most cases the
the worker falls down, a single Inertial Measurement Unit connection to the Internet for the wearable device is
(IMU) sensor can detect the fall. A mobile application is
through the technologies such as Wi-Fi or Bluetooth Low
also designed where the supervisor can constantly monitor
Energy (BLE). The use of these two technologies which
the worker’s status.
exist in any cellphone device has been discussed in many
3) ENVIRONMENTAL CONDITION MONITORING: papers. The fact that all people carry a cellphone these
Hazardous environmental situations may result in serious days, even though not completely true, has become the
health problems and it is crucial to deploy fast response assumption that has been used in many papers as
systems in such environments to alert the workers.
justification for using cellphone as part of connectivity to
Moreover, the data gathered by the environmental wearable
the Internet. We have studied BLE technology and its
sensors provide useful insight about the environmental
practical strengths and limitations extensively [99, 100,
impact on subjects' health. The wearables may also be
installed on vehicles or animals to form a mobile wireless 101, 102, 103, 104, 105, 106, 107] and believe that BLE is
sensor networks to predict hazardous conditions. For one of the enablers of IoT technology that has a good
example the data gathered from the animals in different potential to be used in wearable devices. But, it is quite
farms may help scientist to predict earthquake in a specific possible that CIoT will be the most widely used technology
area. in the future for wearable devices. We believe the use of
The researchers in [89], presented a safety wearable cellular IoT will be increased in a fast pace in future years
equipped with a self-powered sensor network which uses and it will replace other types of IoT communication
Maximum Power Point Control (MPPT) solar energy technologies such as Wi-Fi, or unlicensed Low Power Wide
harvesting, several low power environmental sensing Area Networks (LPWAN) solutions (e.g., LoRa and
nodes, and unlicensed log range wireless modules (i.e., Sigfox) in many use cases.
LoRa). The wearable sensors measure the temperature, It is shown in the past that if the cost of equipment as well
relative humidity, and ultraviolet (UV) index in the as data plan for cellular systems implemented by
surrounding area. The collected measurements are sent to a telecommunication companies are decreased, the users tend
gateway and a cloud server through a LoRa connectivity to prefer to use cellular technology. For example, in 90’s
system. when telecommunication companies provided cellphones
In [90], a wearable safety application is proposed to detect with limited incoming and outgoing voice usage hours in
hazardous conditions in the early stages and alert the their plan, the users would use cellphones mostly outside of
workers. The system detects CO2, the temperature and their homes and offices, where landline systems were not
relative humidity and then the data is communicated using available. As telecommunication companies reduced the
an XBee DigiMesh module. price and offered unlimited calls during evenings and

VOLUME XX, 2020 9

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
10.1109/ACCESS.2020.2986329, IEEE Access

weekends and consequently unlimited incoming and low power consumption systems or energy harvesting
outgoing voice calls, users showed interest to use their techniques such as micro-magneto-electric, thermoelectric,
cellphone even in places where landline systems were piezoelectric, or photoelectric harvesting methods are some
available. We have seen the same user behavior in data possible options used in [108]. Among all energy
usage over the Internet. In the past, due to the cost of data harvesting methods, solar energy is considered as a strong
usage and communication speed, users preferred to switch candidate as it provides the highest power density. The
to Wi-Fi and connect to hotspots instead of cellular disadvantage of solar energy is its limitation to day times
connection. However, as the cost of data usage is and outdoor places.
decreasing, more and more mobile users use their cellular 3) WEARABILITY:
connection to connect to the Internet in places where Wi-Fi The wearable IoT devices have to be comfortable when
connection is also available. Therefore, while there exist worn by the user. It is important that they are light and
some IoT enabling technologies today that may be used for designed in such a way that they do not disturb the normal
wearable devices in local area or solve the wide area activities of the user. The tradeoff between the complexity
coverage requirement of the IoT devices by using a cellular of the computations and the weight of the wearable is one
connection through a cellphone, they fall short as compared of the major challenges. In [109], the idea of smart clothing
or wearable 2.0 for human-cloud integration is introduced
to two existing CIoT technologies of the 3rd Generation
which tries to solve the problem with discomfort caused by
Partnership Project (3GPP), which are LTE-M and Narrow
wearing multiple sensors in different part of the body for
Band IoT (NB-IoT) in terms of coverage, scalability,
health care application.
interoperability, QoS, and security. LTE-M and NB-IoT are
introduced in Release 13 of LTE technology. In LTE 4) SAFETY:
Release, 14, and 15, the enhancements of LTE IoT All wearable IoT devices use wireless technologies to
transmit their sensed data to another node, gateway or a
continued to provide cellular IoT connectivity to more IoT
base station. This wireless transmission involves
devices and in more diverse applications. It should be noted
radiofrequency radiation which could have negative impact
that the performance of CIoT will further hugely be
on the user’s health because the transceiver antennas are
advanced with more implementation of 5G cellular very close to his/her body. In those wearables that are worn
technology. on the head or eyes the radiation risks could be significantly
higher. This safety concern is well addressed in [110] by
IV. IOT- ENABLED WEARABLE CHALLENGES AND
reviewing the standard limits of human exposure to radio
FUTURE POSSIBILITIES
frequency electromagnetic energy and analyzing the
The major challenges of the wearable IoT devices as shown
radiation level of CIoT antennas. It is shown that the
in Figure 5, are listed below:
problem can be worse when the wearable CIoT device is
transmitting in areas with poor coverage.
5) SECURITY:
The complexity of the wearable IoT devices is typically
reduced due to lightweight and less power consuming
designs. Consequently, there could be less strong security
features on such devices. One of the challenges in wearable
IoT devices is how to implement security policies while
keeping the complexity of the system as low as possible. In
general, wearables are easy hacking targets due to poor
encryption and protection.
6) REGULATION:
Figure 5. Major challenges of wearable IoT technology There is currently a limitation in using wearable IoT
devices in many industries due to lack of existence of
1) DATA RESOLUTION OF WEARABLE SENSORS proper regulations. For example, in sports fields and arenas,
Since it is of high importance that the wearable device must the use of wearable IoT devices are technologically
be comfortable when worn by the user and should consume feasible, but it is not being used due to the leagues
low amount of energy, they are typically small in size and regulations.
the sensors have lower resolution compared to non- 7) PRIVACY
wearable devices. The constant exchange of personal data such as vital health
2) POWER CONSUMPTION signals, dosage, and location between the wearable and the
To minimize the human interaction and for wearable IoT hub can create an environment for privacy breaches.
devices to operate for long hours without replacing or Typically, wearable IoT devices are on broadcast mode
charging the battery, special considerations are to be taken which makes them easily discoverable by other nodes in the
into account while designing the wearables. For example, network. Unauthorized nodes can steal the personal data if

VOLUME XX, 2020 9

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
10.1109/ACCESS.2020.2986329, IEEE Access

appropriate privacy policies are not applied. In such 1-10, 2018.

broadcast modes, the built-in hardware security technology [6] A. Ghorbel, S. Bouguerra, N. B. Amor, M. Jallouli, “Cloud based mobile
application for remote control of intelligent wheelchair,” 14th
of the IoT devices might not guarantee the protection of International Wireless Communications & Mobile Computing Conference
personal data against breaches. In [111] a broadcast- (IWCMC), Limassol, Cyprus, 2018.
subscriber IoT model is proposed where the users’ personal [7] M. Opoku, A. Al-Mahmood, “Design and implimentation of a wearable
device for motivating patients with upper and/or lower limb disability via
data are only shared with intended nodes such as healthcare gaming and home rehabiliation,” Fouth international conference on fog
facilities or devices authorized by the user. and mobile edge computing , 2019.
There will be enormous number of applications and [8] C. F. Pasluosta, H. Gassner, J. Winkler, J. Klucken and B. M. Eskofier,
“An Emerging Era in the Management of Parkinson's Disease: Wearable
opportunities for wearable IoT devices. Using 3D printing, Technologies and the Internet of Things,” Biomedical and Health
the next generation of these wearables will be boosted by Informatics, vol. 19, pp. 1873-1881, 2015.
lateralization of their design, fabrication, and distribution. [9] S. B. Baker, W. Xiang, and I. Atkinson, “Internet of Things for Smart
For example by using this technology for health Healthcare: Technologies, Challenges, and Opportunities,” IEEE Access,
pp. 26521-26544, 2017.
application, the patient can upload their disease profiles and S. Jayanth, M. B. Poorvi, R. Shreyas, B. Padmaja and M. P. Sunil,,
[10]
download his/her individualized treatment tools to print “Wearable device to measure heart beat using IoT,” International
them at home with low cost. By improvement in new Conference on Inventive Systems and Control (ICISC), Coimbatore, 2017.
energy harvesting technologies and existence of longer life [11] K. Majumder, Y. ElSaadany, R. Young, “An Energy Efficient Wearable
Smart IoT System to Predict Cardiac Arrest,” Advances in Human-
batteries, the greater computational power will be allowed Computer Interaction, vol. 2019, 2019.
in wearable IoT devices. This increases their autonomy, [12] A. Brezulianu, “IoT Based Heart Activity Monitoring Using Inductive
implementation of feature extraction and classification Sensors,” Journal of Sensors (Basel, Switzerland) , vol. 19, 2019.
and/or prediction, and allows an efficient employment of [13] S. Milici, J. Lorenzo, A. Lázaro, R. Villarino, and D. Girbau, “Wireless
breathing sensor based on wearable modulated frequency selective
hardware resources. The most important opportunity in surface,” IEEE sensors , vol. 17, pp. 1285-1292, 2017.
werable IoT will be realized when IoT data from various [14] S. Syed Tauhid, B. Faizan, D. Faheem, A. Nouman, & A. Jan, “Cloud-
sources and devices are combined there will be an Assisted IoT-Based Smart Respiratory Monitoring System for Asthma
Patients”, Applications of Intelligent Technologies in Healthcare, 2019.
integrated IoT system available.
[15] I. Mahbub et al., “A low-power wireless piezoelectric sensor-based
respiration monitoring system realized in CMOS process,” IEEE sensors,
V. CONCLUSION vol. 17, pp. 1858-1864, 2017.
Wearable IoT can offer endless new opportunities in many [16] D. N. Hernández et al., “Smart Vest for Respiratory Rate Monitoring of
COPD Patients Based on Non-Contact Capacitive Sensing,” Sensors
real-life applications. This paper provided a survey on the (Basel, Switzerland), vol. 18, 2018.
most important efforts of the research community in the [17] J. Wan, M. Al-awlaqi, M. Li, et al, “Wearable IoT enabled real-time health
area of wearable and IoT. After identifying more than one monitoring system,” Wireless Communnication Network , 2018.
hundred papers as the significant literature in this area, the [18] S. Yoshida, H. Miyaguchi, T. Nakamura, "DEvelopment of tablet-shaped
ingestible core-body thermometer powered by gestric acid battery," IEEE
papers were analyzed and divided in four major clusters sensors, vol. 18, pp. 9755-9762, 2018.
based on their applications. The methods used for each [19] F. Lamonaca et al., “An Overview on Internet of Medical Things in Blood
cluster also were grouped. Pressure Monitoring,” IEEE International Symposium on Medical
Measurements and Applications (MeMeA), Istanbul, Turkey, 2019.
Wearables have lots of potential when integrated IoT
[20] D. Murali, D. R. Rao, S. R. Rao and M. Ananda , “Pulse Oximetry and
system becomes available. For this reason the true power of IOT based Cardiac Monitoring Integrated Alert System,” International
combining wearable and IoT has not been recognized. Also Conference on Advances in Computing, Communications and Informatics
cellular IoT can revolutionize the wearable IoT industry (ICACCI) , Bangalore, 2018.

which at this point has not received lots of attention among [21] B. Sargunam, S. Anusha, “IoT based mobile medical application for smart
insulin regulation,” IEEE International Conference on Electrical,
research community. Computer and Communication Technologies (ICECCT), Coimbatore,
India, 2019.
ACKNOWLEDGMENT [22] S. Yang et al., “IoT Structured Long-Term Wearable Social Sensing for
Mental Wellbeing,” IEEE Internet of Things Journal, vol. 6, pp. 3652-
This work has been supported by the British Columbia 3662, 2019.
Institute of Technology research grant. [23] J. Qi, P. Yang, M. Hanneghan, S. Tang and B. Zhou, “A Hybrid
Hierarchical Framework for Gym Physical Activity Recognition and
Measurement Using Wearable Sensors,” IEEE internet of things journal,
REFERENCES vol. 6, no. 2, April 2019.
[1] S. Seneviratne et al, “A survey of wearable devices and challenges,” IEEE
[24] E. Mencarini, A. Rapp, Lia Tirabeni, and M. Zancanaro, “Designing
COMM. surveys & tutorials, vol. 19, no. 4, pp. 2573-2620, fourth quarter
wearable systems for sports: a review of trends and opportunities in
2017.
Humman – computer interaction,” IEEE Trans. On human machine
[2] K. Hartman, “Make: Wearable Electronics: Design, Prototype, and Wear systems,, vol. 49, no. 4, August 2019.
Your Own Interactive Garments,” Maker Media, Sebastopol, CA, USA,
[25] D. Castro, W. Coral, C. Rodriguez, J. Cabra, and J. Colorado, "Wearable-
2014.
Based Human Activity Recognition Using an IoT Approach," Journal of
[3] E. Sazonov and M. R. Neuman, “Wearable Sensors: Fundamentals, sensor and actuators networks, vol. 6, no. 28, 2017.
Implementation and Applications,” Elsevier, Amsterdam, Netherlands,
[26] H. Huang , X. Li, S. Liu, S. Hu, and Y. Sun, "TriboMotion: A Self-
2014.
Powered Triboelectric Motion Sensor in Wearable Internet of Things for
[4] C. Nave and O. Postolache, “Smart Walker based IoT Physical Human Activity Recognition and Energy Harvesting," IEEE Internet of
Rehabilitation System,” International Symposium in Sensing and Things Journal, vol. 5, no. 6, December 2018.
Instrumentation in IoT Era (ISSI), Shanghai, China, 2018.
[27] W. Lu, F. Fan, J. Chu, P. Jing, and Y. Su, "Wearable Computing for
[5] G. Yang, J. Deng, G. Pang, H. Zhang, J. Li, B. Deng, Z. Pang, et al., “An Internet of Things: A Discriminant Approach for Human Activity
IoT-Enabled Stroke Rehabilitation System Based on Smart Wearable Recognition," IEEE Internet of Things Journal, vol. 6, no. 2, April 2019.
Armband and Machine Learning,” IEEE J. Transl. Eng. Health Med. , pp.

VOLUME XX, 2020 9

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
10.1109/ACCESS.2020.2986329, IEEE Access

[28] L. Atallah, B. Lo, R. King, and G.-Z. Yang, "Sensor positioning for [51] J. Haladjan, M. Reif, and B. Brügge, "VIHapp: A wearable system to
activity recognition using wearable accelerometers," IEEE Trans. Biomed. support blind skiing," ACM Int. Joint Conf. Pervasive Ubiquitous
Circuits Syst., vol. 5, no. 4, p. 320–329, Aug. 2011. Comput./ACM Int. Symp Wearable Comput., pp. 1033–1037, 2017.
[29] M. Mathie, B. G. Celler, N. H. Lovell, and A. C. F. Coster, "Classification [52] E. Niforatos, A. Fedosov, I. Elhart, and M. Langheinrich, “Augmenting
of basic daily movements using a triaxial accelerometer," Med. Biol. Eng. skiers’ peripheral perception,” ACM Int. Symp.WearableComput., pp.
Comput., vol. 42, no. 5, p. 679–687, 2004. 114–121, 2017.
[30] A. Mannini and A. M. Sabatini, "Machine learning methods for classifying [53] E. H. Chi, J. Song, and G. Corbin, “Killer App’ of wearable computing:
human physical activity from on-body accelerometers," Sensors j. , vol. Wireless force sensing body protectors for martial arts,” 17th Annu. ACM
10, no. 2, p. 1154–1175, 2010. Symp. User Interface Softw. Technol., pp. 277–285, 2004.
[31] K. H. Walse, R. V. Dharaskar, and V. M. Thakare, "A study on the effect [54] M. Pan, S. Salvi, and E. Brady, “Designing auditory feedback from
of adaptive boosting on performance of classifiers for human activity wearable weightlifting devices,” Extended Abstr. Hum. Factors Comput.
recognition," Int. Conf. Data Eng. Commun. Technol., pp. 419–429, 2017. Syst., Paper LBW561., 2018.
[32] Y. Chen and C. Shen, "Performance analysis of smartphone-sensor [55] H. Havlucu, I. Bostan, A. Coskun, and Ö. Özcan, “Understanding the
behavior for human activity recognition," IEEE Access, vol. 5, p. 3095– lonesome tennis players: Insights for future wearables,” Extended Abstr.
3110, 2017. Hum. Factors Comput. Syst., pp. 1678–1685, 2017.
[33] Z. He and L. Jin, "Activity recognition from acceleration data based on [56] M. Lapinski, E. Berkson, T. Gill, M. Reinold, J. A. Paradiso, “A
discrete cosine transform and SVM," IEEE Int. Conf. Syst. Man Cybern., Distributed Wearable, Wireless Sensor System for Evaluating Professional
pp. 5041–5044., 2009. Baseball Pitchers and Batters,” September 2009.
[34] L. Liu, S. Wang, G. Su, Z.-G. Huang, and M. Liu, "Towards complex [57] D. K. Arvind, and A. Bates, “The speckled golfer,” 3rd International
activity recognition using a Bayesian network-based probabilistic Conference on Body Area Networks, 2008.
generative framework," Pattern recognition, vol. 68, p. 295–309, Aug. H. Ghasemzadeh, and R. Jafari, “Sport Training Using Body Sensor
[58]
2017. Networks: A Statistical Approach to Measure Wrist Rotation for Golf
[35] K.-C. Liu, C.-Y. Yen, L.-H. Chang, C.-Y. Hsieh, and C.-T. Chan, Swing,” The Fourth International Conference on Body Area Networks
"Wearable sensor-based activity recognition for housekeeping task.," (BodyNets 09),, Los Angeles, CA, April 1-3 2009, April 2009.
IEEE Int. Conf. Wearable Implantable Body Sensor Netw., pp. 67–70, R. Shit, S. Sharma, d. Puthal, and A.Zomaya, “Location of things (LOT): a
[59]
2017. review and taxonomy of sensors localization in IoT infrastructure,” IEEE
[36] B. Logan, J. Healey, M. Philipose, E. M. Tapia, and S. Intille, "A longterm Comm. Surveys & Tutorials, vol. 20, no. 3, pp. 2018-2061, Third quarter
evaluation of sensing modalities for activity recognition," UbiComp 2018.
Ubiquitous Comput., vol. 4717., pp. 483–500, Innsbruck, Austria, 2007. Y. Kim, Hyojeong Shin and H. Cha, “Smartphone-based Wi-Fi
[60]
[37] S. M. Lee, S. M. Yoon, and H. Cho, "Human activity recognition from pedestriantracking system tolerating the RSS variance problem,” IEEE Int.
accelerometer data using convolutional neural network," IEEE Int. Conf. Conf. Pervasive Comput. Commun. (PerCom), pp. 11–19., Lugano,
Big Data Smart Comput., pp. 131–134, 2017. Switzerland,, 2012.
[38] M. Ermes, J. PÄrkkÄ, J. MÄntyjÄrvi and I. Korhonen, "Detection of daily [61] A. Hallquist and A. Zakhor, “Single view pose estimation of mobile
activities and sports with wearable sensors in controlled and uncontrolled devices in urban environments,” IEEE Workshop Appl. Comput. Vis.
conditions," IEEE Trans. Inf. Technol. Biomed. , vol. 12, no. 1, p. 20–26, (WACV), pp. 347–354., Tampa, FL, USA, 2013.
Jan. 2008. B. Wang, G. Wu, S. Wang, and L. T. Yang, “Localization based on
[62]
[39] M. Li et al., "Multimodal physical activity recognition by fusing temporal adaptive regulated neighborhood distance for wireless sensor networks
and cepstral information," IEEE Trans. Neural Syst. Rehabil. Eng., vol. with a general radio propagation model,” IEEE Sensors J., vol. 14, no. 11,
18, no. 4, p. 369–380, Aug. 2010. p. 3754–3762, Nov. 2014.
[40] S. Liu, R. X. Gao, D. John, J. W. Staudenmayer, and P. S. Freedson., [63] H. Chen, P. Huang, M. Martins, H. C. So, and K. Sezaki, “Novel centroid
"Multisensor data fusion for physical activity assessment," IEEE Trans. localization algorithm for three-dimensional wireless sensor networks,”
Biomed. Eng., vol. 59, no. 3, p. 687–696, 2012. 4th IEEE Int. Conf. Wireless Commun. Netw. Mobile Comput. (WiCOM),
I. C. Gyllensten and A. G. Bonomi, “Identifying types of physical activity pp. 1–4, 2008.
[41]
with a single accelerometer: Evaluating laboratory-trained algorithms in [64] H. Yu, X. Chen, and J. Fan, “Gauss–Newton method based on energy
daily life,” IEEE Trans. Biomed. Eng., vol. 58, no. 9, p. 2656–2663, 2012. target localization,” Comput. Eng. Appl., vol. 43, no. 27, p. 124–126,
Z. Wang, M. Guo and C. Zhao, “Badminton stroke recognition based on 2007.
[42]
body sensor networks,” IEEE Trans. Hum.-Mach. Syst.,, vol. 46, no. 5, p. [65] L. Lazos and R. Poovendran, “HiRLoc: High-resolution robust
769–775, Oct. 2016. localization for wireless sensor networks,” IEEE J. Sel. Areas Commun.,
A. Raina, T. G. Lakshmi and S. Murthy, "CoMBaT:Wearable technology vol. 24, no. 2, p. 233–246, Feb. 2006.
[43]
based training system for novice badminton players," IEEE 17th Int. Conf. [66] W. Jichun, H. Liusheng, X. Hongli, X. Ben, and L. Shanliang, “A novel
Adv. Learn. Technol., pp. 153–157, 2017. range free localization scheme based on voronoi diagrams in wireless
J. C. Maglott, J. Xu and P. B. Shull, "Differences in arm motion timing sensor networks,” J. Comput. Res. Develop., vol. 5, no. 1, p. 119–125,
[44]
characteristics for basketball free throwand jumpshooting via a body-worn 2008.
sensorized sleeve," IEEE 14th Int. Conf. Wearable Implantable Body [67] B. Mirmahboub, S. Samavi, N. Karimi, and S. Shirani, “Automatic
Sensor Netw pp. 31–34, 2017. monocular system for human fall detection based on variations in
S. Bogers, C. Megens, "Design for balanced engagement in mixed level silhouette area,” IEEE Trans. Biomed. Eng, vol. 60, no. 2, p. 427–436,
[45]
sports teams," SIGCHI Conf. Extended Abstr. Hum. Factors Comput. Feb. 2013.
Syst., pp. 994–1002., 2017. [68] E. Auvinet, F. Multon, A. Saint-Arnaud, J. Rousseau and J. Meunier, “Fall
Z. Wang, H. Zhao, S. Qiu and Q. Gao, "Stance phase detection for detection with multiple cameras: An occlusion-resistant method based on
[46]
ZUPTaided foot-mounted pedestrian navigation system," IEEE/ASME 3-D silhouette vertical distribution,” IEEE Trans. Inf. Technol. Biomed,
Trans. Mechatron, vol. 20, no. 6, p. 3170–3181, Dec. 2018. vol. 15, no. 2, p. 290–300 , Mar. 2011.
Z. Wang, Jiaxin Wang, H. Zhao, Ning Yang and G. Fortino, “CanoeSense: [69] C. Rougier, J. Meunier, A. St-Arnaud, and J. Rousseau, “Robust video
[47]
Monitoring canoe sprint motion using wearable sensors,” IEEE Int. Conf. surveillance for fall detection based on human shape deformation,” IEEE
Syst., Man, Cybern., pp. 644–649, Budapest,, 2016. Trans. Circuits Syst. Video Technol., vol. 21, no. 5, p. 611–622, May
2011.
[48] J. Pansiot, B. Lo, and G. Z. Yang, “Swimming stroke kinematic analysis
with BSN,” Int. Conf. Body Sensor Netw., , pp. 153–158, Singapore, 2010. [70] M. Daher, A. Diab, M. El Badaoui El Najjar, M. Ali Khalil and F.
Charpillet, “Elder tracking and fall detection system using smart tiles,”
[49] M. Bächlin, K. Förster, and G. Tröster, “SwimMaster: A wearable IEEE Sensors Journal, vol. 17, no. 2, p. 469–479 , Jan. 2017.
assistant for swimmer," Int. Conf. Ubiquitous Comput., pp. 215–224.,
2009. [71] W. Saadeh, S. Butt, and M. Altaf, “A patient-specific single sensor IoT
based wearable fall prediction and detection system,” IEEE Trans. On
[50] J. Häkkilä, M. Alhonsuo, L. Virtanen, J. Rantakari, A. Colley, and T. neural systems and rehabilitation engineering, vol. 27, no. 5, May 2019.
Koivumäki, “MyData approach for personal health - A service design case
for young athletes,” 49th Int. Conf. Syst. Sci., pp. 3493–3502, Hawaii, [72] A. Sorvala, E. Alasaarela, H. Sorvoja, and R. Myllylä, “A two-threshold
2016. fall detection algorithm for reducing false alarms,” 6th Int. Symp. Med. Inf.
Commun. Technol., pp. 1–4., Mar. 2012,.

VOLUME XX, 2020 9

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
10.1109/ACCESS.2020.2986329, IEEE Access

[73] W. Saadeh, M. A. Bin Altaf and S. A. Butt, “A wearable Prattichizzo, and M. A. Otaduy, “Optimization-Based Wearable Tactile
neurodegenerative diseases detection system based on gait dynamics,” Rendering,” IEEE Trans. on haptics, vol. 10, no. 2, pp. 254-264, Apr.
IFIP/IEEE Int. Conf. Very Large Scale Integr, pp. 1–6, Oct. 2017. 2017.
[74] R. Bartlett, “Introduction to Sports Biomechanics: Analysing Human [96] M. Maisto, C. Pacchierotti , F. Chinello, G. Salvietti, A. De Luca, and D.
Movement Patterns,” 2nd ed., Routledge: Abingdon, U.K, 2007. Prattichizzo, “Evaluation of Wearable Haptic Systems for the Fingers in
L.Ren, and Y.Peng, , “Research of fall detection and fall prevention Augmented Reality Applications,” IEEE Trans. on haptics, vol. 10, no. 4,
[75]
thechnologies : a systematic review,” IEEE Access, vol. 7, pp. 77702- pp. 511-522, Oct. 2017.
77722, June 2019. [97] R. Silberglitt, A. Lauland, M. Watson, C. A. Eusebi, J. Lastunen,
N. Otanasap, “Pre-impact fall detection based on wearable device using “Weaable technology and law enforcement,” [Online]. Available:
[76]
dynamic threshold model,” 17th Int. Conf. Parallel Dis- trib. Comput., https://www.rand.org/content/dam/rand/pubs/research_reports/RR2000/R
Appl. Technol. (PDCAT),, pp. 362-365, Guangzhou, China,, Dec. 2016. R2012/RAND_RR2012.pdf. [Accessed 10 12 2019].
Y. Wu, Y. Su, Y. Hu, N. Yu, and R. Feng, “A multi-sensor fall detection [98] C. Wang, J. Liu, X. Guo, Y. Wang , Y. Chen, “WristSpy: Snooping
[77]
system based on multivariate statistical process analysis,” J. Med. Biol. Passcodes in Mobile Payment Using Wrist-worn Wearables,” IEEE
Eng., vol. 39, no. 3, pp. 336-351, May 2018. INFOCOM, pp. 2071-2079, June 2019.

L. R. Shi, “Chameleon: Personalised and adaptive fall detection of elderly [99] F. J. Dian, A. Yousefi, S. Lim, “A practical study on Bluetooth Low
[78]
people in home-based environments,” Int. J. Sens. Network, vol. 20, no. 3, energy (BLE) throughput,” IEEE IEMCON, pp. 768-771, Vancouver,
pp. 163-176, 2016. Nov. 2018.

[79] M. N. Nyan, F. E. H. Tay, and E. Murugasu, “A wearable system for pre- [100] F. J. Dian, “Low-power Synchronized Multi-channel Data Acquisition
impact fall detection,” J. Biomech, vol. 41, p. 3475–3481, 2008. Communication System,” IEEE CCWC, pp. 1027-1031,, Las vegas, Jan.
2019.
[80] S. Yu, H. Chen, and R. A. Brown, “Hidden Markov model-based fall
detection with motion sensor orientation calibration: A case for reallife [101] F. J. Dian, A. Yousefi, K. Somaratne , “A study in accuracy of time
home monitoring,” IEEE J. Biomed. Health Inform, vol. 22, no. 6, pp. synchronization of BLE devices using connection-based event,” IEEE
1847-1853, Nov. 2018. IEMCON, pp. 595 – 601, Vancouver, OCT. 2017.

[81] X. Zhou, L.-C. Qian, P.-J.You, Z.-G. Ding, and Y.-Q. Han, “Fall detection [102] K. Somaratne, F. J. Dian and A. Yousefi, “Analysis of time
using convolutional neural network with multi-sensor fusion,” in IEEE Int. synchronization based on current measurement for Bluetooth Low Energy
Conf. Multimedia Expo Workshops (ICMEW), San Diego, CA, USA, pp.1- (BLE),” IEEE IEMCON, pp. 602 - 607, Vancouver, OCT. 2017.
5, Jul. 2018. [103] K. Somaratne, F. J. Dian, A. Yousefi, "Accuracy analysis of time
W. Min, H. Cui, H. Rao, Z. Li, and L. Yao, “Detection of human falls on synchronization using current consumption pattern of BLE devices," IEEE
[82]
furniture using scene analysis based on deep learning and activity CCWC, pp. 841-844 ,, Las vegas, Jan. 2018.
characteristics,” IEEE Access,, vol. 6, pp. 9324-9335, 2018. [104] F. J. Dian, A. Yousefi, K. Somaratne, "Performance evaluation of time
J. Lee, S. Robinvitch, and E. j. park, “Internal sensing-based pre-impact synchronization using current consumption pattern of BLE devices," IEEE
[83]
detection of falls involving near-fall scenarios,” IEEE Trans. on neural CCWC, pp. 906-910, Las vegas, Jan. 2018.
systems and rehabilitation eng., vol. 23, no. 2, March 2015 . [105] F. John Dian, “An analytical scheme for power consumption of battery-
G. Shi, C. S. Chan, W. J. Li, K. Leung, Y. Zou, and Y. Jin, “Mobile operated peripheral BLE nodes,” 9th IEEE CCWC, pp. 1021-1026, las
[84]
human airbag system for fall protection using MEMS sensors and vegas, Jan. 2019.
embedded SVM classifier,” IEEE Sensors J., vol. 9, no. 5, p. 495–503 , [106] F. J. Dian, A. Yousefi, S. Lim, “Time scheduling of central BLE for
May 2009. connection events,” IEEE IEMCON, pp. 763-767, Vancouver, Nov. 2018.
[85] S. Shan and T. Yuan, “A wearable pre-impact fall detector using feature [107] F. John Dian, R. Vahidnia, “Formulation of BLE Throughput Based on
selection and support vector machine,” IEEE Int. Conf. Signal Process. pp. Node and Link Parameters,” IEEE Canadian journal of Electrical and
1686–1689, Beijing, China, 2010. Computer Enginerring, 2020.
[86] P. Choudhary, R. Sharma, G. Singh, and S. Das, “A survey paper on [108] T. Wu, F. Wu, J. M. Redouté, and M. R. Yuce, “An autonomous wireless
drowsiness detection & alarm system for drivers,” International Research body area network implementation towards IoT connected healthcare
Journal of Engineering and Technology (IRJET), vol. 3, pp. 1433-1437, applications,” IEEE Access, vol. 5, p. 11413–11422, 2017.
2016.
[109] [M. chen, y. Ma, Y. Li, D. Wu, Y. Zhang, and C. Youn, “Wearable 2.0:
[87] W. Chang, L. Chen and Y. Chiou, “Design and Implementation of a Enabling human-cloud integration in next generation healthcare systems,”
Drowsiness-Fatigue-Detection System Based on Wearable Smart Glasses IEEE communications magazine, pp. 54-61, January 2017.
to Increase Road Safety,” IEEE Transactions on Consumer Electronics ,
[110] J. Dian and R. Vahidnia, “Radiation Safety Hazards of Cellular IoT
vol. 64, pp. 461-469, 2018. Devices,” IEEE Technology and Society Magazine (submitted), 2019.
[88] S. Dhole, “A novel helmet design and implementation for drowsiness and
[111] R. K. Lomotey, K. Sofranko, and R. Orji “Enhancing Privacy in Wearable
fall detection of workers on-site using {EEG} and Random-Forest IoT through a Provenance Architecture,” Multimodal Technologies and
Classifier,” in The 10th International Conference on Ambient Systems, Interaction; Basel, 2018.
Networks and Technologies, 2019.
[89] F. Wu, J. Redouté and M. R. Yuce, “A Self-Powered Wearable Body
Sensor Network System for Safety Applications,” IEEE sensors, 2018. F. John Dian received his B.Sc. degree in 1990
New Delhi, 2018, pp. 1-4. and M.S. degrees in 1994 from Tehran
[90] D. Antol´ın, N. Medrano, B. Calvo, and F. P´erez, “A wearable wireless Polytechnics, Tehran and his Ph.D. degree in 2004
sensor network for indoor smart environment monitoring in safety from Concordia University, Montreal all in
applications,” IEEE sensors, vol. 17, p. 365, 2017. Electrical and Computer Engineering. He has over
[91] E. Wilhelm et al., “Wearable Environmental Sensors and Infrastructure for 20 years of experience in the areas of design and
Mobile Large-Scale Urban Deployment,” IEEE sensors, vol. 16, pp. 8111- consulting with various companies. Since 2010, he
8123, 2016. has been a faculty with the school of Energy,
[92] M. Serbanescu, V. M. Placinta, O. E. Hutanu and C. Ravariu, “Smart, low Department of Electrical and Computer
power, wearable multi-sensor data acquisition system for environmental Engineering at British Columbia Institute of Technology (BCIT) in
monitoring,” 10th International Symposium on Advanced Topics in Vancouver, Canada. His research interests include IoT, multimedia
Electrical Engineering (ATEE), Bucharest, 2017.
processing for telecommunication applications and Big data.
[93] J. R. Byrne, K. O’Sullivan, and K. Sullivan, “An IoT and Wearable
Technology Hackathon for Promoting Careers in Computer Science,”
IEEE Trans. on education, vol. 60, no. 1, pp. 50 -58, Feb. 2017.
[94] E. de la Guía, V. L. Camacho, L. Orozco-Barbosa, V. M. Brea Luján, V.
M. R. Penichet and M. Lozano Pérez, “Introducing IoT and Wearable
Technologies into Task-Based Language Learning for Young Children,”
IEEE Trans. on learning technologies, vol. 9, no. 4, pp. 366-378, Oct.
2016.
[95] A. G. Perez, D. Lobo, F. Chinello, G. Cirio, M. Malvezzi, J. Martın, D.

VOLUME XX, 2020 9

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
10.1109/ACCESS.2020.2986329, IEEE Access

Reza Vahidnia (M’12) was born in Tehran, Iran.

He received the B.Sc. and M.Sc. degrees in
Electrical Engineering from Tehran and Tarbiat
Modares University, Tehran, Iran, in 2006 and
2009, respectively. He received his Ph.D. degree
from Ontario Tech University, Oshawa, Canada in
2014. From 2014 to 2015 he was a postdoctoral
fellow at Ryerson University where he conducted
research on energy harvesting systems. Since 2016,
he has served as a senior IoT specialist at Rogers
Communications Inc. and TELUS. He is currently a faculty at the
Department of Electrical and Computer Engineering at British Columbia
Institute of Technology (BCIT) in Vancouver, Canada. His research
interests include wireless communications, Internet of Things (IoT) and
signal processing.

Alireza Rahmati was born in Tehran, Iran. He

received the B.Sc. degree in electrical engineering
from Sharif University of Technology, Tehran, Iran
in 2004, and Ph.D. degree in telecommunication
systems engineering from Tarbiat Modares
University, Tehran, Iran in 2012. Currently, he is a
Post-Doctoral Fellow with the Department of
Electrical and Computer Engineering, University of
Waterloo. He was with Ryerson University from
2016 to 2018. From 2007 to 2015, he was with the
Research and Electronic Center Artaaira, Tehran. His main research
interests are statistical signal processing, Internet of things, 5G networks,
and full-duplex radio.

VOLUME XX, 2020 9

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.