Received February 13, 2019, accepted March 2, 2019, date of publication April 11, 2019, date of current version

April 25, 2019.

Digital Object Identifier 10.1109/ACCESS.2019.2910411

Fault-Tolerant mHealth Framework in the

Context of IoT-Based Real-Time Wearable
Health Data Sensors
O. S. ALBAHRI1 , A. S. ALBAHRI2 , A. A. ZAIDAN 1 , B. B. ZAIDAN1 , M. A. ALSALEM1 ,
A. H. MOHSIN1 , K. I. MOHAMMED1 , A. H. ALAMOODI1 , SHAHAD NIDHAL3 ,
ODAI ENAIZAN4 , M. A. CHYAD1 , KARRAR HAMEED ABDULKAREEM5 ,
E. M. ALMAHDI1 , GHAILAN A. AL. SHAFEEY1 , M. J. BAQER1 , ALI NAJM JASIM1 ,
N. S. JALOOD1 , AND ALI. H. SHAREEF1
1 Department of Computing, Universiti Pendidikan Sultan Idris, Tanjong Malim 35900, Malaysia
2 College of Engineering, University of Information Technology and Communications, Baghdad 10013, Iraq
3 Department of Computer Technology Engineering, Dijlah University, Baghdad 10022, Iraq
4 Faculty of Business, Middle East University, Amman 11133, Jordan
5 Faculty of Computer Science and Information Technology, Universiti Tun Hussein Onn Malaysia, Parit Raja 86400, Malaysia

Corresponding author: A. A. Zaidan (aws.alaa@gmail.com)

This work was supported by the Universiti Pendidikan Sultan Idris, Malaysian Ministry of Higher Education, through the Fundamental
Research Grant Scheme, under Grant FRGS/2016-0066-109-02.

ABSTRACT The emerging technology breakthrough of the Internet of Things (IoT) is expected to offer
promising solutions for indoor/outdoor healthcare, which may contribute significantly to human health and
well-being. In this paper, we investigated the technologies of healthcare service applications in telemedicine
architecture. We aimed to resolve a series of healthcare problems on the frequent failures in telemedicine
architecture through IoT solutions, particularly the failures of wearable body sensors (Tier 1) and a medical
center server (Tier 3). For improved generalisability, we demonstrated an effective research approach,
the fault-tolerant framework on mHealth or the so-called FTF-mHealth-IoT in the context of IoT, to resolve
essential problems in current investigations on healthcare services. First, we propose a risk local triage
algorithm known as the risk-level localization triage (RLLT), which can exclude the control process of
patient triage from the medical center by using mHealth and can warn about failures related to wearable
sensors. RLLT performs this initial step towards detecting a patient’s emergency case and then identifying
the healthcare service package of the risk-level. Second, according to the risk-level package, our framework
can aid decision makers in hospital selection through multi-criteria decision making (MCDM). Accordingly,
mHealth can connect directly with the servers of distributed hospitals to ascertain available healthcare
services for the risk-level package in those hospitals. The time of arrival of the patient at the hospital (TAH)
is considered for each hospital to reach a final decision and select the appropriate institution in case of
medical center failure. This paper used two datasets. The first dataset involved 572 patients with chronic
heart disease. Their triage levels were evaluated using our RLLT algorithm. The second dataset included
hospital healthcare services with two levels of availability within distributed hospitals to show variety
when testing the proposed framework. The former dataset is an actual dataset of services collected from
12 hospitals located in the capital Baghdad, which represents the maximum level of availability. The latter
is an assumption dataset of the services within the 12 hospitals located in the capital Kuala Lumpur, which
represents the minimum level of availability. Subsequently, the hospitals were prioritized using a unique
MCDM method for estimating small power consumption, namely, the analytic hierarchy process (AHP),
based on a crossover between the ‘‘healthcare services package/TAH’’ of each hospital and the ‘‘hos-
pital list’’. The results showed that the AHP is effective for solving hospital selection problems within
mHealth. The implications of this study support the patients, organizations, and medical staff in a modern
lifestyle.

INDEX TERMS Telemedicine, Internet of thing, healthcare, triage, hospital selection, medical centre
failure.

2169-3536
2019 IEEE. Translations and content mining are permitted for academic research only.
50052 Personal use is also permitted, but republication/redistribution requires IEEE permission. VOLUME 7, 2019
O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
I. INTRODUCTION
This decade has witnessed the rapid growth of techno-
logical developments in many research fields. The Inter-
net of Things (IoT) is the most recent development that
has already started to substantially impact every aspect
of our lives, including sports, healthcare, industry and the
environment [1]. IoT has a promising future in healthcare
for managing patients remotely and seamlessly under health-
care services and medical care. Remote healthcare service
is the most crucial area in relation to patients’ lives. Stud-
ies on the invention of intensive methods are continually
conducted towards the evolution of novel applications in
telemedicine and medicinal sciences [2]. IoT plays a key
part in making telemedicine a major centre of interest in the
research domain [3], [4]. The Interaction, Things, Process
and Data (ITPD) parameters, are often regarded as the funda-
mental aspects of IoT sectors. ITPD was presented by Ray [5]
in 2014. Interaction denotes how a user becomes attached
to a system (e.g. network technologies 3G/4G, Wi-Fi, blue-
tooth low energy [BTLE] and NFC). Meanwhile, an example
of Things are wearable devices that act for several objects
that collect Data from hardware, cloud-processed data and
raw data related to sensors from environments through the
intervention of sensors. Process involves the operations of
hardware platforms and the cloud and is the core of all
the parameters in effectively managing and engaging with
patients and IoT-enabled technology.
Telemedicine is the medical application of information
technology for consultations with patients outside hospitals
by using digital imaging systems or video conferencing.
Telemedicine has a three-tier architecture: Tier 1 represents
the medical body sensor; Tier 2 is mobile health (mHealth)
(both Tier 1 and Tier 2 denote the client side); and Tier 3
represents the medical centre server side, which connects
and manages the distributed hospitals’ servers [6]–[13].
Recently, IoT showed potential in generating exceptional
changes to distributed hospitals and medical centres, espe-
cially in managing day-to-day clinical operation efficiency
and remotely monitoring the health conditions of hospitalised
patients [2]. Researchers are currently trying to expand such
solutions around IoT technologies and seeking superior con-
trol over operational processes and gain efficiencies to reduce
the time for providing superior healthcare. Safety, reliabil-
ity and cost efficiency are the desired goals attained with
IoT apps in remote health sectors [2], [14]–[19]. By 2020,
the global IoT-enabled healthcare service market is expected
to increase from 32.4 billion USD in 2015 to 163.2 billion
USD, at a compound annual growth rate of 38.1% through
this forecast period [2].
In addition, the burden of cardiovascular disease is grow-
ing worldwide and is projected to emerge as the number
one cause of death worldwide by 2020 [20], [21]. Triag-
ing patients to detect their emergency levels is performed
The associate editor coordinating the review of this manuscript and
approving it for publication was Lin Wang.
VOLUME 7, 2019
after evaluating their vital signs [22], [23]. Triaging is
required to link with compatible healthcare service pack-
ages and complete the processing of healthcare service pro-
visions [22], [150]. Five healthcare service packages can
be provided remotely to patients with chronic heart disease
(CHD): Packages 1, 2, 3, 4 and 5 correspond to the emer-
gency levels of risk, urgent, sick, cold state and normal [24].
In the current research, only the first package (package
1) are adopted according to our previous work [24] which
considered the most important package of services in the
patients’ live [22], [24], [150]. The benefits of telemedicine
include a vast bibliography, but practical challenges remain in
organising risk management in the context of the continuous
improvement of remote healthcare services [25]. Several
challenges addressed in telemedicine related to healthcare
services cause failures in telemedicine architecture; these
failures can significantly affect inpatient life and lead to
link outage, potentially leading to severe consequences, as
discussed below.
Medical centre server (Tier 3) failures are caused by
(i) scalability challenges when patients increase, which
occurs in different aspects, namely disasters and mass casu-
alty incidents (MCIs), aging population (corresponding to a
rise in demand for healthcare services, followed by visiting
doctors online) and network congestion [22], [26], [27]
that causes either medical centre failure or network failure
between Tiers 2 and 3. (ii) Server failure challenges also
constitute a complex issue because of the many possible
configurations of the client–server environment and the fail-
ure modes of client, server and network devices, where the
availability of such structure is a complex matter [28]–[32].
Wearable body sensor (Tier 1) failures occur when sensor
characteristics indicate partial or complete failure, which
can degrade the performance or even destroy the stability
of the overall telemedicine systems [33]. Network con-
gestion caused by network failure between Tiers 1 and 2
will also cause a shortage in data transmission in the client
side [22]. In this case, measuring the patient’s condition
is either inaccurate or was not carried out in the first
place.
Additionally, developing countries usually continue to suf-
fer from a wide shortage of physicians and hospitals. Thus,
patients in these countries endure the physical and mone-
tary burdens of traveling around the country to see physi-
cians. However, these countries may not easily agree to
increasing the number of hospitals because of their eco-
nomic conditions [34]. Hence, instead of adding a few new
hospitals, they would rather deploy as many telemedicine
facilities-which generally cost much less than hospitals-as
they can [34].
The motivation for this research is to continue provid-
ing healthcare services within mHealth for remote patients
with CHD, especially for patients with risk-level emergency
case, and recover from the aforementioned failures. Provid-
ing health consultancy indoor/outdoor can support patients
with CHD through a distinctive quality of care in a modern
50053
 O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
FIGURE 1. High-level abstract of telemedicine architecture during various failures [24].
lifestyle and maintain their independence in a normal living
environment.
The fault-tolerant concept is defined as ‘the property that
enables a system to continue operating properly in the event
of the failure (or one or more faults within) some of its
components’ [35]. Fault-tolerance in a distributed system is
the ability to isolate and recover from failures, self-heal and
have no single point of failure [36], [37]. This property can
be implemented in different ways [37]. In this context, a
fault-tolerant framework in telemedicine architecture should
be proposed to recover telemedicine system parts from the
addressed failures. On the basis of our previous work in [24],
Fig. 1 shows the high-level abstract of telemedicine architec-
ture failures related to Tiers 1 and 3. The figure also illustrates
the direct provision of services from distributed hospitals’
servers to mHealth without the medical centre server.
mHealth represents the attractive parts in telemedicine
architecture and offers the potential for combining sensor
networks and information to improve patient care and pro-
vide healthcare services. The definition from [38] states,
‘M-health is the application of mobile computing, wire-
less communications and network technologies to deliver or
enhance diverse healthcare services and functions in which
the patient has the freedom to be mobile, perhaps within a lim-
ited area’. mHealth is an important link between Tiers 1 and 3,
and it focuses on the mobility of patients in relation to the
healthcare system. This paper presents a fault-tolerant frame-
work on mHealth (FTF-mHealth-IoT) within telemedicine
architecture in the context of the four fundamental sectors
of IoT parameters (ITPD). Fig. 2 illustrates the interaction
50054
between IoT-enabled technologies and the multi-layer archi-
tecture of FTF-mHealth-IoT. The core of these parameters is
to efficiently manage and correlate patients and IoT-enabled
technologies as follows:
Interaction involves the acquaintance of patients with sen-
sors in and around them (e.g. through network technologies
2G, 3G, 4G, GPRS, Wi-Fi, NFC and BTLE). It can represent
the GPS inside mHealth to interact with distant hospitals in
calculating the TAH.
Things are wearable CHD sensors (ECG, SPO2 and BP)
that can be attached to the patient’s body indoor/outdoor and
connects to the network and sharing information in WBAN.
Process concerns the technological processes executed
firstly by the RLLT algorithm to automate data gathering
from CHD sensors and text to improve triage accuracy. It also
encompasses the accumulation, communication and analysis
of healthcare service data and TAH gathered from distributed
hospitals as processed by our decision-making technique
(AHP) in real time to select the best hospital.
Data can be collected from sensors and can include health-
care service data and TAH collected from distributed hospi-
tals. mHealth can stream data online in a bidirectional way
with the distributed hospitals’ servers through the Internet.
The remaining parts of this article are as follows. Section 2
discusses the Literature review. Section 3 reports the Method-
ology of the FTF-mHealth-IoT framework. Section 4 presents
the Results and discussion. Section 5 highlights the Lim-
itations. Section 6 focuses on the Recommendations for
future work. Finally, Section 7 presents the Conclusion of the
research.
VOLUME 7, 2019
O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors

FIGURE 2. Multi-layer architectural framework of FTF-mHealth-IoT.

II. LITERATURE REVIEW confidentiality and fine-grained access to outsourced medical

This section presents a detailed literature review related to the
categories Tiers 1, 2 and 3.
A. TIER 1
Sensors play an ever more important role in medical tech-
nologies, with the aim of simplifying their operation and
making medical devices safe and effective. According to [39],
‘The common problems of control congestion in many data
networks, such as WSNs, result in packet loss, increasing
end-to-end delay and excessive energy consumption due to
retransmission’. In [40], a protocol for congestion detection
was proposed by adopting multi-biosensors based on the
type-2-fuzzy logic system. In [41], the Healthcare Aware
Optimized Congestion Avoidance and Control Protocol was
introduced for a medical health app for active queue manage-
ment, and multi-path QoS aware routing was adopted. In [39],
an optimised congestion management protocol was proposed
for a wireless sensor network in two stages: 1) by avoiding
congestion through an AQM scheme and providing QoS and
2) by using three mechanisms for congestion control.
B. TIER 3
This category includes two areas: (1) security and privacy and
(2) remote monitoring. Under security and privacy, the study
of [37] proposed a mechanism that focused on integrity,
data generated by medical sensor networks. The focus on [42]
is the security of the ehealth society through the incorporation
of a low-cost and secure communication system. In [43],
a remote monitoring system that includes nano networks
inside the body of the patient was presented, and the current
trends, security challenges and requirements in the WBAN
were investigated.
Remote monitoring includes studies on E-triage servers
and the provision of services. The E-triage server refers
to the triage process located in the onsite rescue control
centre for managing and monitoring patients’ vital signals.
In [44], a monitoring system that uses a photoplethysmo-
graph in emergency rooms, which allows healthcare profes-
sionals/physicians to collect the pulse rate and temperature of
patients in a comfortable and constant manner, was proposed.
In [45], a wireless pulse oximeter prototype was developed
to measure data gathered from patients. Clinical analysis
by a central unit assists and coordinates first-aid teams and
updates information on the clinical status and locations of
patients.
Service provision is an attractive and important part of
telemedicine because it facilitates the treatment process of
patients. The investigations of [46]–[51] focused on support-
ing an alert emergency service generated when patients’ vital
signals change to abnormal levels to notify the emergency
medical teams and caregivers. In [52]–[55], several services

VOLUME 7, 2019 50055

 O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
to patients, such as recommendations, tips, drug prescriptions
and telehealth consultation, were presented. The study of [56]
presented a system to promote emergency treatments and
confirmed the necessity of a cloud computing system for
emergency rural health to mitigate deaths from time delays
during patient transportation and the shortage of appropriate
and timely first-aid. Furthermore, [57] provided healthcare
services on-the-fly through vehicles in the case of travel-
ling patients. First-aid operations and services that can be
provided while the patient is in an ambulance were devel-
oped in [58]. Finally, [59] proposed a novel telehealth elderly
healthcare service that connects remote physical therapists to
seniors at home by providing verbal, auditory and visual cues
to support correct exercise movements.
C. TIER2 (mHealth)
The category of mHealth can be divided into three areas as
follows.
1) TREATMENT SUPPORT AND DISEASE SURVEILLANCE
The main role of both aspects is to ‘observe, predict and
minimise the harm caused by outbreaks, epidemics, and pan-
demics and increase knowledge about the factors that con-
tribute to such circumstances’ [24], [60]. Studies in this topic
support patients in monitoring and managing their body tem-
perature, blood pressure, chronic diseases and heart rate to
obtain services from Tier 3 (server side) as a response to using
mHealth. A mobile machine learning model for monitoring
cardiovascular diseases (M4CVD) from a clinical database
was presented in [61]. A monitoring system based on the new
Arduino mega micro-system device that uses blood pressure,
body temperature and heart rate by implementing algorithms
to analyse real-time signals, fuse multi-sensor data and trans-
mit these signals through an Xbee module was presented
in [62]. MobiPatterns, a mobile monitoring app that contin-
uously monitors patients with diabetes via smartphones and
a biometric device, was put forward in [63]. In [64], a diabetic
mobile app was introduced which provides patients with the
proper services through low-level invasive impact technolo-
gies using new process models meant to integrate the software
components and share information. In [64], a ‘Jeev software
app’ for tracking the vaccination coverage of children in rural
communities was highlighted. The study of [65] presented a
Mo-Buzz system using social media for preventing dengue in
Sri Lanka and Southeast Asian regions. A treatment mech-
anism that provides personalised daily interactive sessions
for patients with major depression according to the patient’s
history data, clinical requirement and current responses was
presented in [66]. A coaching approach known as a personal
coaching system was developed in [67] using body sensors
that are integrated with smart reasoning and context-aware
feedback to support patients’ healthy behaviour. In [6], a gen-
eral approach was introduced to assist in managing patients
with an acute coronary syndrome. This approach was built via
a data-driven platform for an urgent decision support system
for ambulance and emergency medical services.
50056
2) TRIAGE OVER mHealth
The term ‘triage’ comes from the French word ‘trier’, which
means ‘to sort’. The concept was used in warfare sys-
tems to prioritise all casualties and give urgent care to the
most critically injured. The triage process generates accu-
rate information on patients’ health conditions, such as risk,
urgent, sick, cold state and normal [7], [22]. According to
the triage environment, the studies can be classified into two
directions: triage-based disaster casualties and incidents and
triage-based home monitoring.
a: TRIAGE-BASED DISASTER CASUALTIES AND INCIDENTS
In MCIs, ‘the first triage of all injured individuals is essential
in the processes of the medical team. Due to the challenges
in the accident area, the medical team should implement
correct and advisable steps when encountering casualties in
the field, and this usually causes inefficient treatments of
casualties or misleading information forwarded to the exec-
utive emergency physician (EEP)’ [68], [69]. The research
of [69] presented a project (AUDIME) based on a hands-
free approach for mobile or wearable devices and capable
of evaluating the social acceptance and usability of wearable
devices in the context of MCI management. The study of [70]
proposed a platform for body-worn vital signal monitoring of
ECG, SpO2, body temperature and multichannel auscultation
for developing a field accident and emergency centre intel-
ligent monitoring system to support the allocation of medi-
cal resources in a disaster environment. In [71], the authors
utilized machine learning techniques to develop a real-time
system that can help patients using mHealth E-triage accom-
plished via crowd-sourced and sensor detector information.
A diorama-based system was presented in [72] to enable
awareness searching and rescue operations in urban environ-
ments in indoor/outdoor settings and for triaging patients by
RFID to mark the patients’ locations and points of interest.
A ripple project presented in [73] involved the creation of a
medical body area network of sensors for the triaging process
in disaster scenarios.
b: TRIAGE-BASED HOME MONITORING
The study of [22] presented a multi-source healthcare archi-
tecture framework for telemonitoring systems amongst inte-
grated multiple sources and text by an adaptive data fusion
method. The study improves healthcare scalability efficiency
by triaging remotely in Tiers 1 and 2 and enhancing priori-
tisation for patients with CHD in Tier 3. The result of the
triage process includes five emergency levels: risk, urgent,
sick, cold state and normal.
3) NETWORK FAILURE BETWEEN TIERS 2 AND 3
The studies in this section focused on tracking and interacting
with patients through mHealth when communication links are
lost between Tiers 2 and the medical server side. The work
in [74] introduced a tracking prototype of WPAN technology
for pilgrims when disasters occur and enables communication
VOLUME 7, 2019
O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
even in tower-less areas by adopting ZigBee. The system
incorporates two modules: surveillance-side and person-side
control. The investigation in [75] considered an mHealth
system used in tsunami-stricken disaster scenario recovery
in environments without functional telecommunication ser-
vices. In this system, vital signals are transmitted over D2D
and LTE-direct technologies by ad-hoc networks to restore
the lost communication links. In [76], the authors presented
power management protocols and a framework that models
the complex decision logic involved in leveraging mobile
ad-hoc networks for diverse patient monitoring scenarios.
Finally, [77] presented a temporary ad-hoc network for the
technical feasibility of medical alarm dissemination in urban
environments by using mobile devices to the nearest hospital
or by medical field personnel when failure or congestion
occurs within infrastructure-based communication networks.
D. GAP ANALYSIS FOR TELEMEDICINE APPLICATIONS
In Tier 2 (mHealth), all studies presented various health-
care systems and applications that are completely controlled
from the server side. Only a few investigations introduced
restricted solutions in the case of interruption or breakdown
of communications between Tiers 2 and 3 by using an ad-hoc
network; however, efficient QoS provision for the mobile
ad-hoc network represents a challenging task, particularly for
different types of traffic [78]. Thus, modern healthcare sys-
tems that use ad-hoc networks have several critical require-
ments and challenges, such as reliability and timely access
to diagnostic information without failure, compared with the
traditional wireless network [79].
In conclusion, none of the studies in the literature consid-
ered a fault-tolerant framework when the mentioned failures
occurred in the telemedicine architecture. Studies that inves-
tigated the delivery of healthcare services by mHealth still
experienced challenges when network failure [22] or medical
centre server failures occur at Tier 3 [7], [80]. Delivering
healthcare services in accordance with the abovementioned
problems remains unaddressed [7], [61], [80]. Moreover,
a new local triage algorithm for risk emergency case by
mHealth should be proposed to exclude the control process
of triage from Tier 3 when failures happen at this side and to
stimulate (alarm) the patient when failures happen at Tier 1.
Failures at Tier 3 or even in its network mean that mHealth
should be connected directly with distributed hospitals to
select the best one. Thus, an understanding of the exact
hospital selection criteria and their weights is important [81],
[82]. Issues about hospital selection are described in the next
section.
E. OPEN ISSUES AND CHALLENGE
FOR HOSPITAL SELECTION
The descriptions of the specific problems of the general prob-
lem, namely hospital selection, within the FTF-mHealth-IoT
framework in terms of issues are as follows:
First issue: As a study case, the first attribute, which is
healthcare service package for CHD which is package 1 (as
VOLUME 7, 2019
mentioned in the Literature Review Section), can cater to
patients from hospitals through several services according
to the triage risk-level [7], [22]. The second attribute is
TAH, which is an important factor for selecting the appro-
priate hospital spatially with the crucial conditions of CHD
patients [83]–[85]. Therefore, hospital selection regarding
multi-attribute facets (healthcare services package and TAH)
with respect to the proper weight assigned for each attribute
is considered a multi-attribute decision matrix [6], [86].
Second issue: Different weights are often given by decision
makers (doctors) to the mentioned attributes, thereby further
increasing the complexity of the task [87], [88].
Third issue: Whenever the service availability within hos-
pitals is at a high level and the TAH takes a little period of
time, this situation has a significant impact in the selection of
the best hospital [89]–[91]. Essentially, this challenge comes
from conflicting terms, specifically, the conflict amongst
attributes and amongst data. Thus, this inverse relationship
occurs between both attributes causing a trade-off.
Fourth issue: The TAH and the availability of services
vary from one hospital to another [83], [87], [92]. Therefore,
the selection process involves simultaneous consideration
from multiple attributes of distributed hospitals in different
situations, which generate data variation. The changing of
TAH happens when the patient moves. Hence, the hospitals’
data represented by services and TAH are frequently changed
and cause a data variation problem.
In conclusion, the selection process of hospitals within
mHealth is a complex multi-attribute decision-making prob-
lem [93], [94], in which hospitals are considered available
alternatives for the decision makers. The configuration of the
problem statement is illustrated in Fig. 3.
Accordingly, the multiple decision-making method
(MCDM) must be used as a recommended solution to solve
this complex situation. MCDM is defined by Keeney and
Raiffa [95] as ‘an extension of decision theory that covers
any decision with multiple objectives’. ‘A methodology for
assessing alternatives on an individual, often conflicting cri-
teria, and combining them into one overall appraisal.’ Stewart
and Belton [96] defined MCDM as ‘an umbrella term to
describe a collection of formal approaches, which seek to take
explicit account of multiple criteria in helping individuals or
groups explore decisions that matter’.
In any rank of the MCDM, a definition of funda-
mental terms is required and must contain the DM and
its attribute [97]–[102]. An evaluation matrix includes n
attributes and m alternatives, which must be identified. The
intersection of an attribute and alternative is defined as z_ij.
Hence, we have a matrix (z_ij)_ (m∗n)’ explained as follows:
 X1 X2 . . . Xn 
Y1 Z11 Z12 . . . Z1 n
DM = Y2
 Z21 Z22 . . . Z2 n 
,
..  ..

.. .. .. 
.  . . . . 
Ym Zm 1 Zm 2 . . . Zmn
50057
 O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
FIGURE 3. Problem statement configuration.
where Y1 , Y2 , . . . ,Ym are suitable alternatives that decision
makers must rank. X1 , X, . . . ,Xn are the attributes/criteria
against which the performance of all alternatives are evalu-
ated. zij is the rating of alternative Yi with respect to crite-
rion Xj , and Wj is the weight of criterion Xj . For example,
DM is assumed to be the DM utilised to score and rank the
alternatives Yi according toXj . Table 1 illustrates an example
of the multiple attribute problems expressed in [103]–[106],
[152]. Note that the data in the flowchart cannot be eas-
ily evaluated due to the large numbers of X2 and X3
(Fig. 4).
Various MCDM theories have been discovered. The advan-
tages and limitations of these MCDM methods are organised
in Table 2 [87], [107]–[119].
50058
MCDM methods have been applied to different health-
care applications. However, none of these methods has been
used in mHealth to rank hospitals as a fault-tolerant sys-
tem [24]. In the present research, the AHP method is rec-
ommended for use within mHealth reviewing the relevant
literature was reviewed. Saaty developed AHP, an effective
analytic tool widely used in several academic and practical
fields. Moreover, Saaty introduced AHP as three levels: the
objective (goal) of the model, the criteria and the alternatives.
The aim of AHP is to identify the relative superiority of
all alternatives and prioritise them. Accordingly, the relative
importance (weights) modelled the criteria (of healthcare ser-
vices package and TAH) that are estimated using numerical
data and are applied to evaluate each alternative (hospitals).
VOLUME 7, 2019
O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
TABLE 1. Example of multi-attribute problems.
TABLE 2. Advantages and limitations of the MCDM methods.
The crucial characteristic related to AHP is the use of pair-
wise comparison based on experts’ judgments, which are
used to compare the alternatives concerning the different
criteria and estimate the criteria weights [120]. AHP is widely
adopted for the ranking matter on medical scatter. It is suitable
for cases with attributes and alternatives and rapidly identifies
VOLUME 7, 2019
the most suitable alternative to mHealth. Furthermore, AHP
is capable of mitigating the complexity of decision making
in a reliable way. Thus, the current trend with respect to
the MCDM uses the AHP method to compensate for the
strength point in a single method [3], [121]–[125], [151]. The
two stages of AHP for fully ranking hospitals must be used.
50059
 O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
FIGURE 4. Graphical representation of the multi-attribute problems
presented in Table 1.
Both stages will be explained in detail in the Methodology
section.
III. METHODOLOGY
The presented Methodology section are derived from the two
phases discussed in our previous study [24] and demonstrated
in Fig. 5.
A. IDENTIFICATION PHASE
The following subsections discuss the six stages presented in
this phase to propose the DM.
1) IDENTIFY THE TARGETED TIER WITHIN THE
TELEMEDICINE ARCHITECTURE
Tier 2 is the targeted tier. The following requirements for
the proposed framework should be achieved to overcome the
mentioned issues and the general problem.
◦ Detect the triage risk-level to identify healthcare service
package of this level for patients with CHD.
◦ Alert the patient in case of failure at Tier 1.
◦ Provide a specific weight to healthcare services package
and TAH for experts to evaluate hospitals.
◦ Possess the capability to rank and order each hospital
based on their available services and TAH and prioritise
them in a queue.
◦ Build two datasets. The first dataset is for the patients
with CHD to feed the triage process. The second dataset
is for the hospitals’ healthcare service dataset to feed the
decision matrixes for hospital selection.
2) IDENTIFICATION OF PATIENTS WITH
CHD AND THE DATASET
Patients were identified as remote home monitoring patients
with CHD. Three sensors were utilised to measure and
monitor the patients’ vital signals, consisting of ECG,
SpO2 and blood pressure, and the texts additionally used.
Text data (complaints) were used in this research as a medical
source to ask patients four questions that can be answered by
‘yes’ or ‘no’. A dataset comprising 572 patients was involved
50060
and adopted in this research from [7], [22]. For all datasets
used, males constituted 60% and females the remaining 40%.
Furthermore, 50% of the patients were 40–65 years old, 40%
were over 65 years old and 10% of the patients were under
40 years old.
3) PROPOSAL OF A NEW ALGORITHM FOR RLLT WITHIN
mhealth FOR TRIAGE AND DETECTION OF TIER 1 FAILURES
A new triage method known as RLLT for the telemedicine
architecture is proposed and derived from our previous
study [24]. The requirement for constructing the RLLT algo-
rithm, general scheme, workflow, data fusion module and
considered theory for data processing are derived from [24].
The mathematical representation of RLLT was also estab-
lished using evidence theory and if–then statements accord-
ing to the mathematical representation of Dempster–Shafer
theory. The outputs of this module include four types of
decisions: triage code (TC) value, triage level, healthcare
services package 1 and alarm (Table 3).
4) IDENTIFICATION OF DISTRIBUTED HOSPITALS
This study adopts 12 hospitals as ‘a proof of concept’ to
represent the alternatives in the DM. To test the framework
in several directions, 12 hospitals located in the capital Bagh-
dad (Al-Karkh) are first used to test the first scenario, and
another 12 hospitals located in the capital Kuala Lumpur
are employed to test two other scenarios. The locations of
hospitals within each city are specified on the map.
5) IDENTIFICATION OF HEALTHCARE SERVICE
PACKAGE DATASET AND TAH
To represent the attributes (or criteria) in the DM, the first
attribute, namely healthcare service package, is explained in
Table 3. TAH represents the second attribute.
Healthcare services in distributed hospitals vary in several
aspects, with some services available in some hospitals but
unavailable in others. This situation is a natural scenario
in hospital work when the demand for healthcare services
increases. According to [126], ‘HeRAMS (Health Resources
& services Availability Mapping System) is a standardized
approach supported by a software-based platform that aims
at strengthening the collection, collation, and analysis of
information on the availability of health resources and ser-
vices in humanitarian emergencies. The key information that
HeRAMS is assessing includes the availability of the health
services, accessibility, functionality status, health infrastruc-
ture and human resources at PHC centers and secondary care
level. The current assessment considered the public hospitals
(MOH general and autonomous hospitals) and public Pri-
mary health care PHC centers’. Therefore, health services
in hospitals vary in terms of availability at a maximum or
minimum level. In this context, service availability in both
maximum and minimum levels should be identified to show
the variation between the 12 hospitals. Therefore, a dataset
services is collected from 12 hospitals located in the capital
Baghdad (Al-Karkh) to represent the maximum capacity of
VOLUME 7, 2019
O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
FIGURE 5. Research methodology phases.
TABLE 3. Healthcare service package 1, triaging levels and TC value.
services (fully available). Conversely, the minimum level of
service availability of package is represented by the assump-
tion dataset within 12 hospitals located in the capital Kuala
Lumpur.
To identify the TAH, the motion detector app uses a built-in
GPS to measure the movements of the patient in a mobile
environment via social network services. Online social net-
work services can easily obtain the patient’s location and
TAH owing to the increasing popularity of GPS-enabled
mobile devices [127]. The justification of TAH as the sec-
ond criterion in the DM can be specified by using three
assumptions on the map. The first assumption is Location
VOLUME 7, 2019
A in the 12 hospitals located in Al-Karkh in Baghdad. The
second and third assumptions are Locations B and C in the
12 hospitals located in Kuala Lumpur. Three scenarios are
identified through the two hospital datasets, and the three
patient locations are shown in detail in the Results and Dis-
cussion section. The proposed DM is illustrated in the next
section.
6) PROPOSED DM IN mHealth
The proposed DM is based on the crossover of (1) healthcare
services package /TAH and (2) hospital lists according to the
availability of multiple services in each hospital and the TAH.
50061
 O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
TABLE 4. DM for hospitals.
FIGURE 6. Framework of identifying DM in the FTF-mHealth-IoT architecture.
a: DESCRIPTION OF DM FOR PACKAGE
A decision matrix is proposed to rank the 12 hospitals. The
DM will evaluate each hospital according to service availabil-
ity and TAH, as explained in Table 4.
b: DM EVALUATION
The services and TAH exert different effects on the hospi-
tal evaluation. Fig. 6 shows the framework for identifying
the DM. The process of DM evaluation is as follows. The
RLLT algorithm is used to measure the triage level based
on the patient’s vital signs received from Tier 1. The output
of this algorithm is represented as a TC value to identify
the risk-level triage, which is linked to healthcare services
package. The FTF-mHealth-IoT framework can then iden-
tify the DM of the package, and requests are sent to hos-
pitals to obtain data representing services based on the IoT
parameters (ITPD). If the services are available, they will be
evaluated as 1; otherwise, they are evaluated as 0. TAH will
50062
also be evaluated according to four groups of time measures:
the shortest time (SstT) is 0–15 minutes, the short time (ST)
is 15–30 minutes, the long time (LT) is 30–45 minutes and
the longest time (LstT) is 45–200 minutes.
B. DEVELOPMENT PHASE
Ranking hospitals is a multi-attribute decision-making prob-
lem; thus, weights should be set based on the objectives
of experts. In this situation, six evaluators should set the
preference weights for the two attributes in the DM.
1) DECISION-MAKING SOLUTION DEVELOPMENT IN
mHealth FOR RANKING HOSPITALS BASED ON AHP
First, the MLAHP calculates the weights of the main criteria
(healthcare service package and TAH) and sub-criteria (mul-
tiple services). Thereafter, the AHP is used to determine the
weights of the grade of criteria importance (GCI) for (1) each
service criteria (available and unavailable) and (2) TAH cri-
teria, namely SstT, ST, LT and LstT. Then, the AHP is used
VOLUME 7, 2019
O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors

FIGURE 7. MLAHP-AHP method for ranking hospitals.

to score and rank the hospitals on the basis of the quanti- of package used in the MLAHP pairwise comparison is
tative information through which the criteria are measured. demonstrated in Fig. 8. The first layer has two main criteria,
Therefore, hospitals with the SstT to the patient with the namely healthcare service package and TAH. The second
maximum level of available services should have high pri- layer in packages 1 has six multi-service sub-criteria. A pair-
ority, whereas those with the LstT to the patient with the wise comparison is performed between the main criteria with
minimum level of available services should have low priority. respect to the main goal to obtain the weights. Moreover,
The MLAHP-AHP method structure is presented in Fig. 7. the sub-criteria of the same parent are compared with the
criteria of their parent.
a: MLAHP FOR SETTING WEIGHTS FOR THE
MAIN CRITERIA AND SUB-CRITERIA
ii) CONSTRUCT THE PAIRWISE COMPARISONS
Several steps are implemented to assign the proper The MLAHP builds a pairwise matrix comparison to establish
weights to the main criteria and sub-criteria using the a decision, as follows:
MLAHP. The MLAHP procedure includes the following
steps [114], [128]–[134].
... ...
 
x11 x12 x1n
i) DECOMPOSE A DECISION PROBLEM
... ...
(
 x21 x22 x2n  xii = 1
INTO A DECISION HIERARCHY A= . .. .. ..  where,
..
 
Problem modeling needs to be designed as a hierarchy con-  .. . . . .  xji = x1ij .
sisting of the decisive goal, the main criteria and the sub- xn 1 xn 2 ... ... xnn
criteria. The hierarchy of the main criteria and sub-criteria (3.1)

VOLUME 7, 2019 50063

 O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
FIGURE 8. Hierarchy of MLAHP for the main and sub-critena.
TABLE 5. Nine scales of pairwise comparisons [107], [108].
FIGURE 9. Sample evaluation form.
Elements Xji are obtained from Fig. 8. ‘The comparisons
(relative importance) of each criterion in the first layer or
sub-criteria in the second layer are measured according to
a numerical scale from 1 to 9’ [135]–[139]. These relative
scales (from 1 to 9), as presented in Table 5, are used to
show the expert judgments for all the comparisons. Each
expert should critically set these judgments based on their
experience and knowledge.
iii) OBTAIN PRIORITY JUDGMENT RANKING SCORES
A pairwise comparison questionnaire is designed and dis-
tributed to a geographically diverse convenience sample
of experts, namely cardiologists with expertise in CHD,
to show the relative importance of the main criteria and
50064
the sub-criteria. The experts are asked to provide their judg-
ments on the main criteria and sub-criteria by using scales
(1 to 9) for comparison. They are also asked to rank the rela-
tive importance of package. A sample of the criteria pairwise
comparison is presented in Fig. 9.
The experts reveal their judgments for each sub-criterion
with respect to the healthcare service package (parent).
In other words, the multiple services in the second layer are
compared with their parents in the first layer by following
the same pattern for the criterion. The number of required
pairwise comparisons is n× (n− 1)/2 (15 comparisons),
where n is the number of criteria used during the evaluation.
At this stage, the MLAHP extracts the weight of importance
of the healthcare service package and TAH and the related
VOLUME 7, 2019
O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors

FIGURE 10. Design of the MLAHP measurement steps for the weight preferences for healthcare Services package.

multiple services. The MLAHP is technically valid and does

not require a large sample size [109], [140]–[142]. Hence, six
experts with over 10 years of experience are selected in this
study. All comparisons for the healthcare and related services
of the package are made at this point.
FIGURE 11. Sample hierarchy for one criterion of services.
iv) BUILD THE NORMALISED DM
Every element of matrix A is normalised by dividing each
element in the column by the sum of the elements in the
same column to create the normalised pairwise comparison v) CALCULATE ALL PRIORITY VALUES (EIGENVECTOR)
matrix A_norm. A_norm is the normalised matrix of A(1), In this step, the MLAHP pairwise comparison uses mathe-
where A(x_ij) is given by Equation (3.2). A_norm (aij) is matical calculations to convert the judgments for the weights
expressed as of the main criteria and sub-criteria. A reciprocal matrix is
xij created after obtaining the responses of the pairwise compar-
aij = Pn , (3.2) isons. The MLAHP pairwise derives the local priorities for
i=1 xij
each group at each level, which represents the importance
a11 a12 . . . · · · a1n
 
 a21 a22 . . . . . . a2n  of each service in each package with respect to the parent.
Anorm =  .

.. .. . . .. 

(3.3) Then, the global priority for each service, which represents
 .. . . . .  the importance of each service with respect to the goal,
an 1 an 2 . . . . . . ann is obtained. The weights of the decision factor i can be

VOLUME 7, 2019 50065

 O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors

FIGURE 12. Hierarchy for the TAH criteria.

FIGURE 13. Sample evaluation form for GCI for Services.

FIGURE 14. Sample evaluation form for the GCI criteria of TAH.

calculated as Equation by Equation (3.7)

Pm n
j=1 aij CI
X
Wi = and Wi = 1 (3.4) CR = (3.7)
n RI
j=1
The CR is the ratio of the CI to the RI. The CR, which
where n is the number of compared elements. The MLAHP was proposed by [116] and [117], is a quantitative measure
measurement steps should be designed to obtain the weights for the degree of inconsistency of a pairwise comparison
based on the preference of the evaluator. Fig. 10 presents the matrix. A pairwise comparison matrix with a corresponding
MLAHP measurement steps for the weight preferences that CR of not more than 10% or 0.1 is acceptable [116], [118],
are used by the six evaluators for package. [143]–[147]. If the level of inconsistency is unacceptable,
then the decision maker should revise the pairwise compar-
vi) CALCULATE THE CONSISTENCY RATIO isons; otherwise, these comparisons will be ignored.
The consistency ratio (CR), which expresses the internal con-
sistency of the judgments, is calculated. The study of [118] b: AHP FOR RANKING HOSPITALS
defined the terms to develop a quantitative measure for the
The AHP obtains the GCI weights for each service (available
degree of inconsistency within a pairwise comparison matrix.
and unavailable) and TAH criteria (SstT, ST, LT and LstT).
The consistency index (CI) is calculated by Equation (3.5).
In the same sequence, the AHP includes the same MLAHP
λ max −n steps above to calculate the weights for both procedures.
CI = (3.5) At this point, the overall weights of the main criteria and
n−1
sub-criteria are derived from the MLAHP. Thereafter, the
The random index (RI) is calculated by Equation (3.6). AHP is used to score and rank the hospitals on the basis
1.98(n − 1) of the quantitative information through which the criteria
RI = .CI (3.6) are measured. The available alternative scores are ranked
n
in descending order, and the hospitals are prioritised based
The degree of inconsistency is measured by the CI. The on their available healthcare services and TAH. Aggregate
corresponding measure for the degree of inconsistency of a scores only provide an idea of which hospitals are more
pairwise comparison matrix is the RI. The CR is calculated appropriate than others. The AHP steps in this section to

50066 VOLUME 7, 2019

O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
FIGURE 15. Structure measurements of the integrated MLAHP-CGI weights for ranking hospitals.
obtain the GCI weights for the service and TAH criteria are
as follows:
i) DECOMPOSE A DECISION PROBLEM
INTO A DECISION HIERARCHY
A sample hierarchy for one service criterion is provided
in Fig. 11, while the hierarchy for TAH is presented in Fig. 12.
A pairwise comparison is performed between each service
criterion (available and unavailable) and the TAH criteria
(SstT, ST, LT and LstT) to obtain the weights.
ii) CONSTRUCT THE PAIRWISE COMPARISONS
AHP builds a pairwise matrix comparison by using
Equation 3.1. Elements Xji in this equation are obtained from
each service hierarchy and TAH criteria. The comparison of
each criterion is measured according to the numerical scales
(1 to 9) presented in Table 5.
iii) OBTAIN PRIORITY JUDGMENT RANKING SCORES
The same experts are asked to provide their judgments for
the service criteria and TAH by using the nine scales for
comparison. A sample of the criteria pairwise comparison
in the evaluation form distributed to the experts is presented
in Figs. 13 and 14.
VOLUME 7, 2019
In other words, each service is compared with itself
with respect to availability and unavailability. At this stage,
the AHP identifies the weights of the GCI for each service
criterion and TAH, and the experts’ judgements show the
importance. The number of required individual comparisons
is n× (n− 1) /2, where n is the number of criteria used during
the evaluation. The decision-making team is set up at this
stage. The total number of GCI comparisons is 6 for the
services criteria and 6 for the TAH criteria.
iv) BUILD THE NORMALISED DM
All the elements in matrix A that represent the service and
TAH criteria are normalised by dividing each one in the
column by the sum of the elements in the same column to
create the normalised pairwise comparison matrix A_norm.
A_norm is the normalised matrix of A(1), where A(x_ij) is
specified by Equation (3.2). A_norm (aij) is expressed as
shown in Equation (3.3).
v) CALCULATE ALL PRIORITY VALUES (EIGENVECTOR)
The AHP pairwise comparison uses mathematical calcula-
tions to convert the judgments to give the weights for the
service and TAH criteria. The weights of the decision factor
i can be calculated as Equation (3.4). The AHP measurement
50067
 O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors

FIGURE 16. Design of the measurement steps for ranking hospitals.

TABLE 6. Real healthcare services package data set of 12 hospitals within Baghdad-Al Kakh.

TABLE 7. Assumption healthcare Services package data set of 12 hospitals within Kuala Lumpur.

steps should be designed to obtain the weights based on the
evaluator’s preference.
vi) CALCULATE THE CR
CI is calculated by Equation (3.5), RI is calculated by
Equation (3.6) and CR is defined in Equation (3.7).
vii) AGGREGATION
All the required weights for the MLAHP and the GCI are
calculated in this step. The result of the GCI weights for
each service gain two weight values, with the first value
50068
representing the service in the DM which is available at the
hospital and the second value representing the service in the
DM which is unavailable. Moreover, the GCI values for TAH
gains four weight values (SstT, ST, LT and LstT) to represent
one in the DM. The structure measurements of the integrated
MLAHP-CGI weights for ranking hospitals are presented
in Fig. 15. In the same context, Fig. 16 presents the DM
measurement step design and the mathematical operations to
determine the appropriate hospital. Aggregation is calculated
by the summation of the service and TAH values (V) in
each row. This value (VS_H) represents the final hospital
score.
VOLUME 7, 2019
O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors

FIGURE 17. Overview of the evaluation and ranking results of the hospitals selection process.

TABLE 8. TAH and distances of location A towards 12 hospitals within the
capital Baghdad-Al Karkh.
score after the aggregation process for each hospital. The
aggregation value is obtained by multiplying the MLAHP
and the CGI weights, as presented in Figure 16. The fig-
ure gives the full description and calculations of the hospital
rankings.

IV. RESULTS AND DISCUSSION

The results of the second and third phases are presented in
this section. The ranking process results used in this research
are presented in Fig. 17.

A. SEQUENCE RESULTS FOR DM EVALUATIONS

As shown in Fig. 15, the MLAHP weights, which were
settled in Section 3.2.1.1, are integrated with the CGI weights,
which were settled in Section 3.2.1.2, to obtain the final
Several sequences are first presented in this section to
evaluate DM with different scenarios. These scenarios are
based on the patient dataset for linking patients of risk-level
with healthcare services package, the high and low hos-
pital healthcare service level datasets from 12 hospitals

VOLUME 7, 2019 50069

 O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors

TABLE 9. TAH and distances of locations A and B towards 12 hospitals within the capital Kuala Lumpur.

TABLE 10. Ranking hospital scenarios. TABLE 11. Evaluation results of the three scenarios.

in different cities and the patient locations within these

cities.

1) PATIENT DATASET
The CHD patient dataset contains 572 patients. According to
the results of the TC values, which are calculated by our pro-
posed RLLT algorithm, there are 66 patients in the risk level.
In this context, the results of the package 1 based on patient
risk-levels are 66 package 1. Table 1 in the Appendix shows
the complete patient dataset and the results of the TC values
and healthcare services.

2) HOSPITAL HEALTHCARE SERVICE

PACKAGE DATASET
The healthcare services package in the 12 hospitals is rep-
resented by the maximum and minimum levels of service
availability as follows.

a: REAL HEALTHCARE SERVICE PACKAGE DATASET

The results of the data collection of real healthcare ser-
vices package in the 12 hospitals are shown in Table 6.
The number of services below represents the maximum
level of services in hospitals in Al-Karkh, Baghdad. For In addition, the symbols of all the services within the
privacy reasons, this research did not mention the names package are mentioned and predefined in the Methodology
of the hospitals and labeled them from 1 to 12 instead. section.

50070 VOLUME 7, 2019

O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
FIGURE 18. Location A and 12 hospitals within the capital Baghdad–Al Karkh.
TABLE 12. MLAHP results of the weight calculated for six experts.
TABLE 13. Maximum and mínimum weight of the criteria obtained from six experts.
b: ASSUMPTION FOR LOW-LEVEL HEALTHCARE
SERVICE PACKAGE DATASET
The assumption dataset shows the various hospital rankings
in terms of normal and abnormal service availability levels.
The assumption dataset is used to represent abnormal service
capacity in the evaluation process if low services occurred in
the hospitals, as shown in Table 7.
3) PATIENT LOCATION STATUS FOR DETERMINING TAH
A total of three assumptions for the patient locations in the
two cities are presented based on our hospital datasets. The
first assumption is Location A, which is in Al-Karkh in
Baghdad. This location is evaluated with the real healthcare
service dataset collected from Baghdad hospitals, as shown
in Fig. 18 and Table 8. The other two assumptions are
VOLUME 7, 2019
Locations B and C, which are in Kuala Lumpur. These loca-
tions are evaluated with the assumption healthcare service
package dataset. B and C differ from each other in terms
of distance and time of the 12 hospitals in Kuala Lumpur,
as shown in Fig. 19 and Table 9.
4) EVALUATION SCENARIOS
Table 10 illustrates three scenarios to show more than one sce-
nario and cover the general evaluation states of the hospitals
based on the previous subsections. Among the scenarios, first
are dependent on patients in Location A in the real healthcare
service dataset, and the rest are dependent on patients in
Locations B and C in the low-level assumption of healthcare
service dataset. The evaluation process for each location is
50071
 O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
FIGURE 19. Locations B and C and 12 hospitals within the capital Kuala
Lumpur.
likewise tested for patients in emergency risk case. The final
evaluation results of the three scenarios are shown in Table 11.
B. HOSPITAL RANKING RESULTS
The calculation of the weights is presented in the next
subsections.
1) MLAHP SETTING WEIGHTS RESULTS FOR THE MAIN
CRITERIA AND SUB-CRITERIA
The pairwise uses mathematical calculations to convert the
judgments of the six experts into weights for the main cri-
teria and sub-criteria. The original and normalised matrixes
and the aggregation calculation measurements to obtain the
weights of the main criteria and sub-criteria from the six
experts are shown in Table 2 and Tables 3 in the Appendix.
Furthermore, Tables 4 in the Appendix present the package
multi-service local and global priorities from the six experts.
This table shows that the overall CR for the six expert scores
is an acceptable ratio of less than 0.1 [116], [118] and [135].
The weights are listed below in Table 12; six services from
the MLAHP sequence process results are included.
The maximum and minimum criteria weights obtained
from the six experts are shown in Table 13.
To summarise, the weight results of the MLAHP for each
expert are presented. Clearly, the results show variances
among the weights obtained from the six experts. Applying
50072
TABLE 14. Final MLAHP local and global weights arithmetic mean for six
experts.
TABLE 15. Final GCI arithmetic mean weight of the services criteria for six
experts.
TABLE 16. Final GCI arithmetic mean weight of the TAH criteria for six
experts.
the arithmetic mean for the final weighs of the six experts is
required to provide a ranking for the hospitals considering
the overall DM according to [103], [113] and [148]. The
calculation of the arithmetic means for the final weights of the
six experts for ranking the scenarios is illustrated in Table 14.
2) AHP RESULTS IN RANKING HOSPITALS
The pairwise uses mathematical calculations to convert the
judgments of the six experts into the GCI weights of the
service (available and unavailable) and TAH criteria (SstT,
ST, LT and LstT). The weight results of the GCI for each
expert are presented in the following subsections. Clearly,
the results show variances among the weights obtained from
the six experts. Therefore, applying the arithmetic mean for
the final weighs of the six experts is required to rank the
hospitals considering the overall DM. The calculation of the
arithmetic means for the weights of the six experts for ranking
the scenarios is illustrated in the tables in each subsection.
a: GCI WEIGHTS FOR SERVICE CRITERIA
The final GCI arithmetic means weights of the service from
the six experts are presented in Table 15. The original and
normalised matrixes and the aggregation calculation of the
service criteria of packages 1 from the six experts are shown
in Tables 5 in the Appendix.
VOLUME 7, 2019
O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
TABLE 17. Final ranking results.
b: GCI WEIGHTS FOR TAH CRITERIA
In the same context, the final GCI arithmetic means weights
of the TAH criteria from the six experts are presented
in Table 16. The original and normalised matrixes and the
aggregation calculation for the TAH criteria of packages 1
from the six experts are shown in Tables 6 in the Appendix.
3) HOSPITAL RANKING RESULTS FOR SCENARIOS
The obtained MLAHP weights of the main criteria and
sub-criteria and the obtained GCI weights of the service
criteria and TAH are used with the scenario evaluations pre-
sented previously in Table 11. The final results of the hospital
rankings are shown in Table 17.
The available hospitals’ scores are ranked in descending
order for all the scenarios, which are dependent on the AHP
method. The AHP allocates the scores for all the hospitals
VOLUME 7, 2019
dependent on the arithmetic means of the final weights of the
MLAHP and the GCI from the six experts. The hospitals are
ranked from highest to lowest based on service availability
and relative TAH. In addition, the healthcare service datasets
are classified into two levels in terms of service availability.
The real hospital healthcare service dataset represents the
maximum availability of services (high level) in Location A,
and the assumption healthcare service dataset represents the
minimum availability of services (low level) in Locations B
and C. The results are debated on to show the differences in
the hospital rankings regarding the varied service availability
levels and patient locations, as explained below.
The obtained results provide conclusive evidence that the
hospital-ranking process depends not only on the TAH cri-
teria but also on the availability of services, which play a
crucial role in the process of selecting the best hospital.
50073
 O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
The hospital-ranking process cannot be determined by a spe-
cific situation because numerous factors combine to impact
the priority setting at the hospital level. For example, a hospi-
tal with three available services with the shortest TAH crite-
rion may be better than a hospital with four available services
with a short or long TAH criterion. In another example, two
hospitals with the same number of available services with the
same TAH criterion may gain different priority levels because
one hospital may have services with a higher weight priority
than the other. The first hospital may also have PSR, PST and
PSD services with the MLAHP weights of 0.123, 0.138 and
0.145, respectively, whereas the second hospital may have
POS, SA and PM services with weights of 0.109, 0.077 and
0.141, respectively. Although both hospitals in this case have
the same number of services, their scores will differ based on
the variety of the weights.
In addition, the results of scenarios 1 prove that when
hospitals have the maximum level of service availability,
the selection of the best hospital depends only on the TAH cri-
terion. In these scenarios, the services of this package datasets
are available, and thus the results of the hospital sequences for
all the packages are the same. The results of scenarios 2 and
3 prove that when hospitals change their levels of service (the
minimum level), the selection of the best hospital depends on
the service availability and the TAH criteria. In conclusion,
the ranking results of the AHP based on the arithmetic mean
from the six experts are introduced, and the hospital rankings
for more than one scenario are described and discussed.
V. RESEARCH LIMITATIONS
Most research has limitations which can be solved in future
studies. The scope of this research includes a few limitations.
Second generation telemedicine for real-time situations for
the proposed framework have not yet been implemented.
In addition, the framework of this study only connects
patients with the best hospital but does not notify them of the
types of services available after a hospital has been selected.
The procedures for assessing health conditions also change
frequently, which in turn affects the healthcare service pro-
visions from the hospital to the patient. In other words, the
changing health condition of the patient leads to changes in
the appropriate hospital, but the proposed framework does not
provide a solution for such situations.
VI. RECOMMENDATIONS FOR FUTURE WORK
The recommendations for future work are as follows:
1. A future trend is to consider multiple chronic dis-
eases in case studies. Therefore, medical guidelines should
be improved, the types and numbers of medical services
involved identified and a new triage algorithm for multiple
chronic diseases developed.
2. Resolving the aforementioned issue will pave the way
for designing an adaptive decision-making platform for hos-
pital selection within mHealth for multiple chronic diseases,
such as heart disease, diabetes and hypertension. The plat-
form should designate priority hospitals in case of multiple
50074
bookings by several patients and consider the diversity of
healthcare service packages for patients with multiple chronic
diseases.
3. Implement the proposed framework in a real-time appli-
cation in telemedicine architecture.
4. In the event of a failed connection between mHealth
and the designated hospitals, this research only provides tem-
porary service recommendations for patients. Other services,
such as practical first aid based on the triage algorithm, could
equip the patient with the skills and knowledge to help save
his life. As an example, videos or graphic directions can save
a patient during a heart attack. The application can also create
a patient profile containing the phone numbers of relatives for
sending alarm messages in certain situations.
VII. CONCLUSION
The main goal of this research is to improve and continue
providing healthcare services to risk patients with CHD.
The improvements are achieved in the context of IoT by
an FTF-mHealth-IoT framework in Tier 2. The state-of-the-
art and multifaceted contributions of this research are as
follows: (i) designed a fault-tolerant framework for mHealth
to ensure continuous healthcare services, (ii) identified a new
algorithm (RLLT), (iii) proposed a decision matrix based on
healthcare service package/TAH and hospital list crossover
for hospital selection and (iv) used MCDM in hospital selec-
tion by adopting the unique AHP method for estimating small
power consumption.
In conclusion, hospital selection provides conclusive evi-
dence that ranking hospitals depend not only on the TAH
criterion but also on the availability of services, which play
a key role in the selection of the best hospital. Ranking
hospitals cannot be determined by a specific situation because
numerous factors combine to impact the priority setting at
the hospital level, as discussed in the Results section. Finally,
the implications of this study support indoor/outdoor patients
with specific care requirements for their modern lifestyle and
commercial and organisational medical services, as well as
assist medical staff in terms of time support.
REFERENCES
[1] A. A. Zaidan et al., ‘‘A survey on communication components for
IoT-based technologies in smart homes,’’ Telecommun. Syst., vol. 69,
no. 1, pp. 1–25, Sep. 2018.
[2] P. P. Ray, ‘‘Understanding the role of Internet of Things towards smart
e-healthcare services,’’ Biomed. Res., vol. 28, no. 4, pp. 1604–1609, 2017.
[3] H. Çalışkan, ‘‘Selection of boron based tribological hard coatings
using multi-criteria decision making methods,’’ Mater. Des., vol. 50,
pp. 742–749, Sep. 2013.
[4] M. Talal et al., ‘‘Smart home-based IoT for real-time and secure remote
health monitoring of triage and priority system using body sensors:
Multi-driven systematic review,’’ J. Med. Syst., vol. 43, no. 3, p. 42,
Mar. 2019.
[5] P. P. Ray, ‘‘Home health hub Internet of Things (H3 IoT): An architectural
framework for monitoring health of elderly people,’’ in Proc. Int. Conf.
Sci. Eng. Manage. Res. (ICSEMR), 2014, pp. 1–3.
[6] S. V. Kovalchuk, E. Krotov, P. A. Smirnov, D. A. Nasonov, and
A. N. Yakovlev, ‘‘Distributed data-driven platform for urgent decision
making in cardiological ambulance control,’’ Future Gener. Comput.
Syst., vol. 79, pp. 144–154, Feb. 2018.
VOLUME 7, 2019
O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
[7] N. Kalid et al., ‘‘Based on real time remote health monitoring systems:
A new approach for prioritization ‘large scales data’ patients with chronic
heart diseases using body sensors and communication technology,’’
J. Med. Syst., vol. 42, no. 4, p. 69, Apr. 2018.
[8] M.-Y. Chang, C. Pang, J. M. Tarn, T.-S. Liu, and D. C. Yen, ‘‘Exploring
user acceptance of an e-hospital service: An empirical study in Taiwan,’’
Comput. Standards Interfaces, vol. 38, pp. 35–43, Feb. 2015.
[9] M. V. M. Figueredo and J. S. Dias, ‘‘Mobile telemedicine system for home
care and patient monitoring,’’ in Proc. 26th Annu. Int. Conf. IEEE Eng.
Med. Biol. Soc., vol. 2, Sep. 2004, pp. 3387–3390.
[10] J. Wang, M. Qiu, and B. Guo, ‘‘Enabling real-time information service
on telehealth system over cloud-based big data platform,’’ J. Syst. Archit.,
vol. 72, pp. 69–79, Jan. 2017.
[11] O. S. Albahri, A. A. Zaidan, B. B. Zaidan, M. Hashim, A. S. Albahri,
and M. A. Alsalem, ‘‘Real-time remote health-monitoring systems in a
medical centre: A review of the provision of healthcare services-based
body sensor information, open challenges and methodological aspects,’’
J. Med. Syst., vol. 42, no. 9, p. 164, 2018.
[12] O. S. Albahri et al., ‘‘Systematic review of real-time remote health
monitoring system in triage and priority-based sensor technology: Taxon-
omy, open challenges, motivation and recommendations,’’ J. Med. Syst.,
vol. 42, no. 5, p. 80, 2018.
[13] O. H. Salman, A. A. Zaidan, B. B. Zaidan, Naserkalid, and M. Hashim,
‘‘Prioritizing using ‘big data’ patients with chronic heart diseases through
telemedicine environmental,’’ Int. J. Inf. Technol. Decis. Making, vol. 16,
no. 5, pp. 1211–1245, 2017.
[14] S. Iqbal et al., ‘‘Real-time-based E-health systems: Design and implemen-
tation of a lightweight key management protocol for securing sensitive
information of patients,’’ Health Technol., vol. 9, no. 2, pp. 93–111,
Mar. . 2019.
[15] M. Hussain et al., ‘‘Conceptual framework for the security of mobile
health applications on Android platform,’’ Telematics Inform., vol. 35,
no. 5, pp. 1335–1354, Mar. 2018.
[16] A. H. Mohsin, ‘‘Real-time remote health monitoring systems using body
sensor information and finger vein biometric verification: A multi-layer
systematic review,’’ J. Med. Syst., vol. 42, no. 12, p. 238, 2018.
[17] O. Enaizan et al., ‘‘Electronic medical record systems: Decision support
examination framework for individual, security and privacy concerns
using multi-perspective analysis,’’ in Health and Technology. Berlin,
Germany: Springer, 2018.
[18] A. H. Mohsin et al., ‘‘Real-time medical systems based on human biomet-
ric steganography: A systematic review,’’ J. Med. Syst., vol. 42, no. 12,
p. 245, 2018.
[19] M. A. Alsalem et al., ‘‘Systematic review of an automated multiclass
detection and classification system for acute Leukaemia in terms of eval-
uation and benchmarking, open challenges, issues and methodological
aspects,’’ J. Med. Syst., vol. 42, no. 11, p. 204, 2018.
[20] D. K. Moser et al., ‘‘Reducing delay in seeking treatment by patients
with acute coronary syndrome and stroke: A scientific statement
from the American Heart Association Council on cardiovascular nurs-
ing and stroke council,’’ Circulation, vol. 114, no. 2, pp. 168–182,
2006.
[21] Report of the Adhoc Committee on Health Research Relating to
Future Intervention Options (TDR/Gen/96.1). Investing in Health
Research and Development, World Health Org., Geneva, Switzerland,
1996.
[22] O. H. Salman, M. F. A. Rasid, M. I. Saripan, and S. K. Subramaniam,
‘‘Multi-sources data fusion framework for remote triage prioritization in
telehealth,’’ J. Med. Syst., vol. 38, no. 9, p. 103, Sep. 2014.
[23] R. W. Derlet, D. Kinser, L. Ray, B. Hamilton, and J. McKenzie, ‘‘Prospec-
tive identification and triage of nonemergency patients out of an emer-
gency department: A 5-year study,’’ Ann. Emergency Med., vol. 25, no. 2,
pp. 215–223, 1995.
[24] A. S. Albahri, A. A. Zaidan, O. S. Albahri, B. B. Zaidan, and
M. A. Alsalem, ‘‘Real-time fault-tolerant mHealth system: Comprehen-
sive review of healthcare services, opens issues, challenges and method-
ological aspects,’’ J. Med. Syst., vol. 42, no. 8, p. 137, 2018.
[25] A. Sene, B. Kamsu-Foguem, and P. Rumeau, ‘‘Telemedicine framework
using case-based reasoning with evidences,’’ Comput. Methods Programs
Biomed., vol. 121, no. 1, pp. 21–35, Aug. 2015.
[26] L. van Dyk, ‘‘A review of telehealth service implementation frame-
works,’’ Int. J. Environ. Res. Public Health, vol. 11, no. 2, pp. 1279–1298,
2014.
VOLUME 7, 2019
[27] S. Jeong, C.-H. Youn, E. B. Shim, M. Kim, Y. M. Cho, and L. Peng,
‘‘An integrated healthcare system for personalized chronic disease care
in home–hospital environments,’’ IEEE Trans. Inf. Technol. Biomed.,
vol. 16, no. 4, pp. 572–585, Jul. 2012.
[28] J. L. B. Cineros and O. Lund, ‘‘KmerFinderJS: A client-server method for
fast species typing of bacteria over slow Internet connections,’’ BioRxiv,
p. 145284, 2017. doi: 10.1101/145284.
[29] T. Duong-Ba, T. Nguyen, B. Bose, and D. A. Tran, ‘‘Distributed
client-server assignment for online social network applications,’’ IEEE
Trans. Emerg. Topics Comput., vol. 2, no. 4, pp. 422–435, Dec. 2014.
[30] A. Wood, ‘‘Predicting client/server availability,’’ Computer, vol. 28, no. 4,
pp. 41–48, Apr. 1995.
[31] M. L. Shuwandy, B. B. Zaidan, A. A. Zaidan, and A. S. Albahri, ‘‘Sensor-
based mhealth authentication for real-time remote healthcare monitoring
system: A multilayer systematic review,’’ J. Med. Syst., vol. 43, no. 2,
p. 33, Feb. 2019.
[32] A. H. Mohsin et al., ‘‘Blockchain authentication of network applications:
Taxonomy, classification, capabilities, open challenges, motivations, rec-
ommendations and future directions,’’ Comput. Standards Interfaces,
vol. 64, pp. 41–60, May 2019.
[33] J. Dong and G.-H. Yang, ‘‘Reliable state feedback control of T–S fuzzy
systems with sensor faults,’’ IEEE Trans. Fuzzy Syst., vol. 23, no. 2,
pp. 421–433, Apr. 2015.
[34] Y. Xiao and H. Chen, Mobile Telemedicine: A Computing and Networking
Perspective. New York, NY, USA: Auerbach Publications, 2008.
[35] B. Randell, P. Lee, and P. C. Treleaven, ‘‘Reliability issues in computing
system design,’’ ACM Comput. Surv., vol. 10, no. 2, pp. 123–165, 1978.
[36] S. Murtaza and R. Al, ‘‘Qos taxonomy towards wireless body area
network solutions,’’ Int. J. Appl. Innov. Eng. Manage., vol. 2, no. 4,
pp. 221–234, 2013.
[37] A. Lounis, A. Hadjidj, A. Bouabdallah, and Y. Challal, ‘‘Healing on the
cloud: Secure cloud architecture for medical wireless sensor networks,’’
Future Gener. Comput. Syst., vol. 55, pp. 266–277, Feb. 2016.
[38] P. Pawar, V. Jones, B.-J. F. van Beijnum, and H. Hermens, ‘‘A framework
for the comparison of mobile patient monitoring systems,’’ J. Biomed.
Inform., vol. 45, no. 3, pp. 544–556, Jun. 2012.
[39] A. A. Rezaee, M. H. Yaghmaee, and A. M. Rahmani, ‘‘Optimized con-
gestion management protocol for healthcare wireless sensor networks,’’
Wireless Pers. Commun., vol. 75, no. 1, pp. 11–34, Mar. 2014.
[40] S. Ghanavati, J. Abawaji, and D. Izadi, ‘‘A congestion control scheme
based on fuzzy logic in wireless body area networks,’’ in Proc. IEEE
14th Int. Symp. Netw. Comput. Appl., Sep. 2015, pp. 235–242.
[41] A. A. Rezaee, M. H. Yaghmaee, A. M. Rahmani, and A. H. Mohajerzadeh,
‘‘HOCA: Healthcare aware optimized congestion avoidance and control
protocol for wireless sensor networks,’’ J. Netw. Comput. Appl., vol. 37,
no. 1, pp. 216–228, Jan. 2014.
[42] K. Saleem, A. Derhab, J. Al-Muhtadi, and B. Shahzad, ‘‘Human-
oriented design of secure machine-to-machine communication system
for e-healthcare society,’’ Comput. Hum. Behav., vol. 51, pp. 977–985,
Oct. 2015.
[43] M. Usman, M. R. Asghar, I. S. Ansari, and M. Qaraqe, ‘‘Security in wire-
less body area networks: From in-body to off-body communications,’’
IEEE Access, vol. 6, pp. 58064–58074, 2018.
[44] S. Moreno, A. Quintero, C. Ochoa, M. Bonfante, R. Villareal, and
J. Pestana, ‘‘Remote monitoring system of vital signs for triage and
detection of anomalous patient states in the emergency room,’’ in Proc.
21st Symp. Signal Process., Images Artif. Vis. (STSIVA), Aug./Sep. 2016,
pp. 1–5.
[45] J. R. B. dos Santos, G. Blard, A. S. R. Oliveira, and N. B. de Carvalho,
‘‘Wireless sensor tag and network for improved clinical triage,’’ in Proc.
Euromicro Conf. Digit. Syst. Design, vol. 31, 2015, pp. 399–406.
[46] J. Gómez, B. Oviedo, and E. Zhuma, ‘‘Patient monitoring system based on
Internet of Things,’’ Procedia Comput. Sci., vol. 83, pp. 90–97, Dec. 2016.
[47] A. Hussain, R. Wenbi, A. L. da Silva, M. Nadher, and M. Mudhish,
‘‘Health and emergency-care platform for the elderly and disabled people
in the smart city,’’ J. Syst. Softw., vol. 110, pp. 253–263, Dec. 2015.
[48] W.-T. Sung and K.-Y. Chang, ‘‘Health parameter monitoring via a
novel wireless system,’’ Appl. Soft Comput., vol. 22, pp. 667–680,
Sep. 2014.
[49] S. V. Zanjal and G. R. Talmale, ‘‘Medicine reminder and monitoring
system for secure health using IOT,’’ Procedia Comput. Sci., vol. 78,
pp. 471–476, Dec. 2016.
50075
 O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
[50] J. Mendes, H. Simões, P. Rosa, N. Costa, C. Rabadão, and A. Pereira,
‘‘Secure low-cost solution for elder’s eCardio surveillance,’’ Procedia
Comput. Sci., vol. 27, pp. 46–56, Jan. 2014.
[51] G. C. Lamprinakos et al., ‘‘An integrated remote monitoring plat-
form towards telehealth and telecare services interoperability,’’ Inf. Sci.,
vol. 308, pp. 23–37, Mar. 2015.
[52] A. Ahmed, A. Rebeiro-Hargrave, Y. Nohara, E. Kai, Z. H. Ripon, and
N. Nakashima, ‘‘Targeting morbidity in unreached communities using
portable health clinic system,’’ in IEICE Trans. Commun., vol. E97-B,
no. 3, pp. 540–545, 2014.
[53] K. Ganapathy, B. Priya, B. Priya, Dhivya, V. Prashanth, and V. Vaidehi,
‘‘SOA framework for geriatric remote health care using wireless sensor
network,’’ Procedia Comput. Sci., vol. 19, pp. 1012–1019, 2013.
[54] S. J. Miah, J. Hasan, and J. G. Gammack, ‘‘On-cloud healthcare clinic:
An e-health consultancy approach for remote communities in a develop-
ing country,’’ Telematics Inform., vol. 34, no. 1, pp. 311–322, 2017.
[55] G. Traverso et al., ‘‘Physiologic status monitoring via the gastrointestinal
tract,’’ PLoS ONE, vol. 10, no. 11, 2015, Art. no. e0141666.
[56] R. Rajkumar and N. C. S. N. Iyengar, ‘‘Dynamic integration of mobile
JXTA with cloud computing for emergency rural public health care,’’
Osong Public Health Res. Perspect., vol. 4, no. 5, pp. 255–264, 2013.
[57] N. Kumar, K. Kaur, A. Jindal, and J. J. P. C. Rodrigues, ‘‘Providing health-
care services on-the-fly using multi-player cooperation game theory in
Internet of Vehicles (IoV) environment,’’ Digit. Commun. Netw., vol. 1,
no. 3, pp. 191–203, Aug. 2015.
[58] S. Moretti, S. Cicalò, M. Mazzotti, V. Tralli, and M. Chiani,
‘‘Content/context-aware multiple camera selection and video adaptation
for the support of m-Health services,’’ Procedia Comput. Sci., vol. 40,
pp. 206–213, Jan. 2014.
[59] P. Calyam et al., ‘‘Synchronous big data analytics for personalized and
remote physical therapy,’’ Pervasive Mobile Comput., vol. 28, pp. 3–20,
Jun. 2016.
[60] A. Katib, D. Rao, P. Rao, K. Williams, and J. Grant, ‘‘A prototype of
a novel cell phone application for tracking the vaccination coverage of
children in rural communities,’’ Comput. Methods Programs Biomed.,
vol. 122, no. 2, pp. 215–228, Nov. 2015.
[61] O. Boursalie, R. Samavi, and T. E. Doyle, ‘‘M4CVD: Mobile machine
learning model for monitoring cardiovascular disease,’’ Procedia Com-
put. Sci., vol. 63, no. 2, pp. 384–391, 2015.
[62] M. Fezari, R. Rasras, and I. M. M. El Emary, ‘‘Ambulatory health
monitoring system using wireless sensors node,’’ Procedia Comput. Sci.,
vol. 65, pp. 86–94, Jan. 2015.
[63] V. Villarreal, J. Fontecha, R. Hervas, and J. Bravo, ‘‘Mobile and ubiq-
uitous architecture for the medical control of chronic diseases through
the use of intelligent devices: Using the architecture for patients with
diabetes,’’ Future Gener. Comput. Syst., vol. 34, pp. 161–175, May 2014.
[64] M. Sebillo, G. Tortora, M. Tucci, G. Vitiello, A. Ginige, and
P. Di Giovanni, ‘‘Combining personal diaries with territorial intelligence
to empower diabetic patients,’’ J. Vis. Lang. Comput., vol. 29, pp. 1–14,
Aug. 2015.
[65] M. O. Lwin et al., ‘‘A 21st century approach to tackling dengue: Crowd-
sourced surveillance, predictive mapping and tailored communication,’’
Acta Tropica, vol. 130, no. 1, pp. 100–107, Feb. 2014.
[66] A. Bresó, J. Martínez-Miranda, E. Fuster-García, and J. M. García-
Gómez, ‘‘A novel approach to improve the planning of adaptive and
interactive sessions for the treatment of Major Depression,’’ Int. J. Hum.-
Comput. Stud., vol. 87, pp. 80–91, Mar. 2015.
[67] H. Hermens, H. op den Akker, M. Tabak, J. Wijsman, and M. Vollenbroek,
‘‘Personalized coaching systems to support healthy behavior in people
with chronic conditions,’’ J. Electromyogr. Kinesiol., vol. 24, no. 6,
pp. 815–826, Dec. 2014.
[68] J. Broach et al., ‘‘Usability and reliability of smart glasses for secondary
triage during mass casualty incidents,’’ in Proc. Annu. Hawaii Int. Conf.
Syst. Sci., 2018, pp. 1416–1422.
[69] A. Paulus, P. Meisen, T. Meisen, S. Jeschke, M. Czaplik, and F. Hirsch,
‘‘AUDIME: Augmented disaster medicine,’’ in Proc. 17th Int. Conf.
E-Health Netw., Appl. Services (HealthCom), 2016, pp. 342–345.
[70] C. Beck and J. Georgiou, ‘‘A wearable, multimodal, vitals acquisition unit
for intelligent field triage,’’ in Proc. IEEE Int. Symp. Circuits Syst., no. 3,
May 2016, pp. 1530–1533.
[71] L. I. Besaleva and A. C. Weaver, ‘‘CrowdHelp: M-Health application
for emergency response improvement through crowdsourced and sensor-
detected information,’’ in Proc. Wireless Telecommun. Symp., 2014,
pp. 1–5.
50076
[72] A. Ganz, J. M. Schafer, J. Tang, Z. Yang, J. Yi, and G. Ciottone, ‘‘Urban
search and rescue situational awareness using DIORAMA disaster man-
agement system,’’ Procedia Eng., vol. 107, pp. 349–356, Jan. 2015.
[73] A. Renner et al., ‘‘RIPPLE: Scalable medical telemetry system for sup-
porting combat rescue,’’ in Proc. IEEE Nat. Aerosp. Electron. Conf.,
Jun. 2014, pp. 228–232.
[74] S. Gunasekaran and M. Suresh, ‘‘A novel control of disaster protec-
tion (NCDP) for pilgrims by pan technology,’’ in Proc. IEEE 8th Int. Conf.
Intell. Syst. Control (ISCO), Jan. 2014, pp. 103–107.
[75] S. Adibi, ‘‘A mobile health network disaster management system,’’ in
Proc. 7th Int. Conf. Ubiquitous Future Netw., Jul. 2015, pp. 424–428.
[76] S. Sneha and U. Varshney, ‘‘A framework for enabling patient monitor-
ing via mobile ad hoc network,’’ Decis. Support Syst., vol. 55, no. 1,
pp. 218–234, Apr. 2013.
[77] A. Fratini and M. Caleffi, ‘‘Medical emergency alarm dissemination in
urban environments,’’ Telematics Inform., vol. 31, no. 3, pp. 511–517,
Aug. 2014.
[78] Y. Qin, L. Li, X. Zhong, Y. Yang, and C. L. Gwee, ‘‘A cross-layer QoS
design with energy and traffic balance aware for different types of traffic
in MANETs,’’ Wireless Pers. Commun., vol. 85, no. 3, pp. 1429–1449,
2015.
[79] S. Manfredi, ‘‘Congestion control for differentiated healthcare service
delivery in emerging heterogeneous wireless body area networks,’’ IEEE
Wireless Commun., vol. 21, no. 2, pp. 80–90, Apr. 2014.
[80] J. Soto, S. Queiroz, and M. Nogueira, ‘‘Managing sensing and coopera-
tion to analyze PUE attacks in cognitive radio ad hoc networks,’’ in Proc.
8th Int. Conf. Netw. Service Manage. (CNSM) Workshop Syst. Virtualiz-
tion Manage. (SVM), 2012, pp. 219–223.
[81] H. F. Lingsma et al., ‘‘Comparing and ranking hospitals based on out-
come: Results from The Netherlands stroke survey,’’ Int. J. Med., vol. 103,
no. 2, pp. 99–108, 2009.
[82] J. Leister and J. Stausberg, ‘‘Why do patients select a hospital?’’
J. Hospital Marketing Public Relations, vol. 17, no. 2, pp. 13–31,
Oct. 2007.
[83] L. H. Wizig, ‘‘Method and system for providing a user-selected healthcare
services package and healthcare services panel customized based on a
user’s selections,’’ U.S. Patents 6 735 569 B1, May 11, 2004.
[84] R. G. Robbinson, K. L. Kudos, L. R. Starr, K. Rao, and T. R. Price,
‘‘Mood disorders in stroke patients. Importance of lesion location.,’’
Brain, vol. 107, pp. 81–93, Mar. 1984.
[85] W. G. Barsan et al., ‘‘Time of hospital presentation in patients with acute
stroke,’’ Arch. Internal Med., vol. 153, no. 22, pp. 2558–2561, 1993.
[86] J. Faulin, A. A. Juan, S. E. Grasman, and M. J. Fry, Decision Making in
Service Industries: A Practical Approach. Boca Raton, FL, USA: CRC
Press, 2012.
[87] Q. M. Yas, A. A. Zaidan, B. B. Zaidan, B. Rahmatullah, and H. A. Karim,
‘‘Comprehensive insights into evaluation and benchmarking of real-time
skin detectors: Review, open issues & challenges, and recommended
solutions,’’ Measurement, vol. 114, pp. 243–260, Jan. 2018.
[88] O. Zughoul et al., ‘‘Comprehensive insights into the criteria of student
performance in various educational domains,’’ IEEE Access, vol. 6,
pp. 73245–73264, 2018.
[89] N. F. Berglas, M. F. Battistelli, W. K. Nicholson, M. Sobota, R. D. Urman,
and S. C. M. Roberts, ‘‘The effect of facility characteristics on patient
safety, patient experience, and service availability for procedures in non-
hospital-affiliated outpatient settings: A systematic review,’’ PLoS ONE,
vol. 13, no. 1, 2018, Art. no. e0190975.
[90] J. Nicholl, J. West, S. Goodacre, and J. Turner, ‘‘The relationship between
distance to hospital and patient mortality in emergencies: An observa-
tional study,’’ Emergency Med. J., vol. 24, no. 9, pp. 665–668, Sep. 2007.
[91] L. Wei et al., ‘‘Impact on mortality following first acute myocardial
infarction of distance between home and hospital: Cohort study,’’ Heart,
vol. 94, no. 9, pp. 1141–1146, 2008.
[92] R. Busse, J. Schreyögg, and P. C. Smith, ‘‘Variability in healthcare
treatment costs amongst nine eu countries—Results from the healthbasket
project,’’ Health Econ., vol. 17, pp. S1–S8, Jan. 2008.
[93] D. H. Akdag, T. Kalaycı, S. Karagöz, H. Zülfikar, and D. Giz, ‘‘The eval-
uation of hospital service quality by fuzzy MCDM,’’ Appl. Soft Comput.,
vol. 23, pp. 239–248, Oct. 2014.
[94] A. M. R. Khan, P. N. Prasad, and S. Rajamanoharane, ‘‘A decision-making
framework for service quality measurements in hospitals,’’ Int. J. Enter-
prise Netw. Manage., vol. 4, no. 1, p. 80, 2010.
VOLUME 7, 2019
O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
[95] R. L. Keeney and H. Raiffa, Decisions with Multiple Objectives: Prefer-
ences and Value Trade-Offs. Cambridge, U.K.: Cambridge Univ. Press,
1976.
[96] B. Valerie and T. J. Stewart, Multiple Criteria Decision Analysis:
An Integrated Approach. Boston, MA, USA: Kluwer, 2002.
[97] M. Whaiduzzaman, A. Gani, N. B. Anuar, M. Shiraz, M. N. Haque, and
I. T. Haque, ‘‘Cloud service selection using multicriteria decision analy-
sis,’’ Sci. World J., vol. 2014, Feb. 2014, Art. no. 459375.
[98] A. A. Zaidan, B. B. Zaidan, Z. Kadhem, M. Larbani, M. B. Lakulu, and
M. Hashim, ‘‘Challenges, alternatives, and paths to sustainability: Better
public health promotion using social networking pages as key tools,’’
J. Med. Syst., vol. 39, no. 2, p. 7, Feb. 2015.
[99] B. B. Zaidan, A. Haiqi, A. A. Zaidan, M. Abdulnabi, M. L. M. Kiah, and
H. Muzamel, ‘‘A security framework for nationwide health information
exchange based on telehealth strategy,’’ J. Med. Syst., vol. 39, no. 5, p. 51,
May 2015.
[100] M. L. M. Kiah, M. S. Nabi, B. B. Zaidan, and A. A. Zaidan, ‘‘An enhanced
security solution for electronic medical records based on AES hybrid
technique with SOAP/XML and SHA-1,’’ J. Med. Syst., vol. 37, no. 5,
p. 9971, Oct. 2013.
[101] M. S. Nabi, M. L. Mat Kiah, A. A. Zaidan, and B. B. Zaidan, ‘‘Suitability
of adopting S/MIME and OpenPGP email messages protocol to secure
electronic medical records,’’ in Proc. 2nd Int. Conf. Future Gener. Com-
mun. Technol. (FGCT), 2013, pp. 93–97.
[102] M. L. M. Kiah, A. Haiqi, B. B. Zaidan, and A. A. Zaidan, ‘‘Open
source EMR software: Profiling, insights and hands-on analysis,’’ Com-
put. Methods Programs Biomed., vol. 117, no. 2, pp. 360–382, 2014.
[103] C.-L. Hwang and K. Yoon, Multiple Attribute Decision Methods and
Applications. Springer Science & Business Media, 2011.
[104] B. B. Zaidan, A. A. Zaidan, and M. L. M. Kiah, ‘‘Impact of data privacy
and confidentiality on developing telemedicine applications: A review
participates opinion and expert concerns,’’ Int. J. Pharmacol., vol. 7, no. 3,
pp. 382–387, 2011.
[105] M. L. M. Kiah, B. B. Zaidan, A. A. Zaidan, M. Nabi, and R. Ibraheem,
‘‘MIRASS: Medical informatics research activity support system using
information mashup network,’’ J. Med. Syst., vol. 38, no. 4, p. 37,
Apr. 2014.
[106] M. L. M. Kiah, S. H. Al-Bakri, A. A. Zaidan, B. B. Zaidan, and
M. Hussain, ‘‘Design and develop a video conferencing framework for
real-time telemedicine applications using secure group-based communi-
cation architecture,’’ J. Med. Syst., vol. 38, no. 10, p. 133, Oct. 2014.
[107] E. Triantaphyllou, Multi-criteria Decision Making Methods: A Compar-
ative Study, vol. 44. Boston, MA, USA: Springer, 2000, pp. 5–12.
[108] A. A. Zaidan, ‘‘A review on smartphone skin cancer diagnosis apps
in evaluation and benchmarking: Coherent taxonomy, open issues and
recommendation pathway solution,’’ Health Technol., vol. 8, no. 4,
pp. 223–238, Sep. 2018.
[109] B. B. Zaidan, A. A. Zaidan, H. A. Karim, and N. N. Ahmad, ‘‘A new
approach based on multi-dimensional evaluation and benchmarking for
data hiding techniques,’’ Int. J. Inf. Technol. Decis. Making, vol. 16, pp. 1–
42, Mar. 2017.
[110] F. M. Jumaah, A. A. Zaidan, B. B. Zaidan, R. Bahbibi, M. Y. Qahtan,
and A. Sali, ‘‘Technique for order performance by similarity to ideal
solution for solving complex situations in multi-criteria optimization of
the tracking channels of GPS baseband telecommunication receivers,’’
Telecommun. Syst., vol. 68, no. 3, pp. 425–443, Jul. 2018.
[111] B. Rahmatullah, A. A. Zaidan, F. Mohamed, and A. Sali, ‘‘Multi-complex
attributes analysis for optimum GPS baseband receiver tracking channels
selection,’’ in Proc. 4th Int. Conf. Control, Decis. Inf. Technol. (CoDIT),
2017, pp. 1084–1088.
[112] E. Triantaphyllou, B. Shu, S. N. Sanchez, and T. Ray, ‘‘Multi-criteria
decision making: An operations research approach,’’ Encyclopedia Elect.
Electron. Eng., vol. 15, no. 1998, pp. 175–186, 1998.
[113] K. Yoon and C.-L. Hwang, Multiple Attribute Decision Making:
An Introduction, vol. 104. Thousand Oaks, CA, USA: Sage Publications,
1995.
[114] A. A. Zaidan, B. B Zaidan, A. Al-Haiqi, M. L. M. Kiah, M. Hussain, and
M. Abdulnabi, ‘‘Evaluation and selection of open-source EMR software
packages based on integrated AHP and TOPSIS,’’ J. Biomed. Inform.,
vol. 53, pp. 390–404, Feb. 2015.
[115] H.-S. Shih, H.-J. Shyur, and E. S. Lee, ‘‘An extension of TOPSIS for group
decision making,’’ Math. Comput. Model., vol. 45, nos. 7–8, pp. 801–813,
2007.
VOLUME 7, 2019
[116] T. L. Saaty and M. S. Ozdemir, ‘‘Why the magic number seven plus or
minus two,’’ Math. Comput. Model., vol. 38, nos. 3–4, pp. 233–244, 2003.
[117] D. Lesmes, M. Castillo, and R. Zarama, ‘‘Application of the analytic
network process (ANP) to establish weights in order to re-accredit a
program of a University,’’ in Proc. 10th Int. Symp. Analytic Hierarchy
Process, vol. 29, 2009, pp. 1–14.
[118] L. Pecchia, P. A. Bath, N. Pendleton, and M. Bracale, ‘‘Analytic hierarchy
process (AHP) for examining healthcare professionals’ assessments of
risk factors,’’ Methods Inf. Med., vol. 50, no. 5, pp. 435–444, Dec. 2011.
[119] M. Aruldoss, T. M. Lakshmi, and V. P. Venkatesan, ‘‘A survey on multi
criteria decision making methods and its applications,’’ Amer. J. Inf. Syst.,
vol. 1, no. 1, pp. 31–43, 2013.
[120] N. H. Zardari, K. Ahmed, S. M. Shirazi, and Z. Bin Yusop, ‘‘Weighting
methods and their effects on multi-criteria decision making model out-
comes in water resources management,’’ Int. J. Oper. Res., vol. 10, no. 2,
pp. 56–66, 2015.
[121] M. A. O. Barrios, F. De Felice, K. P. Negrete, B. A. Romero, A. Y. Arenas,
and A. Petrillo, ‘‘An AHP-topsis integrated model for selecting the most
appropriate tomography equipment,’’ Int. J. Inf. Technol. Decis. Making,
vol. 15, no. 4, pp. 861–885, 2016.
[122] M. A. Ortiz, J. P. Cómbita, A. D. L. La Hoz, F. De Felice, and
A. Petrillo, ‘‘An integrated approach of AHP-DEMATEL methods
applied for the selection of allied hospitals in outpatient service,’’ Int. J.
Med. Eng. Inform., vol. 8, no. 2, pp. 87–107, 2016.
[123] H. O. Alanazi, A. A. Zaidan, B. B. Zaidan, M. L. M. Kiah, and
S. H. Al-Bakri, ‘‘Meeting the security requirements of electronic medical
records in the ERA of high-speed computing,’’ J. Med. Syst., vol. 39, no. 1,
p. 165, 2015.
[124] Q. M. Yas, A. A. Zaidan, B. B. Zaidan, M. Hashim, and C. K. Lim,
‘‘A systematic review on smartphone skin cancer apps: Coherent tax-
onomy, motivations, open challenges and recommendations, and new
research direction,’’ J. Circuits, Syst. Comput., vol. 27, no. 5, May 2018,
Art. no. 1830003.
[125] M. Abdulnabi, A. Al-Haiqi, M. L. M. Kiah, A. A. Zaidan, B. B. Zaidan,
and M. Hussain, ‘‘A distributed framework for health information
exchange using smartphone technologies,’’ J. Biomed. Inform., vol. 69,
pp. 230–250, May 2017.
[126] World Health Organization. (2016). Health Resources Availability
Mapping System (HeRAMS). Manila, Philippines: WHO
Regional Office for the Western Pacific. [Online]. Available:
https://www.who.int/hac/herams/en/
[127] M.-J. Lee and C.-W. Chung, ‘‘A user similarity calculation based on the
location for social network services,’’ in Database Systems for Advanced
Applications (Lecture Notes in Computer Science), vol. 6587. Berlin,
Germany: Springer, 2011, pp. 38–52.
[128] M. Lahby, L. Cherkaoui, and A. Adib, ‘‘A novel ranking algorithm based
network selection for heterogeneous wireless access,’’ J. Netw., vol. 8,
no. 2, pp. 263–272, 2013.
[129] K. Saksrisathaporn, A. Bouras, N. Reeveerakul, and A. Charles, ‘‘Appli-
cation of a decision model by using an integration of AHP and TOPSIS
approaches within humanitarian operation life cycle,’’ Int. J. Inf. Technol.
Decis. Making, vol. 15, no. 4, pp. 887–918, 2016.
[130] A. A. Zaidan, B. B. Zaidan, M. Hussain, A. Haiqi, M. L. M. Kiah, and
M. Abdulnabi, ‘‘Multi-criteria analysis for OS-EMR software selection
problem: A comparative study,’’ Decis. Support Syst., vol. 78, pp. 15–27,
Oct. 2015.
[131] B. B. Zaidan and A. A. Zaidan, ‘‘Software and hardware FPGA-based
digital watermarking and steganography approaches: Toward new
methodology for evaluation and benchmarking using multi-criteria
decision-making techniques,’’ J. Circuits, Syst. Comput., vol. 26, no. 7,
Jul. 2017, Art. no. 1750116.
[132] M. Hussain et al., ‘‘A security framework for mHealth apps on Android
platform,’’ Comput. Secur., vol. 75, pp. 191–217, Jun. 2018.
[133] H. O. Alanazi, G. M. Alam, B. B. Zaidan, and A. A. Zaidan, ‘‘Securing
electronic medical records transmissions over unsecured communica-
tions: An overview for better medical governance,’’ J. Med. Plants Res.,
vol. 4, no. 19, pp. 2059–2074, 2010.
[134] M. S. A. Nabi, M. L. M. Kiah, B. B. Zaidan, A. A. Zaidan, and
G. M. Alam, ‘‘Suitability of using SOAP protocol to secure electronic
medical record databases transmission,’’ Int. J. Pharmacol., vol. 6, no. 6,
pp. 959–964, 2010.
[135] Y. Wind and T. L. Saaty, ‘‘Marketing applications of the analytic hierarchy
process,’’ Manage. Sci., vol. 26, no. 7, pp. 641–658, 1980.
50077
 O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
[136] T. L. Saaty, ‘‘A scaling method for priorities in hierarchical structures,’’
J. Math. Psychol., vol. 15, no. 3, pp. 213–281, 1977.
[137] B. N. Abdullateef, N. F. Elias, H. Mohamed, A. A. Zaidan, and
B. B. Zaidan, ‘‘An evaluation and selection problems of OSS-LMS pack-
ages,’’ SpringerPlus, vol. 5, no. 1, p. 248, Dec. 2016.
[138] Q. M. Yas, A. A. Zadain, B. B. Zaidan, M. B. Lakulu, and B. Rahmat-
ullah, ‘‘Towards on develop a framework for the evaluation and bench-
marking of skin detectors based on artificial intelligent models using
multi-criteria decision-making techniques,’’ Int. J. Pattern Recognit. Artif.
Intell., vol. 31, no. 3, Mar. 2017, Art. no. 1759002.
[139] B. B. Zaidan, A. A. Zaidan, H. A. Karim, and N. N. Ahmad, ‘‘A new
digital watermarking evaluation and benchmarking methodology using
an external group of evaluators and multi-criteria analysis based on
‘large-scale data,’’’ Softw., Pract. Exper., vol. 47, no. 10, pp. 1365–1392,
Oct. 2017.
[140] H. Ahmadi, M. Nilashi, and O. Ibrahim, ‘‘Organizational decision to
adopt hospital information system: An empirical investigation in the
case of Malaysian public hospitals,’’ Int. J. Med. Inform., vol. 84, no. 3,
pp. 166–188, 2014.
[141] I. M. Ar and A. Kurtaran, ‘‘Evaluating the relative efficiency of commer-
cial banks in Turkey: An integrated AHP/DEA approach,’’ Int. Bus. Res.,
vol. 6, no. 4, p. 129, 2013.
[142] K. Lam and X. Zhao, ‘‘An application of quality function deployment to
improve the quality of teaching,’’ Int. J. Qual. Rel. Manage., vol. 15, no. 4,
pp. 389–413, 1998.
[143] V. Sherekar and M. Tatikonda, ‘‘Impact of factor affecting on labour
productivity in construction projects by AHP method,’’ Int. J. Eng. Sci.
Comput., vol. 6, no. 6, pp. 6771–6775, 2016.
[144] B. B. Zaidan and A. A. Zaidan, ‘‘Comparative study on the evaluation and
benchmarking information hiding approaches based multi-measurement
analysis using TOPSIS method with different normalisation, separation
and context techniques,’’ Measurement, vol. 117, pp. 277–294, May 2018.
[145] S. Kubler, J. Robert, W. Derigent, A. Voisin, and Y. Le Traon, ‘‘A state-of
the-art survey & testbed of fuzzy AHP (FAHP) applications,’’ Expert Syst.
Appl., vol. 65, pp. 398–422, Dec. 2016.
[146] F. M. Jumaah, A. A. Zadain, B. B. Zaidan, A. K. Hamzah, and R. Bahbibi,
‘‘Decision-making solution based multi-measurement design parameter
for optimization of GPS receiver tracking channels in static and dynamic
real-time positioning multipath environment,’’ Measurement, vol. 118,
pp. 83–95, Mar. 2018.
[147] I. Tariq et al., ‘‘MOGSABAT: A metaheuristic hybrid algorithm for
solving multi-objective optimisation problems,’’ Neural Computing and
Applications, Oct. 2018.
[148] S.-J. Chen and C.-L. Hwang, ‘‘Fuzzy multiple attribute decision mak-
ing methods,’’ in Fuzzy Multiple Attribute Decision Making. Berlin,
Germany: Springer, 1992, pp. 289–486.
[149] M. A. Qader, B. B. Zaidan, A. A. Zaidan, S. K. Ali, M. A. Kamaluddin,
and W. B. Radzi, ‘‘A methodology for football players selection problem
based on multi-measurements criteria analysis,’’ Measurement, vol. 111,
pp. 38–50, Dec. 2017.
[150] O. S. Albahri et al., ‘‘Based multiple heterogeneous wearable sensors:
A smart real-time health monitoring structured for hospitals distributor,’’
IEEE Access, vol. 7, pp. 37269–37323, 2019.
[151] M. Khatari, A. A. Zaidan, B. B. Zaidan, O. S. Albahri, and M. A.
Alsalem, ‘‘Multi-criteria evaluation and benchmarking for active queue
management methods: Open issues, challenges and recommended path-
way solutions,’’ Int. J. Inf. Technol. Decis. Making, vol. 18, 2019.
doi: 10.1142/S0219622019300039.
[152] B. N. Abdullateef, N. F. Elias, H. Mohamed, A. A. Zaidan,
and B. B. Zaidan, ‘‘An evaluation and selection problems of OSS-
LMS packages,’’ SpringerPlus, vol. 5, no. 1, p. 248, 2016.
O. S. ALBAHRI received the B.Sc. degree in com-
puter science from the Al Turath University Col-
lege, Baghdad, Iraq, in 2011, the M.Sc. degree in
computer science and communication from Arts,
Sciences and Technology University in Lebanon,
Beirut, Lebanon, in 2014, and the Doctor of Phi-
losophy (Ph.D.) degree in artificial intelligence
from Universiti Pendidikan Sultan Idris (UPSI),
Tanjung Malim, Malaysia, in 2019. He led or
member for many funded research projects and has
published more than 20 papers at various ISI/WOS international journals. His
research interests include decision theory, artificial intelligent, and medical
informatics.
50078
A. S. ALBAHRI received the M.Sc. degree in
ICT specialist in medical informatics and health
sciences from Arts, Sciences and Technology Uni-
versity in Lebanon, Beirut, Lebanon, in 2014,
and the Ph.D. degree in artificial intelligence
from Universiti Pendidikan Sultan Idris (UPSI),
Malaysia. He is currently a Lecturer with the
University of Information Technology and Com-
munications (UOITC) and Iraqi Commission for
Computers and Informatics (ICCI). He has pub-
lished many papers in WoS (ISI) databases under affiliation A. S. Albahri.
His research interests include AI applications on telemedicine, dis-
aster management, E-health, mHealth, machine learning, multi-criteria
decision-making (MCDM), the IoT, bigdata, and student evaluation. He is
a member and reviewer in lots of prestige journals by Clarivate Analytics.
A. A. ZAIDAN received the B.Eng. degree (Hons.)
in computer engineering from the University of
Technology, Baghdad, Iraq, in 2004, the M.Sc.
degree in data communications and computer net-
work from the University of Malaya, Malaysia,
in 2009, and the Ph.D. degree in artificial intel-
ligence from Multimedia University, Malaysia,
in 2013. He is currently a Senior Lecturer with the
Department of computing, Universiti Pendidikan
Sultan Idris. He led or member for many funded
research projects and has published more than 150 papers at various inter-
national conferences and journals. His research interests include artificial
intelligent, decision theory, data communication and networks, AI applica-
tions on telemedicine, and E-health.
B. B. ZAIDAN received the B.Sc. degree in
applied mathematics from Al-Nahrain University,
Baghdad, Iraq, in 2004, and the M.Sc. degree
in data communications and information security
from the University of Malaya, Malaysia, in 2009.
He is currently a Senior Lecturer with the Depart-
ment of Computing, Universiti Pendidikan Sultan
Idris. He led or member for many funded research
projects and has published more than 150 papers at
various international conferences and journals. His
research interests include artificial intelligent, decision theory, information
security and networks, and multi-criteria evaluation, and benchmarking.
M. A. ALSALEM received the B.Sc. and M.Sc.
degrees in computer Science from the University
of Mosul, Iraq, in 2010 and 2014, respectively,
where he is currently a Lecturer. He is currently
pursuing the Candidate Doctor of Philosophy
(Ph.D.) degree with Universiti Pendidikan Sultan
Idris (UPSI), Tanjung Malim, Malaysia. He led or
member for many funded research projects and has
published more than 15 papers at various inter-
national journals. His research interests include
machine learning, telemedicine, and multi-criteria decision making.
VOLUME 7, 2019
O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
A. H. MOHSIN received the B.Sc. degree in soft-
ware engineering from the Al-Sadiq University of
Baghdad, Iraq, in 2008, and the M.Sc. degree in
software engineering from Hamdard University,
New Delhi, India, in 2013. He is currently pursu-
ing the Candidate Doctor of Philosophy (Ph.D.)
degree with Universiti Pendidikan Sultan Idris
(UPSI), Tanjung Malim, Malaysia. He led or mem-
ber for many funded research projects and has
published more than five papers at various interna-
tional conferences and prestige journals. His research interests include data
security, software engineering, and medical informatics.
K. I. MOHAMMED received the M.Sc. degree in
computer science specialist in medical informatics
and health sciences from Anbar University, Iraq,
in 2014, and the Ph.D. degree in artificial intel-
ligence from Universiti Pendidikan Sultan Idris
(UPSI), Malaysia. He is currently a Lecturer with
Sunni Endowment Diwan. He has published many
papers in WoS (ISI) databases under affiliation
K. I. Mohammed. His research interests include
AI applications on telemedicine, disaster manage-
ment, E-health, mHealth, machine learning, multi-criteria decision-making
(MCDM), the IoT, bigdata, and student evaluation. He is a member and
reviewer in lots of prestige journals by Clarivate Analytics.
A. H. ALAMOODI received the B.Sc. degree in
information and communication technology and
the M.Sc. degree in computer networking from
Limkokwing University, Malaysia, in 2015 and
2017, respectively. He is currently pursuing the
Candidate Doctor of Philosophy (Ph.D.) degree
with Universiti Pendidikan Sultan Idris (UPSI),
Tanjung Malim, Malaysia. He has published more
than three papers at various international confer-
ences and prestige journals. His research inter-
ests include artificial intelligent, data science and prediction, and machine
learning.
SHAHAD NIDHAL received the B.Sc. degree
in electrical and electronics engineering from the
University of Technology, Iraq, in 1999, the M.Sc.
degree in electrical engineering from UKM Uni-
versity, Malaysia, in 2005, and the Ph.D. degree
from UKM Malaysia, in 2012. He is currently
a Lecturer with MSU University. He led or
member for many funded research projects and
has published more than seven papers at vari-
ous international conferences and journals. His
research interests include pattern recognition, digital signal processing, sig-
nal processing, biomedical signal processing, and renewable energy.
VOLUME 7, 2019
ODAI ENAIZAN received the B.Sc. degree in
computer science from Middle East University,
Jordon, in 2005, and the M.Sc. and Ph.D. degrees
in computer science from Universiti Sains Islam
Malaysia (USIM), Malaysia, in 2011 and 2017,
respectively. He is currently a Lecturer with Mid-
dle East University. He led or member for many
funded research projects and has published more
than seven papers at various international confer-
ences and journals. His research interests include
privacy and security health information systems MCDME-business.
M. A. CHYAD received the B.Sc. degree in com-
puter science from the University of Dialya, Iraq,
in 2010, and the M.Sc. degree in computer sci-
ence and communication from IUKL University,
Malaysia, in 2013. He is currently pursuing the
Candidate Doctor of Philosophy (Ph.D.) degree
with Universiti Pendidikan Sultan Idris (UPSI),
Tanjung Malim, Malaysia. His research interests
include artificial intelligent, decision theory, and
image classification.
KARRAR HAMEED ABDULKAREEM received
the B.S. degree in computer science (artificial
intelligence) from the University of Technology,
Iraq, in 2007, and the M.S. degree in com-
puter science (internetworking technology) from
the Universiti Teknikal Malaysia Melaka (UTeM),
Malaysia, in 2016. He is currently pursuing the
Ph.D. degree in computer science and information
technology with the Universiti Tun Hussein Onn
Malaysia (UTHM), Malaysia. His research inter-
ests multi-criteria decision making, image dehazing, and computer security.
E. M. ALMAHDI received the B.Sc. degree in
computer science from the University of Baghdad,
Iraq, in 2000, and the M.Sc. degree in com-
puter science/artificial intelligence from Universiti
Pendidikan Sultan Idris (UPSI), Tanjung Malim,
Malaysia, in 2018, where he is currently pursuing
the degree. His research interests include medical
informatics and decision making.
GHAILAN A. AL. SHAFEEY received the B.Sc.
degree in computer science from Al Mustan-
siriya University, Iraq, in 2003, and the M.Sc.
degree in computer science from IUKL University,
Malaysia, in 2014. He is currently pursuing the
Candidate Doctor of Philosophy (Ph.D.) degree
with Universiti Pendidikan Sultan Idris (UPSI),
Tanjung Malim, Malaysia. His research interests
include virtual reality, augmented reality, and data
communication.
50079
 O. S. Albahri et al.: Fault-Tolerant mHealth Framework in the Context of IoT-Based Real-Time Wearable Health Data Sensors
M. J. BAQER received the B.Sc. degree in soft-
ware engineering from the Baghdad College of
Economics Sciences University, Iraq, in 2013, and
the M.Sc. degree in information technology from
UPSI University, Malaysia, in 2018. His research
interest includes multi-criteria decision making.
ALI NAJM JASIM received the bachelor’s degree
in software engineering from the University of
Imam Jaafar Al-Sadiq, Iraq, in 2012, and the M.Sc.
degree in computer science/artificial intelligence
from Universiti Pendidikan Sultan Idris (UPSI),
Tanjung Malim, Malaysia, in 2018, where he is
currently pursuing the degree. His research inter-
ests include education application and decision
making.
50080
N. S. JALOOD received the bachelor’s degree
in software engineering from the University of
Imam Jaafar Al-Sadiq, Iraq, in 2013, and the M.Sc.
degree in computer science/artificial intelligence
from Universiti Pendidikan Sultan Idris (UPSI),
Tanjung Malim, Malaysia, in 2018, where he is
currently pursuing the degree. His research inter-
ests include education application and decision
making.
ALI. H. SHAREEF received the bachelor’s degree
in software engineering from the University of
Imam Jaafar Al-Sadiq, Iraq, in 2013, and the M.Sc.
degree in computer science/artificial intelligence
from Universiti Pendidikan Sultan Idris (UPSI),
Tanjung Malim, Malaysia, in 2018, where he is
currently pursuing the degree. His research inter-
ests include education application and decision
making.
VOLUME 7, 2019