Towards sustainable smart IoT applications architectural elements and design: opportunities, challenges, and open directions

Abstract

The convergence initiatives of cloud technology and the Internet of Things (IoT) have demonstrated a massive rise of the futuristic technologies which ensure sustainable production development of IoT applications such as intelligent transportation, smart cities, smart healthcare, and other applications. These real-time applications are employing the information and communication technology as an efficient utility to support a ubiquitous approach in the industry so that any apparatus can seamlessly interact with different cyber-physical systems and may also share its information effortlessly. Although the amalgamation of cloud and IoT offers robustness pervasive computing scenarios that can achieve the composite service with high quality and acceptable cost, it may face several challenges and issues not only in performance but also in architecture. The main aim of this study is to offer a stander framework to investigator and application designer that how sustainability can be achieved using different quality of service metrics. This is an overview of some crucial issues of IoT and the relation between it and other technologies based on various platforms. Focus is given in this study to three of the most important IoT applications that require highly successful sustainability procedures which are presented. A brief discussion and statistical categories of the research techniques published from 2000 to 2020, which highlight sustainable IoT applications, are also introduced. The contributions of this study are as follows: (i) summarizing the organization and methodology of the most important sustainable IoT applications, (ii) discussing the main challenges of the sustainable growth of IoT applications, (iii) highlighting future research directions in the field of IoT, and introducing guidelines for open business opportunities.

Full text

Vol:.(1234567890)
The Journal of Supercomputing (2021) 77:5668–5725
https://doi.org/10.1007/s11227-020-03477-7
1 3
Towards sustainable smart IoT applications architectural
elements anddesign: opportunities, challenges, andopen
directions
ZainabH.Ali1 · HeshamA.Ali1
Accepted: 20 October 2020 / Published online: 9 November 2020
© Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
The convergence initiatives of cloud technology and the Internet of Things (IoT)
have demonstrated a massive rise of the futuristic technologies which ensure sus-
tainable production development of IoT applications such as intelligent transporta-
tion, smart cities, smart healthcare, and other applications. These real-time applica-
tions are employing the information and communication technology as an efficient
utility to support a ubiquitous approach in the industry so that any apparatus can
seamlessly interact with different cyber-physical systems and may also share its
information effortlessly. Although the amalgamation of cloud and IoT offers robust-
ness pervasive computing scenarios that can achieve the composite service with
high quality and acceptable cost, it may face several challenges and issues not only
in performance but also in architecture. The main aim of this study is to offer a
stander framework to investigator and application designer that how sustainability
can be achieved using different quality of service metrics. This is an overview of
some crucial issues of IoT and the relation between it and other technologies based
on various platforms. Focus is given in this study to three of the most important
IoT applications that require highly successful sustainability procedures which are
presented. A brief discussion and statistical categories of the research techniques
published from 2000 to 2020, which highlight sustainable IoT applications, are
also introduced. The contributions of this study are as follows: (i) summarizing the
organization and methodology of the most important sustainable IoT applications,
(ii) discussing the main challenges of the sustainable growth of IoT applications,
(iii) highlighting future research directions in the field of IoT, and introducing guide-
lines for open business opportunities.
Keywords Cloud computing· Cyber-physical systems· Internet of things·
Pervasive computing· Quality of service· Service composition· Smart applications
Extended author information available on the last page of the article

5669
1 3
Towards sustainable smart IoT applications architectural…
1 Introduction
Information and communication technology (ICT) affects the shaping of IT envi-
ronment so that information can be conveniently exchanged via wireless global
village to provide a broad range of industrial networking solutions, such as
remote monitoring and control tools, real-time road conditions visibility, acceler-
ating decision-making and power reduction, and real-time data analysis.
Internet of Things (IoT) has been recognized as a scalable complex machine-
to-machine (M2M) system that allows delivering digital data among "things" in
ultra-large-scale networks [58]. This type of connection offers opened opportuni-
ties to bridge the gap between the IT industry and economic sectors. Moreover, it
has planted the seed for utilizing the ICT concept as an attempt to reduce the cost
of data acquiring and therefore achieving the long-term sustainable development
goals in IoT applications such as smart cities, smart homes, intelligent farming
environments, and intelligent transport systems (ITS) [50]. Figure1 shows sam-
ple of smart applications with IoT portfolio.
The proliferation of IoT applications accelerates the transformation of harsh
environments by supporting real-time data transmission. The experts estimate
that the connected devices rely on smart context-aware applications that will
reach approximately 50 billion devices in the current year, and the data produced
by these devices will reach 500 zettabytes [81]. As a result, the data transmission
rate over the internet has rapid growth and it should be controlled. Therefore, the
new requirements of network development have emerged as remote monitoring,
local power computation, high interoperability, greater operational efficiency, and
energy conservation. Consequently, providing end-to-end business networking
Fig. 1 A sample of IoT applications that used to stream out of digital information

5670
Z.H.Ali, H.A.Ali
1 3
solutions can support sustainable development with adjustable Quality of Service
(QoS) capability that has become imperative [19, 102].
The cloud computing technology realizes pervasive information among ubiq-
uitous computing sensors and IoT devices for enabling sustainable production
and on-demand service management [63, 89]. It can offer elastic storage capac-
ity with intelligent scale service deployment. With the increasing usage of the
pervasive information that is generated by IoT devices, the integration between
cloud computing and IoT has become a better solution to provide robustness net-
working business solutions that can intensively use on-demand. Surely, this type
of integration paradigm is a suitable way to offer robustness networking business
solutions that can intensively use from anywhere at any time. Thus, it not only
satisfies the fast growth of customer requirements toward providing end-to-end
composite service provisioning that can meet the user requirements without com-
promising its quality and cost, but it also achieves other benefits such as elastic
computation, efficient data analytic tool, and accelerate decision making using
real-time data [7, 8].
Due to the relevance of the convergence of cloud and IoT in keeping the sustain-
able growth of IoT applications, the concept has been attracting significant attention
from both academic and industry researchers within multiple genre studies. We sup-
pose that the current academic studies on sustainable IoT have reached a level of
maturity, and it has offered researchers a timely and succinct appraisal of the critical
areas and current research directions. But the comprehensive reviews which survey
the economic networking architectures and platforms are still required to offer a val-
uable information source for academic researchers on evolving and significant topic.
With our recommendation of developing countries with low resource utiliza-
tion, lower motorization, less financial resources, and less institutional and technical
capacity, we believe that the governmental institutions need to strongly support the
tendency toward applying the concept of IoT with taking into account the neces-
sity of QoS requirements on all applications. Herein, the ability to support sustain-
able development and increase productivity as well as establishing collaborative and
transparent eco-systems with high quality and low cost will be increased.
The main objective of this study is to offers an up-to-date survey of research
efforts with technical taxonomy for IoT applications focusing on their basic QoS
requirements that can achieve sustainability. According to the content of the existing
academic researches selected in this study, we intended to highlight IoT applications
from various points of view with a deep focus on three of most important IoT appli-
cations including smart city, intelligent transportation, and smart healthcare due to
their ability to ensure successful sustainable production development. In the last few
years, the concept of IoT has been undergone a lot of studies since early 2000 in an
attempt to beneficial for both academics and practitioners in improving the qual-
ity of life. In this study, the definitions of IoT terms that are released from 2005 to
2018 are detailed according to the ubiquitous computing and QoS perspectives. A
brief discussion and statistical categorization of the research techniques published
from 2000 to 2020 that focused on the wider sense and prominence on QoS, tech-
nologies, and applications along with related issues of the IoT applications are also
introduced. New guidelines for open business opportunities and future research

5671
1 3
Towards sustainable smart IoT applications architectural…
directions are demonstrated in the last section of this study. The fundamental contri-
butions of this review paper can be considered as follows:
• Summarizing the organization and methodology of the sustainable IoT applica-
tions.
• Providing a comprehensive study and analysis review of the most important IoT
applications.
• Demonstrating a brief discussion and statistically categorization of the research
techniques emerging by the sustainable growth of IoT applications.
• Addressing the fundamental challenges that face the sustainable growth of IoT
applications.
• Highlighting clear guidelines for open business opportunities and future research
directions in the field of IoT.
The rest of this review paper is organized as follows: Sect.2 discusses the main
motivation, selection methodology, and organization of IoT applications. Section3
demonstrates the challenges and limitations that face the sustainable growth of IoT
applications. Section4 provides the open research directions and proposed business
opportunities. Section5 presents the paper conclusion.
2 Motivation andorganization ofsustainable IoT applications
In this section, the motivation toward supporting the sustainable growth of IoT appli-
cations is introduced. The selection methodology that is applied to selecting papers
in this review is also illustrated. Moreover, a valuable comprehensive review based
on recent literature studies with the current technical taxonomy is demonstrated.
2.1 Motivation
Nowadays, IoT applications are not considered as stand-alone applications, but they
cooperate with other state-of-the-art computing technologies for sharing their infra-
structure and resources over public/private networks. The need to support the sus-
tainable growth of IoT applications that play an important role in developing coun-
tries that face a lack of motorization, less-developed infrastructure, less financial
resources, and insufficient institutional and technical capacity [26]. One solution
to give sustainable production development with an acceptable QoS level is to the
amalgamation of cloud computing and IoT called the "CloudIoT" paradigm [108].
The CloudIoT paradigm has brought significant benefits to internet vendors and end-
users alike, due to its capability to provide robustness ubiquitous services and deliv-
ering utility models at pay-as-you-go context [24]. Surely, the CloudIoT paradigm
not only promotes the IT industry in the fields of extremely resources availability,
elastic business models, and direct exploitation of cloud infrastructure by ubiquitous
IoT devices [22] but also enhances the overall human life quality. However, time-
sensitive applications such as ITS and healthcare may suffer from real-time adaptive

5672
Z.H.Ali, H.A.Ali
1 3
sensing and performance degradation for a long time. The main aim of this work is
to offer a stander framework to investigator and application designer that how highly
successful sustainability development can be achieved in the field of IoT.
2.2 Selection methodology
As shown in Fig.2, the procedures of our selection methodology were conducted
on three electronic databases for selection and analyzed the related review studies
on IoT applications. These procedures are fuelled by the following steps: (i) the use
of the leading scientific databases for searching, and Table1 depicts the electronic
databases include, Science Direct (Elsevier), IEEE Xplore, and Springer that are
used in the review process; (ii) paper filtration according to eligibility criteria such
as published year, paper context, and QoS evaluation metrics; (iii) the investigation
of the eligibility criteria; and finally (iv) the eligible papers are selected. With the
context of selection methodology, we are keen to display a large scale of literature
reviews without duplication to guide academic researchers for the importance of the
field of sustainability for IoT applications. Meanwhile, we ignored several papers
that have ineligible criteria include white papers, book chapters, short papers, sur-
veyed papers, papers have not considered the amalgamation of cloud and IoT, papers
have not considered QoS performance metrics, non-English written papers, and low-
quality papers.
2.3 Organization ofIoT applications
Thus far IoT is still considered a novel term and a fuzzy idea among internet users,
although it has strongly imposed itself in the recent period as one of the leading
technologies in the sensory world. The term IoT was started life in 1997 by the Inter-
national Telecommunication Union (ITU) under the title "Challenges to the Net-
work: Telecommunications and the Internet." After that, the ITU submitted a series
of issues under various addresses to clear the notion of IoT, including Internet for
Development (1999), IP Telephony (2001), Internet for a Mobile Generation (2002),
The Birth of Broadband (2003), and The Portable Internet (2004) [73]. The defini-
tion with extreme precision under the phrase of "Internet of Things" or "Internet
of Object" was coined by Kevin Ashton as "the ability of the networked devices to
propagate their information about the physical world objects through the web" [47].
Table2 provides the summarization of the IoT definitions published in 2005–2018.
These definitions are classified according to several themes that are explored and
analyzed in the context of the wider literature.
After the amalgamation of cloud computing and IoT, the IoT has become an
integral part of the ubiquitous computing scenario. So that it allows establishing an
on-demand direct communication among a set of physical and virtual entities from
everywhere at anytime [24]. From a ubiquitous computing point of view, the cloud
computing technology not only enables shifting from the traditional services to the
IoT-composite services but also allows enhancing interactions of the smart things
and providing a flexible place can store, compute, and manage the outcomes from

5673
1 3
Towards sustainable smart IoT applications architectural…
these things [12]. Surely, this amalgamation have a high effect on supporting the
sustainable growth of smart applications such as smart cities, smart mobility, smart
agriculture environment, and smart healthcare.
As shown in Fig. 3, from the QoS perspective, realizing sustainable produc-
tion development of the smart application is required to measure different QoS
Fig. 2 The flowchart of selection methodology for literature reviews

5674
Z.H.Ali, H.A.Ali
1 3
levels such as reliability, robustness, interoperability, scalability, resource utiliza-
tion, power consumption, and security. When we take a look at the smart applica-
tions from the ubiquitous computing perspective, we will find that each one of these
applications needs to integrate with at less one of the computing platforms to ensure
continuity and future sustainability. For instance, the application of ITS can inte-
grate efficiently with all types of computing platforms such as cloud computing,
mobile cloud computing, vehicular cloud computing, and edge computing. How-
ever, smart healthcare applications can integrate with the cloud, fog, mobile, and
edge computing only. This integration typically creates trust services and improves
overall performance. From a ubiquitous computing perspective, the cloud comput-
ing technology not only enables shifting from the traditional services to the IoT-
composite services but also allows simplifying interactions of the smart things with
these services and providing a flexible place can store, compute, and manage the
outcomes from these things.
2.4 Smart city: overview
A smart city utilizes digital transformation to connect, protect, expand, and enhance
the lives of citizens and quality. The IoT sensors, surveillance cameras, social media,
and other harsh environments act as a digital nervous system, increasing automotive
operations, and reducing IoT devices deployment risk.
Several alternative common terminologies include "Digital City", "Ubiquitous
City", "Information City", and "Knowledge-based City" that have been used in this
review paper to obtain all possible academic articles relevant to the term of "smart
city". According to the Elsevier publisher, there are up to 34,410 published articles
under the name of "Smart City" and its counterparts from the duration from 2000
to 2020. While there are 22,596 papers published in the online Springer database,
estimated approximately 19,140 research articles have also been published in the
IEEE database during the same period. Figure4 shows the statistic of the distribu-
tion number of academic research articles published by the ScienceDirect, Springer,
and IEEE Explore databases within a fixed period from 2000 to 2020. For more
details, Fig.5 depicts the statistic depending on the article’s type published by the
ScienceDirect, Springer, and IEEE databases.
Recently, the concept of smart cities has been extensive research articles on using
ICT to disseminate ubiquitous sensing information. As a result, a lot of traditional
Table 1 Electronic databases used in the review process
Online database No. of published papers/main topics URL
2000–2020 2000–2020 2000–2020
Smart city ITS Smart healthcare
ScienceDirect (Elsevier) 34,410 30,423 21,333 https ://www.scien cedir ect.com/
Springer 22,596 11,1375 11,119 http://link.sprin ger.com/
IEEE Explore 19,140 28,467 3367 http://ieeex plore .ieee.org/

5675
1 3
Towards sustainable smart IoT applications architectural…
Table 2 The summarization of the IoT definitions published in 2005–2018
References Themes Description Year
Kevin Ashton [47] Ubiquitous computing context The potential capacity of the networked devices to propagate their real information about the
physical objects through the web
2005
Atzori etal. [13] QoS-context Combining embedded technologies such as semantic, middleware, RFID and sensor to estab-
lish the service orientation
2010
Atzori etal. [14] Pervasive computing context The IoT incorporates a wide range of technologies and envisages a variety of entities or
artifacts around us that are capable of communicating with each other via specific addressing
schemes and standard communication protocols, and collaborating with their neighbors to
achieve shared objectives
2012
Gubbi etal. [39] Smart environment Interconnecting sensing and actuating devices which allow information to be shared across
platforms through a unified framework, creating a common operating picture for innovative
applications
2013
Whitmore etal. [97] Intelligent environment/ Pervasive computing
context
Build a global network to assist devices with ubiquitous computing and context-awareness. In
addition, it provides an ambient intelligence that would allow everyday objects to understand
their surroundings, communicate with people and make decisions. A world of smart objects
has tremendous potential to boost business processes and people’s lives
2015
Ali etal. [4] Intelligent environment/Ubiquitous computing
context
A dynamic global network infrastructure with self-configuration and interoperable communica-
tion. Additional, it makes everything around us starting from (i.e., machine, devices, mobile
phone and cars) even (cities and roads) are expected to be connected to the internet with an
intelligent behavior as well as taking into account the existence of the kind of autonomy and
privacy
2015
Razzaque etal. [81] Pervasive computing context Part of a pervasive computing system involving an ultra-wide network of items (e.g., Wire-
less Sensor Network (WSN), RFID, M2 M, and Supervisory Control and Data Acquisition
(SCADA)) and the huge number of events that these items will produce spontaneously
2016
Cluster of European
research projects
[48]
QoS/Physical natural context Objects are active participants in industry, knowledge and social processes where they can
engage and communicate with each other and with the environment through the exchange
of environmentally sensitive data and information, while responding autonomously to the
real/physical entities and influencing it by running processes that trigger actions and create
services with/without direct human intervention
2018

5676
Z.H.Ali, H.A.Ali
1 3
services update their behavior using sensor-based devices technique to be smarter
and intelligent. A significant number of studies indicate that the proper management
of the amount of data streaming out of sensors affects services into smart cities.
Teng etal. [90] proposed a novel cost-effective code dissemination model using
mobile vehicles with an opportunistic communication style. This proposed is suit-
able for maximizing coverage of code deployment over the smart city, but it lacks
reliability. In [2], the urban IoT framework based on hybrid routing technique for
supporting a new vision of digital smart cities was introduced. A new methodology
in this framework was studied to overcome the challenges stream out of the amalga-
mation of ICT and IoT such as the huge amount of heterogeneous data, transmission
power consumption, and delay time. This type of methodology is suitable not only
for the real-time applications but also for ensuring the sustainable growth of digital
services and dynamic realization of new updates of cities.
Several research articles on smart cities are closely connected with controlling
excessive data transmission over the Internet to enable power reduction and network
stability. Mora et al. [72] found that providing more elaborate semantic informa-
tion of sensing data is the main issue in smart cities. The amalgamation of cloud
computing and mobile computing techniques was a promising solution to design
a three-layer distributed architecture for intensive urban computing. Although the
amalgamation of cloud computing and mobile computing techniques deals well
with the excessive data produced by sensor devices and offers high storage capa-
bility and computation power, the latency and the network bandwidth-consuming
were clear deficiencies in the proposed architecture. In [8], a real-time application
is proposed with a combination of adaptive routing and privacy-preserving mecha-
nisms to reduce summary vector sizes without compromising minimal QoS level.
By controlling the packet transmission size, the overall performance of the proposed
technique was achieved high throughput and low power consumption rate. However,
the improvement of resource utilization is still required.
Montori etal. [70] suggested a service composition architecture focused on com-
puting context for smart cities called SenSquare. This provided environmental mon-
itoring for the smart city through semantic aware Mobile Crowd Sensing (MCS).
But it was a lack of scalability, a lack of data quality, and users’ privacy is limited.
Seo et al. [83] proposed a new computing model focused on the cloud platforms
for promoting ubiquitous computing techniques. This work described the ubiquitous
computing as referring to the model of computer collaboration in which knowledge
is systematically incorporated into everyday events and things. The context knowl-
edge is performed as an important issue in the suggested solution, according to the
considerable complexities of ubiquitous environments. The adaptation of the activ-
ity of distributed applications with the context deviations is considered in the intro-
duced architecture. The simulation results demonstrated that the introduced model
provided an effective technique for mobile applications. The defects of this model
not only result in high power depletion but also in low rate of reliability.
Distefano etal. [33] demonstrated an analytical delicate framework focused on
stochastic Petri nets to assess QoS indicators of a class of Mobile Crowd Sensing
(MCS). The proposed framework used a MCS application as a simple model to
acquire some specified QoS classes so as to achieve the quantitative analysis and

5677
1 3
Towards sustainable smart IoT applications architectural…
description of the crowd comportment. A real-time case study was used to investi-
gate the effectiveness of the framework in the response time and energy consump-
tion alike. The limitation of this framework is high communication cost. In [28],
a practical framework for benchmarking cities based on the smartness in their
transportation systems is presented. The framework approach includes: formulat-
ing a proper smartness definition, creating a common matrix of smartness indica-
tors for an urban transportation system, calculating smartness indices for different
urban transportation system sub-systems, and taking into account the use of smart
technologies. The proposed framework improved end-to-end communication and
scalability, while the defect of this framework is high rate of power consumption.
Fig. 3 A comprehensive technical taxonomy of IoT applications

5678
Z.H.Ali, H.A.Ali
1 3
In [99], a cloud real-time energy monitoring system using big data technology was
proposed. This proposed system was built upon an environmental ecosystem called
"Hadoop" to improve processing, analyzing, and data storage. The benefits of the
proposed system were focused on the reduction of emission and power consumption
rate as well as providing high scalability. The limitation of this system is to the delay
time must be considered.
In [66], the real
CO2
measuring system in the urban area by sensor networks was
introduced. This system is called "CitySee," and it is used to address several network
issues including the deployment of sensors, data collection, data processing, and net-
work management. The substantial issue highlighted in this study was about the
deployment of sensors with some other related problems such as data representability
Fig. 4 The statistic of online research articles of the "smart city" that are published by ScienceDirect,
Springer, IEEE databases from 2000 to 2020
Fig. 5 The ratio of the distribution of "smart city" articles that are published by ScienceDirect, Springer,
and IEEE database from 2000 to 2020

5679
1 3
Towards sustainable smart IoT applications architectural…
reliability, coverage, and connectivity. Results of the assessment showed the efficiency
of the proposed solution in a real environment. Lingling etal. [62] established an IoT-
based vehicle monitoring system. The proposed device architecture is composed of five
layers: the application, data, network, sensing, and object layers. In order to assess this
architecture, the Intelligent Vehicle Monitoring System (IVMS) was installed in Nan-
jing city, based on the IoT technology that composes video recognition, Global Posi-
tioning System (GPS) and Radio-frequency identification (RFID) to provide driver
identity information and vehicle traffic information. The proposed architecture has
gained the significant advantages in numerous directions including growing into the
dynamic resources management of the road network and improving the management
monitoring of emergency conditions. The main defects of this architecture are that
power consumption and reliability are not considered.
The information acquisition system using four-meter centralized reading was intro-
duced in [105]. This proposed system can read data from household electricity, water,
gas, and heat based on remote meter reading technology. Meanwhile, it can provide the
reading electrical data to each company for five consecutive months. Depending on the
meter, 1 yuan/household per month is charged, and then, each year has a 34,800-yuan
profit. Water, gas, and heat enterprises reduce the workload of manual meter reading.
According to the calculation of 0.5 person/month, the enterprises can reduce around
42,500 yuan cost each year according to the monthly average wage (3000 yuan) of Tai-
yuan. The advantages of this proposed system are communication reliability, resources
utilization, and cost-effective. In [7], the Internet of Things-Based Urban Waste Man-
agement System was presented to adjust waste management issues in smart cities. This
proposed system used Cuckoo Search Optimized Long Short Term Neural Networks
(CLSTRNN) for the analysis of IoT-based waste data. This system allows for increasing
resource utilization and response time. However, the disadvantage of this system is that
the control parameters of the cuckoo search algorithm are kept constant for a certain
duration. Thus, the efficiency of the algorithm is decreased.
The effective management of power consumption is still a focal point in provid-
ing the sustainability of smart cities due to its high ability to avoidance of the network
lifetime ends prematurely. In [102], a decentralized energy-aware scheme for mobility
management of mobile nodes in smart cities was introduced. A hybrid data transmis-
sion capability and low power consumption rate are the main advantages of this study.
However, scalability and reliability are needed to consider. Another article studied the
energy depletion rate by Arora etal. [11]. The power-aware framework was proposed,
which efficiently switches the disk into standby, active and idle states, leading to the
least power consumption. The transmission delay time and power depletion ratio are
reduced, while the communication reliability is still needed to improve.
2.5 Evaluation ofsmart city studies
The evaluation process in this section is conducted based on three tables. Table3
summarizes the crucial aspects of smart environment solutions using real-time data
intelligence, which considers a better way to increase sustainability deployment.
Table 4 explains in a brief the main context of the previous literature studies on

5680
Z.H.Ali, H.A.Ali
1 3
smart city and determines the published year of each proposed architecture and
design in these papers. Moreover, the significant QoS performance metrics that are
required to allow highly successful sustainability in smart cities are presented in
Table5, which includes availability, scalability, reliability, response time, and power
consumption.
2.6 Intelligent transportation systems (ITS): overview
Real-time informative traffic systems, on-demand service mobility, and automation
operations are the long-term motivations for sustainable intelligent transportation.
The ability to create a global vehicular network overview to allow remote monitor-
ing traffic-flow issues with real-time data-driven is a primary goal.
The "Smart Transportation", "Intelligent Transportation", "Intelligent Mobility",
and "Smart Mobility" are the concrete expressions used as substituted words for
the term "ITS" to obtain all possible published papers relevant to this concept. The
search operation was in three engineering databases Elsevier, Springer, and IEEE.
As shown in Fig.6, the peak of published papers was Springer database, and it has
around 11,1375 articles in the duration time from 2000 to 2020. The second place
was the Elsevier database with estimated at approximately 30,423 research articles.
The third was the IEEE Explore database, and it reached 28,467. For more clarifica-
tion, Fig.7 depicts the overall ratio of total published articles according to their type
whether review paper, research article, conference or even book chapter.
Providing adequate transport services with limited financial resources and less
developed infrastructure is the biggest challenge toward ensuring the sustainability
of transportation systems in developing countries. One solution to avoid traffic con-
gestion on the roads and improve the overall use of existing transport infrastructure
is to maximize the information utilization with adjusting the data transmission rate
over the Internet instead of hardware installation. This solution not only achieves
more sustainable and better systems growth but also improves real-time perfor-
mance and decision [6, 29].
Mohan etal. [68] suggested a new methodology to obtain the position of chuck-
hole and resend the information to the central point (server), sensors like Accelerom-
eter, General Packet Radio Services (GPRS), and GPS are used in the smart phone.
When the vehicle node goes into the chuckhole, variation in the accelerometer is
noted and GPS value will be recorded on that particular location. The whole infor-
mation will be sent to the server through GPRS for more analysis. When another
vehicle is moving on that same road, the vehicle will be reported via an application
that there is a chuckhole ahead. If the chuckhole exists for a long time which triggers
congestion on the roads, it will be notified to the construction company for repairing
this road. One benefit of GPS is to determine the position; however, it cannot record
the history of vehicle navigation in real time. In [9], the camera in the smart phone is
used as a sensor to collect data in order to obtain real navigation and assessment for
the objects. This scenario has given the live broadcast of roads to the server, where
the server makes analysis and predict traffic on the roads. This proposed method can
record traffic road history.

5681
1 3
Towards sustainable smart IoT applications architectural…
Supporting the nowcasting technique in order to improve immediate action and
real-time decision making instead of the forecasting technique is necessary to shift
from conventional ITS to smart mobility. Therefore, with this update, delay time
caused by cloud computing is not acceptable with this change. Several computing
technologies are being proposed to this purpose. The infrastructure facilities for
OpenNebula cloud computing supporting resource decentralization are presented in
[82]. This method was dealing with data generated by sensors locally so as to save
network bandwidth but it is limited storage, and therefore, a lot of data are lost. Hsu
etal. [44] introduced a transportation cloud-based elastic service platform using the
virtualization and cluster computation features of cloud computing. This platform
not only provides convenient transport services but also offers cloud-based software
development and storage services. Improving navigation and real energy conserva-
tion are still required.
For using the vehicular resources efficiently, a modern distributed architecture
called Vehicle Fog Computing (VFC) was studied in [43] to optimize the use of
vehicle-as-infrastructure technology. This architecture has high capacity for com-
puting and communicating to support a multitude of vehicles working together.
Lin etal. [60] introduced a semi-Markov decision processes (SMDP) for Vehicular
Cloud Computing (VCC) resource allocation that considers heterogeneous vehicles
and Road Side Units (RSUs). This proposed scenario improves location-awareness
but it is limited storage. On the other hand, Li etal. [59] provided an analytic model
that can estimate the power depletion rate of Edge Cloud-based IoT Platforms for
IoT. This model is composed of several elements such as the IoT infrastructure, the
collecting point to gather sensory data streamed out by sensors, the cloud computing
for processing and storing collected data, and the network layer links the collecting
point and cloud center. The proposed model reduced the network energy overhead.
However, scalability is lack.
From the data-driven point of view, after the amalgamation of vehicular networks
and IoT, the amount of data has become "big" because they are streamed out from
thousands of sensor devices. Surely, the real-time decision on the roads is heavily
related to achieving reliable information dissemination from these data. Several aca-
demic studies on controlling data generated from sensors have appeared, and various
frameworks and architectures have been proposed to improve the overall vehicular
networks. Ansari etal. [10] suggested computing architecture called Cloud Comput-
ing on Cooperative Cars (C4S) is mainly based on Navigation-as-a-Service (NAV-
aaS) for building Cooperative ITS (C-ITS) that ensure interoperability among C-ITS
users of single- and multi-Global Navigation Satellite Systems (GNSS) receivers.
The proposed architecture design supported elastic computing services, enhancing
interoperability, and improves system navigation among vehicles. However, it suf-
fered from communication overhead and time latency.
An efficient continuous event-monitoring system as well as a Two-Phase Event-
monitoring and data-Gathering scheme (TPEG) using fog technology in VANETs
were investigated in [56]. This type of schema was appropriate for the suppression
of excessive uploading and transmission of information. Pereira etal. [77] suggested
a new generic architecture, and a proof-of-concept system for deploying the com-
puting technology has the ability to deal locally with vehicular networks for traffic

5682
Z.H.Ali, H.A.Ali
1 3
Table 3 The list of the literature reviews and other information in smart cities
Research Design/propose Feature Weakness
Teng etal. [90] A smart urban model High scalability
Low power consumption
Enhancing response time
Enhancing the balance between
object tracking precision and
energy usage
Not support reliability
Ahmed etal. [2] A real-time vehicular framework/ hybrid routing algorithm Reliability
Scalable system
Resource utilization
Enhancing accuracy rate
Real-time decision making
High power consumption
Mora etal. [72] A semantic ontology based on cloud computing Availability
High interoperability
Enhancing response time
Resource utilization
End-to-End communication model
Network bandwidth consuming
High power consumption
High computation transmission consumption
Amah etal. [8] An adaptive routing algorithm Secure system
High throughput
Low delay time
Low power consumption
Decreasing the execution time
Resource utilization is lack
Montori etal. [70] A service composition architecture End-to-End communication model
Reliable communication
Enhancing response time
Not support scalability
Seo etal. [83] An intelligent computing architecture Elastic on-demand services
Resource utilization
High transparency
High interoperability
High power consumption
Not support reliability

5683
1 3
Towards sustainable smart IoT applications architectural…
Table 3 (continued)
Research Design/propose Feature Weakness
Distefano etal. [33] An analytical modeling framework Enhancing response time
Low power consumption
High interoperability
Decreasing the execution time
High communication cost
Not support scalability
Not support reliability
Debnath etal. [28] An intelligent practical framework End-to-End communication model
Reliable communication
Scalable system
Resource utilization
Enhancing response time
Not support real-time location awareness
High power consumption
Yang etal. [99] A real-time energy monitoring system End-to-End communication model
Reliable communication
Scalable system
Resource utilization
Emission reduction
Low power consumption rate
Not support real-time location awareness
High response time
Mao etal. [66] An air pollution measuring system Low power consumption
Emission reduction
Not support scalability
Not support availability
Lingling etal. [62] A vehicular monitoring system Enhancing response time
Improving real-time decision mak-
ing process
High interoperability
Increasing the dynamic resources
management of the vehicular
network
High power consumption
Zhang etal. [105] A remote meter reading system Reliability
Scalability
Resource Utilization
Cost-effective
Emission reduction
Data capacity is limited

5684
Z.H.Ali, H.A.Ali
1 3
Table 3 (continued)
Research Design/propose Feature Weakness
Alqahtani etal. [7] An urban Waste Management System Availability
Resource utilization
Emission reduction
The control parameters are constant
Zamanifar etal. [102] An energy-aware scheme for mobility management of mobile
nodes
Availability
Enhancing response time
Low power consumption
Emission reduction
Reliability is lack
Scalability is lack
Arora etal. [11] An energy-aware framework for enhancing the real-time deci-
sion
Enhancing response time
Low power consumption
Emission reduction
Reliability is lack
Scalability is lack
Availability is limited

5685
1 3
Towards sustainable smart IoT applications architectural…
Table 4 The main context of the smart city literature studies
Research Context Year
[90] A new cost-effective code dissemination framework using mobile vehicles with an opportunistic communication style 2019
[2] A new methodology to control urban vehicular networks 2018
[72] A semantic-aware mobile system based on the combination between cloud computing and mobile computing techniques 2019
[8] A real-time adaptive protocol for power reduction 2020
[70] Service composition architecture based on computing context 2017
[83] The context-aware pervasive computing 2016
[33] The QoS-aware computing services 2015
[28] An urban transportation system based on context-aware service composition 2014
[99] An Energy monitoring system using an environmental ecosystem called "Hadoop" to improve big data processing and analyz-
ing
2020
[66] A real-time monitoring system for environmental sustainability 2012
[62] Traffic monitoring and load balancing management 2011
[105] Information acquisition system using four-meter centralized reading are electricity, water, gas, and heat 2019
[7] Waste management system using Cuckoo Search Optimized Long Short Term Neural Networks (CLSTRNN) to the analysis of
IoT-based waste data
2020
[102] A decentralized energy-aware scheme for management huge amount of data generated from mobile nodes 2020
[11] The power-aware prediction-based disk storage framework containing threshold policies 2020

5686
Z.H.Ali, H.A.Ali
1 3
Table 5 The evaluation QoS performance metrics of the smart city studies
References Availability Scalability Reliability Resource
Utilization
Response
Time
Power
Con-
sumption
Teng etal. [90]
✓
✓
✗
✓
✓
✓
Ahmed etal. [2]✗
✓
✗
✓
✓
✗
Mora etal. [72]
✓
✓
✓
✓
✓
✗
Amah etal. [8]
✓
✓
✓
✗
✓
✓
Montori etal. [70]
✓
✗
✓
✗
✓
✗
Seo etal. [83]
✓
✗ ✗
✓
✓
✗
Distefano etal. [33]
✓
✗ ✗ ✗
✓
✓
Debnath [28]✗
✓
✓
✓
✓
✗
Yang etal. [99]
✓
✓
✓
✓
✗
✓
Mao etal. [66]✗ ✗ ✗ ✗
✓
✓
Lingling etal. [62]✗
✓
✗ ✗
✓
✗
Zhang etal. [105]
✓
✓
✓
✓
✓
✓
Alqahtani etal. [7]
✓
✗
✓
✓
✓
✓
Zamanifar etal. [102]✗ ✗ ✗ ✗
✓
✓
Arora etal. [11]✗ ✗ ✗
✓
✓
✓
Fig. 6 The statistic of online research articles that are published under the title "ITS" and "Smart Mobil-
ity" in ScienceDirect, Springer, IEEE databases from 2000 to 2020

5687
1 3
Towards sustainable smart IoT applications architectural…
anomaly detection and travel time estimation is to Fog Cloud Computing (FCC).
This technique not only improves response time but also promotes data storage and
analytic.
Recently, video cameras /video surveillance are being deployed as a harsh envi-
ronment to monitor city street traffic and to provide real-time delivering data. In
[54], Infrastructure to Vehicle Multi-View Pedestrian Detection Database (I2V-
MVPD), which gets synchronized images from an embedded camera installed in
a car and a static camera in the road infrastructure, was introduced to build a his-
torical public dataset. The article also proposed a multi-view pedestrian detection
framework using smart collaborative between vehicles and infrastructure. Easy-to-
use infrastructure and real-time performance are realized. But, the increasing system
scalability is still required.
2.7 Evaluation ofintelligent transportation systems (ITS) studies
The evaluation process in this section is conducted based on three tables. Table6
summarizes all literature studies mentioned in the previous section and highlights
the weaknesses and strengths of each one. Table7 describes the fundamental context
of each literature studies and determines its published year. The most important QoS
parameters include location-awareness, scalability, latency, response time, power
consumption, interoperability, and communication overhead shown in Table 8 in
order to measure the rate of sustainability in ITS. In addition, a comparison between
the most popular computing platforms that are used in the previous researches is
described in Table9.
2.8 Smart healthcare: overview
As the developed world and mature economic community are rapidly rushed to con-
nect literally any apparatus that can be connected to the internet whether at school,
street, home, or workplace, the IoT technology has succeeded to provide new solu-
tions to enhance healthcare systems’ behavior by enabling clinical operational
efficiency and increasing remote patient monitoring capabilities with expanding
telehealth service access. This section introduces a sample of published papers in
Fig. 7 The ratio of the distribution of articles that are published under the title "ITS" and "smart mobil-
ity" by Springer, ScienceDirect, and IEEE databases from 2000 to 2020

5688
Z.H.Ali, H.A.Ali
1 3
the fields of telehealth and data-driven care delivery to enable the transition to a
smart healthcare system. Figures8 and9 depict the distribution of academic reviews
conducted by leading scientific publishers, including ScienceDirect, Springer, and
IEEE. As shown in these figures, in the period from 2000 to 2020, the first place was
the Springer database with a total of 11,119 papers, the second place was Science-
Direct database with 21,333 papers, and the last place was IEEE Xplore database
with 3367 papers.
The process of the transformation in smart healthcare systems has undergone a lot
of academic studies since the early 2000s in an attempt to promote sustainable pro-
duction development by enhancing clinical operational efficiency and maximizing
leverage data-driven utilization. Donaldson etal. [34] described how clinical gov-
ernance as a quality structure underpins and facilitates child-centered, secure, and
high-quality care delivery. It also recognizes the importance of teams, institutions,
and processes in providing the treatment. The paper is also offered a case study that
provides an insight into how clinical governance can be implemented in the care of
children. Moon etal. [71] proposed three different strategies to control the computa-
tional overhead caused by the fingerprint deployment over large scale of telehealth
services. The proposed scenario was overcoming the communication overhead as
well as improving security and privacy issues.
Vecchia et al. [30] provided an intelligent hospital network and supplied basic
facilities in order to improve medical staff/ patient interactions and provided ubiq-
uitous and transparent access to clinical data stored in standard clinical databases.
The utilization of healthcare resources is investigated. Moreover, the cost is reduced
and scalability is improved. An accurate and efficient monitoring system via reliable
Missing Tag Detection (MTD) protocol for studying a large set of WISP tags and
identifying the missing ones is proposed in [80]. The data reliability and security
are the mean benefits of this proposed protocol. Another type of healthcare monitor-
ing system is demonstrated in [53] to track the health status regularly, and irregular
values are reported to both the transportation authority and the healthcare provider.
Although the proposed system improved end-to-end communication over the net-
work, it suffered from increasing the power consumption ratio.
Deen [53] used several low-cost, noninvasive, user-friendly sensing and actuating
systems to maximize the utilization of ICT and improve the quality of the health-
care. The proposed system has many advantages such as cost-effectively, high inter-
operability, low energy consumption, and high response time. The limitation is the
area scale. A remote mobile physician monitoring IoT device was proposed in [31].
By using intelligent nodes, this paper suggested a model of human activity and
physiological parameters. The proposed system is introduced by storing the essen-
tial emergency details in the hospital database to establish reliable emergency alarm
systems. The major benefit of this study is to increase the speed and precision of the
calculation of physiological factors and also to use low power consumption instru-
ments. This study’s weaknesses are that the expense has not been considered.
Fafoutis etal. [36] built a new framework for the establishment of a monitor-
ing system for the diagnosis and prevention of chronic medical conditions such as
diabetes, obesity, and depression. The question of energy constraints is regarded as
a consequence of the related costs of the associated repair or replacement of the

5689
1 3
Towards sustainable smart IoT applications architectural…
Table 6 The list of the literature studies and other information in ITS
Research Design/proposed Feature Weakness
Mohan etal. [68] A smart navigation system Scalable system
Location-awareness
Accelerating response time
Reducing delay rate
Navigation and decision according to history of
data are must be considered
The energy usage rate is risen
High communication cast
Angin etal. [9] A real-time vehicular monitoring and navigation
system
Improving real time
Enhancing response time
Recording road history
Scalable system
The energy usage rate is risen
High communication overhead
High network bandwidth usage
Ryden etal. [82] The vehicular system based on cloud computing Real-time decision making
Vehicular infrastructure management
Supporting resource decentralization
Reducing the network bandwidth usage
High response time
Costly system
Hsu etal. [44] A transportation cloud-based service platform Enabling the cloud elastic and cluster comput-
ing services
Enabling virtualization technique
Scalable system
The real energy conservation is still required
High network bandwidth usage
Increasing the amount of data transmission power
Location-awareness is lack
The response time is relatively high
Hou etal. [43] A new distributed architecture called "Vehicle
Fog Computing"
Enhancing response time
Low delay time
Lower bandwidth costs
Increasing real-time decision making
Reducing communication and computational
process over the network
Improving location-awareness
High interoperability
The potential storage needs to improved
Costly system
Lin etal. [60] A semi-Markov decision processes for vehicular
cloud computing
Vehicular resources efficiently
Infrastructure management
Accelerating response time
Reducing delay rate
Improving location-awareness
The storage is limited

5690
Z.H.Ali, H.A.Ali
1 3
Table 6 (continued)
Research Design/proposed Feature Weakness
Li etal. [59] A real-time power estimation system Enhancing power depletion ratio
On-demand elastic computing services
Communication overhead
Location-awareness is lack
Scalability is lack
The response time is relatively high
Costly system
Ansari etal. [10] A dynamic data-driven approach based on cloud
computing technology
High interoperability
Scalable system
Enabling cloud elastic services
High power consumption rate
High network bandwidth usage
Location-awareness is lack
Lai etal. [56] A dynamic event-driven approach
Monitoring and tracking system
Enhancing response time
Low delay time
Lower bandwidth costs
Increasing real-time decision making
Reducing communication and computation
process over the network
Improving location-awareness
High interoperability
The potential storage needs to improved
Pereira etal. [77] A proof-of-concept system for deploying the
computing technology
A geographically computational system
Enhancing response time
Low delay time
lower bandwidth costs
Increasing real-time decision making
Reducing communication and computational
process over the network
The storage capacity is limited
Khalifa etal. [54] An Infrastructure to Vehicle Multi-View Pedes-
trian Detection Database (I2V-MVPD)
A multi-view pedestrian detection framework
using smart collaborative between mobile
vehicles and road infrastructure
High Interoperability
Resource Utilization
Increasing real-time decision making
Improving location-awareness
High interoperability
The potential storage needs to improved
The power consumption needs to improved
Highly cost in communication and computational
process over the network

5691
1 3
Towards sustainable smart IoT applications architectural…
batteries in the wearable devices. In this article, only battery-powered solutions are
taken into account. Also, this proposed system has the ability to enhance data relia-
bility, and therefore, it can decrease network bandwidth usage. The limitation of this
framework is that the response time is not enough considered. Aziz etal. [15] pro-
vided an appropriate real rime healthcare monitoring and tracking framework. The
software will track, track, manage patients, and promote their healthcare, allowing
appropriate medical services to be provided at the right time. Using specific sensors,
the data will be collected and compared to a configurable microcontroller thresh-
old established by a specialist doctor who follows the patient; in any case, a short
message service (SMS) will be sent to the Doctor’s mobile number along with the
calculated values via GSM module. The paper illustrated the viability of implement-
ing an integrated end-to-end smart health system that responds to the real require-
ments of health system design by taking into account broader critical human health
parameters.
A miniature wearable cardiopulmonary monitoring system named "Smart Chest
Strap" was presented in [106]. This system composed of an elastic band worn
around the user’s chest with integrated sensors, an acquisition unit for physiological
signals, and a mobile phone. Moreover, it provides sampling, digitization, storage,
and simultaneous transmission to a mobile phone through Bluetooth of physiologi-
cal signals including electrocardiogram, respiratory inductance plethysmograph, and
accelerations (ACC). The results showed the validity of the Smart Chest Strap dur-
ing various strength exercises to calculate heart rate (HR), breathing rate (BR), and
ACC. The limitations of this monitoring system are that power consumption and
security are must be considered. Zulkifli etal. [109] investigated a new design of a
health monitoring system called "Mooble" (Monitoring for Better Life Experience)
to track the condition of patient health and to prevent diseases as soon as possible.
This proposed design consists of three subsystems: web application, database, and
API design, and mobile application on the android platform. It enhances end-to-end
communication, power consumption, and response time. However, it has the limita-
tion of cost.
The architecture of a cloud-based u-healthcare network with a QoS-guaranteed
mobile health service was proposed in [27]. This type of architecture is suitable
for applications that need reliable data transmission and broadband communica-
tion infrastructure. However, the achievement of high response time and low power
consumption is still required. The effectiveness of management of the amount of
data generated by wearable medical devices has relied not only on computing plat-
forms and routing mechanisms but also on privacy and security roles. In [35], the
computing architecture of a fog-based middleware for Internet of Healthcare Things
(IoHT)-based cloud healthcare services was presented. This proposed design is suit-
able for sensitive medical applications that require all data exchange to be accom-
plished under full secure channels. High response time with low power consumption
is also tangible benefits in this direction.
A dynamic and interoperable framework of communication (DICF) was imple-
mented in [19] to govern the operation of wearable healthcare devices. As part of
a smart healthcare tracking system, the framework is responsible for monitoring,
decision-making, and managing the functionalities and running time of wearable

5692
Z.H.Ali, H.A.Ali
1 3
Table 7 The main context of the ITS literature studies
Research Context Year
[68] Applying on real prototype and the data are collected via GPS 2008
[9] Applying on real prototype and using camera in the smart phone as sensor to obtain real navigation 2010
[82] Using resource-efficient cloud computing 2014
[44] A cloud computing scenario supporting the virtualization and cluster computation features 2015
[43] Using resource-efficient fog computing 2016
[60] Applying vehicular cloud computing approach 2017
[10] A cloud computing scenario for providing cooperative transport architecture 2018
[59] A simple power analysis scenario base on edge computing 2018
[56] An efficient continuous event-monitoring scenario 2018
[77] A fog computing approach for establishing normal traffic detection and travel time estimation 2019
[54] A multi-view pedestrian detection framework using smart collaborative vehicular units; this framework supports Infrastructure-
as-a-service concept
2020

5693
1 3
Towards sustainable smart IoT applications architectural…
sensors (WS). Decision-making based on classification and regression helps predict
incidents and define emergency intervals to improve the performance of wearable
sensor operations. DICF harmonizes WS’s data collection, event identification and
interpretation, and communication functionality with the decision-making system
to maximize the advantages of a sensor-dependent personal healthcare system. This
proposed framework not only improves response time and delay but also achieves
reliability.
Research in smart healthcare is also closely connected with research in patient-
oriented medical data analysis, such as in [76]. This article has embraced the idea
of transformation in oriented healthcare society by improving system capabilities in
medical data analysis using Deep Neural Networks (DNN) and the swift of medical
wearable, thus combining the power of high response time and accuracy. Patan etal.
[76] proposed a holistic Deep Neural Network-driven IoT smart healthcare method,
namely Grey Filter Bayesian Convolution Neural Network (GFB-CNN) based on
real-time analytics. This proposed design increases the effectiveness of medical data
analysis by reducing communication overhead and improving accuracy. Although
the use of a cloud server with proposed architecture affects the overall rate of the
response time, it may need to use another computing platform close as possible to
the patient for achieving concrete results in the area of data analytic.
Another study developed an energy-efficient transmission protocol that allows
healthcare systems to maximize the utilization of the most appropriate routing
metrics. This proposed protocol has successfully supported both energy utilization
and network lifetime; however, the scalability is lack with this protocol [79]. Other
studies investigated an efficient multihop routing protocol that controls energy-
constrained wireless networks. In [87], an energy-efficient distributed collabora-
tion mechanism called Voronoi-based collaboration (DVOC) is proposed based on
an Energy-aware Dual-path GR (EDGR) protocol. This mechanism develops the
Table 8 The evaluation QoS performance metrics of the ITS studies
Reference QoS parameters
Location-
aware-
ness
Scalability Latency Response time Power
consump-
tion
Inter-
oper-
ability
Commu-
nication
overhead
Mohan etal. [68]
✓
✓
✓
✓
✗ ✗ ✗
Angin, etal. [9]
✓
✓
✓
✓
✗
✓
✗
Ryden etal. [82]✗ ✗
✓
✗ ✗
✓
✗
Hsu etal. [44]✗
✓
✗ ✗ ✗
✓
✗
Hou etal. [43]
✓
✓
✓
✓
✓
✓
✓
Lin etal. [60]
✓
✗
✓
✓
✓
✓
✓
Li etal. [59]✗ ✗
✓
✓
✓
✓
✓
Ansari etal. [10]✗
✓
✗ ✗ ✗
✓
✗
Lai etal. [56]
✓
✓
✓
✓
✓
✓
✓
Pereira etal. [77]
✓
✓
✓
✓
✓
✓
✓
Khalifa etal. [54]
✓
✗
✓
✓
✗
✓
✗

5694
Z.H.Ali, H.A.Ali
1 3
overall performance of healthcare systems in the following respects: (i) minimizing
the delivery delay time, (ii) reducing power consumption rate, and (iii) enhancing
location-awareness.
2.9 Evaluation ofsmart healthcare studies
The evaluation process in this section is conducted based on three tables. Table10
lists all literature studies mentioned in the previous section and highlights the weak-
nesses and strengths of each one. Table11 describes the fundamental context of
each literature studies and determines its published year. The most important QoS
performance metrics include availability, response time, reliability, scalability,
Table 9 Comparative study of Cloud Computing (CC), Mobile Computing (MC), Vehicular Computing
(VC), and Fog Computing (FC) [6]
Characteristic CC MC VC FC
On-demand elastic [32, 45] Support Possible Possible Support
Pay-as-you-go model [43, 45] Support Support Support Support
Mobility of cloud [45, 104] N/A A N/A A
Location awareness [45, 104] N/A A N/A A
Data processing [25, 43, 45] Pay-as-you-go N/A Medium Medium
Data storage [25, 43, 45, 104] Pay-as-you-go N/A Medium Medium
Computational resources [25,
43, 45]
High N/A Medium Medium
Latency and delay time [32, 45,
100, 104]
High High Medium Very low
Network bandwidth [45, 96,
104]
High Medium Medium Very low
Power consumption [43, 100,
104]
High High Medium Very low
Battery limitation [43, 100, 104] N/A A N/A A
Real-time applications [25, 43,
45, 104]
Possible N/A Possible Support
Ubiquitous services [25, 43, 45,
104]
High A A A
Network architecture [25, 43, 45] Client-server Client-server Peer2Peer (P2P)
or Client-
server
P2P or Client-server
Geographical distribution arch.
[25, 43, 45]
N/A N/A N/A High
Management model [25, 43, 45] Centralized Centralized Decentralized Decentralized
Decentralized data analytic [43,
45]
N/A N/A N/A Support
Computing offloading [43, 45] A A A A
Mobile resources [25, 37, 43,
45, 85]
N/A Support Support Support
Decision-making [32, 43, 45, 96] Remotely Remotely Remotely Local

5695
1 3
Towards sustainable smart IoT applications architectural…
power consumption, security, and system cost shown in Table12 in order to meas-
ure the rate of sustainability in smart healthcare operations. According to the papers
mentioned before in this section, two QoS efficiency indicators have been heavily
influenced by providing a successful healthcare system: response time and data
reliability.
2.10 General aspects IoT applications: overview
The rapid growth of the concept of IoT has gained great significant attention
due to its capability to attract a lot of IT industries and academic fields. Surely,
the rapid deployment of intelligent things plays as an enabler in providing
Fig. 8 The statistic of online research articles of the "Smart Healthcare" that are published in ScienceDi-
rect, Springer, IEEE databases from 2000 to 2020
Fig. 9 The ratio of the distribution number of published paper under the title "Smart Healthcare" from
2000 to 2020 in the international databases of ScienceDirect, Springer, and IEEE

5696
Z.H.Ali, H.A.Ali
1 3
standardization collaborative frameworks that can bridge the gap between IT
domain interaction and industrial platform unification. IEEE P2413 is an out-
growth of a multi-year series of IoT Standards workshops, which is responsible
for putting the standard roles for smart applications such as smart healthcare,
smart home, smart agriculture, intelligent transportation, and so on. The goal of
IEEE P2413 is to build an architectural framework for IoT applications taking
into account the QoS standardization such as transparency and interoperability,
resource utilization, power consumption, system compatibility, privacy and secu-
rity, and functional exchangeability to achieve sustainable applications [63]. In
this section, various aspects of recent smart applications and their QoS needs are
discussed. The classification of this section is based on five themes, namely the
intelligent environmental applications, IT industrial applications, utility services,
commercial IoT applications, and smart agriculture environment.
From the environmental applications point of view, the IoT environmental sur-
veillance and security systems based on the complexity of communication net-
works have studied in [103]. It is important to formulate appropriate and practical
transmission policies because of the various communication costs of transmission
devices in the various communication networks. The benefits of this system are to
reduce the workload of communication by achieving high reliability and by using
the suggested system to reduce energy consumption. The drawback is that there is
no support for scalability, yet it appears to be an important problem for monitor-
ing and protecting the environment. Li etal. [57] suggested an digital IoT moni-
toring system to control environmental conditions, for instance the temperature,
humidity,
CO2
and
NH3
based on a WSN for henhouses. In this paper, it is stated
that most of the concentrations in earlier research are in the production of sys-
tems where the efficiency of wireless data transmission is not taken into account.
According to the loss recovery method, a wireless transport protocol is proposed
as a solution to this problem. The key advantage of this study is the ingenuity in
offering an IoT-henhouse control system that aims to increase the efficiency of
data transmission rate. The biggest weakness of this study is the measurement
of energy consumption which is very limited. Transformation in collaborative
autonomous systems is being deployed to enhance IT industry. A novel perspec-
tive on the Internet of Robotic Things (IoRT) for improving interoperability and
trustworthiness capability was investigated in [92]; however, reliability and delay
time are need to be considered.
From the IT industrial applications point of view, a QoS sensitivity schedul-
ing algorithm for the IoT post was provided by Abdullah etal. [1]. Messages in
this paper can be divided into (i) messages about high precedence that are urgent
and (ii) messages about best effort that are urgent non-duty. The purpose of this
proposed algorithm is providing an improved QoS-aware scheduling and routing
algorithm in order to get rid energy consumption. A simulation to demonstrate
the efficiency of the proposed solution was implemented in Matlab. The benefits
of this algorithm can be considered as reducing latency and saving energy deple-
tion. The weakness of the proposed algorithm is the use of restricted types of
scheduling approaches within the similar service variety. A content-based cross-
layer scheduling approach known as CONCISE for IoT applications was indicated

5697
1 3
Towards sustainable smart IoT applications architectural…
in [52] to provide a new model for managing and centrally collecting data via
Time Synchronized Channel Hop-ping (TSCH) scheduling.
Ali etal. [16] introduced a dynamic QoS Provisioning Framework (QoPF) for
the oriented service using Backtracking Search Optimization (BSO) algorithm.
The design of QoPF aimed to increase the efficiency of composite services in the
IoT application layer; this design not only improves service reliability but also
decreases estimated computing time. The QoPF is composed of three fundamen-
tal stages respectively the preparation, computation, and validation stages. In the
QoPF, the direct link to the sensor layer is established for collecting data pro-
duced by sensors and storing them in a queue. For further review and analysis, all
received requests will be sent to the next level. The next procedure in the QoPF
is dedicated to determining the optimum composition of each service through
the distinction of service states, which involves three different services includ-
ing complete, open, and supple services. To complete distinguish from incom-
ing requests and to identify a better customer service mode (meaning better IoT
service composition patterns), three types of service branches are also used as
follows: series, circulation, and branch services. The advantage of this proposed
framework is to improve delay time and reliability. The disadvantage of this pro-
posal is that power consumption and scalability are not considered.
From the utility services point of view, a new scheme that enables the commu-
nication between IoT devices with less human interference through a standardized
communication architecture was suggested by Iqbal etal. [49]. The architecture in
question consists of four phases. The first phase is about the discovery and detec-
tion of electrical appliances in a smart home or smart building. The second phase
is about the installation of sensors. The third phase is about the operation of the
proposed load balancing on appliances and sensors. The fourth phase is about data
analytic to optimize the use of home and electrical appliances. The scheme under
consideration is evaluated on actual electronic devices, and the energy consumption
is reported using the Electronic System Sleep Scheduling Algorithm (EDSA) sug-
gested. The results indicate that in a heterogeneous setting, the proposed architecture
is better than other state-of-the-art techniques. Using Hadoop Ecosystem, the data
are also processed to maximize efficiency and reduce the time needed to process
the real-time data. The benefits of this proposed architecture are reducing power
consumption and enhancing response time. The weakness of this architecture is in
the cost. In [95], an IoT-based energy consumption control and saving device for
intelligent construction was introduced. The system includes the normal three lay-
ers of sensors for all of the building’s subsystems. The size is confined to local area
network, and fault detection, energy management, and device control are monitored
by the application layer. The study’s pitfall is a lack of effective application of test
cases.
In energy-aware composition, a new multi-cloud IoT service composition algo-
rithm named (E2C2) to enhance IoT composite services and reduce power consump-
tion rate was presented by Baker etal. [17]. The considered E2C2 has been strength-
ened by the method of regional alignment used to identify the best IoT composite
service approach based on the transitional relationships between the customer and
the data centers. The superiority of the considered model has been demonstrated by

5698
Z.H.Ali, H.A.Ali
1 3
Table 10 A list of the literature studies and other information in smart healthcare
Reference Design/Propose Features Weakness
Donaldson etal. [34] A clinical governance framework Improving security and privacy issues
Improving QoS-aware for e-health
Scalable system
Low energy usage rate
The reliability is must be considered
Moon etal. [71] The management and control healthcare
system
Improving scalability
Enhancing the use of power consumption rate
Improving security and privacy issues
The reliability is must be considered.
Costly system
Della Vecchia etal. [30] An intelligent hospital network Maximizing resources utilization
Improving resources management
Scalable system
Cost-effective
Security needs to improve
Power consumption needs to reduce
Rahman etal. [80] The management and monitoring tracking
system
Improving security issue
Accelerating response time
Data is reliable
Availability needs to enhanced
The power depletion is still needed to enhance
Kavitha etal. [53] A tracking healthcare system Increasing system reliability High power consumption
The data is not secure enough
Costly system
Deen etal. [29] The monitoring and tracking healthcare
system
High interoperability
High transparency
Low power consumption
Cost-effective
Low security and privacy
Ding etal. [31] A mobile physician monitoring system Improving reliability
Accelerating response time
Increasing accuracy and precision
Decreasing energy consumption rate
Security and privacy are not considered
Costly system
Fafoutis etal. [36] An adaptive healthcare monitoring system Low power depletion
Decreasing the rate of power transmission
Improving data reliability
The response time needs to be considered

5699
1 3
Towards sustainable smart IoT applications architectural…
Table 10 (continued)
Reference Design/Propose Features Weakness
Aziz etal. [15] A real-time healthcare monitoring and object
tracking framework
Accelerating response time
Reducing delay rate
Providing End-to-End communication mode
Improving service orientation concept
The security and privacy should be considered
Costly system
Zheng etal. [106] A smart monitoring system called "Smart
Chest Strap"
Increasing the response time for aphysiologi-
cal signals
Increasing accuracy and precision
Enabling the cloud elastic services
High power consumption
The data is not secure enough
Zulkifli etal. [109] The tracking healthcare condition system
called "Mooble" or Monitoring for Better
Life Experience
Easy-to-use
Reliable system
High response time
Low power consumption
Costly system
Chung etal. [27] A cloud-based u-healthcare architecture with a
QoS-guaranteed mobile health service
Easy-to-use
Reliable system
Scalable system
Resource utilization
High power consumption
Low response time
Low location-awareness
Elmisery etal. [35] A fog-based middleware for Internet of
Healthcare Things (IoHT) based cloud
healthcare services
High response time
Scalable system
Reliable system
High network security
Low power consumption
High degree of complexity
Costly system
Baskar etal. [19] A dynamic and interoperable framework of
communication called "DICF"
Improving system transparency
Improving reliability
Scalable system
Low security and privacy
High energy depletion
Costly system
Patan etal. [76] The Grey Filter Bayesian Convolution Neural
Network (GFB-CNN) method
Improving system transparency
Improving reliability
Scalable system
Location awareness is low
Response time is still needs to improve
High energy depletion
Costly system

5700
Z.H.Ali, H.A.Ali
1 3
Table 10 (continued)
Reference Design/Propose Features Weakness
Qureshi etal. [79] An Interference Aware Energy Efficient Trans-
mission Protocol (IEETP)
Improving system transparency
Improving resource utilization
Location awareness is high
Improving reliability
low power consumption
Response time is still needs to improve
Scalability must be considered
Security is lack
Sridhar etal. [87] An energy-efficient distributed collaboration
Protocol
Improving system transparency
Location awareness is high
Improving reliability
Low power consumption
Response time is still needs to improve
Scalability must be considered
Security is lack

5701
1 3
Towards sustainable smart IoT applications architectural…
Table 11 The main context of the smart healthcare literature studies
Research Context Year
[34] In the care of children, the concept of clinical governance as a quality framework can be applied to provide a clinical system 2003
[71] Using VLSI technology to provide a new system supporting two-tier authentication protocol 2009
[30] Supporting the pervasive computing context to offer elastic services. Resource–efficient cloud computing computing 2012
[80] Maximizing system reliability. Adaptive security management 2013
[53] Monitor multiple environmental factors of smart healthcare 2014
[29] A QoS perspective and utilizing ICT 2015
[31] Human activity and physiological monitoring 2015
[36] IoT monitoring system for the diagnosis and prevention of chronic medical conditions 2016
[15] Real-time Health monitoring system 2016
[106] Tracking system to record the human physiological signals 2017
[109] E-health monitoring system. Energy Efficient system 2018
[27] A QoS-guaranteed cloud health service 2019
[35] Enhancing security of medical data transmission using fog platform 2019
[19] A dynamic and interoperable IoT framework 2020
[76] Patient-oriented architecture for using deep neural network 2020
[79] Maximizing energy utilization and network lifetime in Wireless Body Area Networks (WBAN) using an interference aware
energy efficient transmission protocol
2020
[87] Applying a collaborative routing mechanism to enhance network lifetime and power consumption 2020

5702
Z.H.Ali, H.A.Ali
1 3
contrasting it with current alternative algorithms such as server, base cloud, smart
cloud, and COM2 in the experimental results. The advantages of this model are that
decreasing the rate of power consumption and saving network bandwidth via han-
dling the amount of data transmission. The cost is the drawback of this considered
model. Zhou et al. [107] introduced a service discovery chain and a new energy
management selection process in WSNs. This method categorized and identified the
sensors based on their functionality and health to choose better composite services.
The experimental results indicated that the validation with the service composition
of the proposed Genetic Algorithm (GA) is minimized energy consumption. The
downsides of this proposed system are that GA’s iteration process is fairly time-con-
suming, besides the response time and delay are high.
Developed smart grid applications are mainly relied on multiple routing and data
transferring services for efficient power management, distribution, pricing, and on-
demand services. In [78], a new distributed Software-Defined Networking (SDN)
approach is proposed to address the scalability and robustness issues and improve
the overall energy efficiency in smart grids systems. This approach is a better soft-
ware solution to meet composite network needs such as frequent fragmentation,
fast dynamic topology, high amount of data transmission, and extensive power
Table 12 The evaluation QoS performance metrics of the smart healthcare studies
Reference QoS parameters
Availability Response time Reliability Scalability Power
Con-
sumption
Security Cost
Donaldson etal.
[34]
✗
✓
✗
✓
✓
✓
✗
Moon etal. [71]✗
✓
✗
✓
✓
✓
✗
Della Vecchia etal.
[30]
✓
✗ ✗
✓
✗ ✗
✓
Rahman etal. [80]✗
✓
✓
✓
✗
✓
✗
Kavitha etal. [53]✗ ✗
✓
✗ ✗ ✗ ✗
Jamal Deen [29]✗
✓
✗ ✗
✓
✗
✓
Ding etal. [31]✗
✓
✗ ✗
✓
✗ ✗
Fafoutis etal. [36]✗ ✗
✓
✗
✓
✗
✓
Aziz etal. [15]
✓
✗
✓
✗ ✗ ✗ ✗
Zheng etal. [106]✗
✓
✓
✓
✗ ✗
✓
Zulkifli etal. [109]
✓
✓
✓
✓
✓
✗ ✗
Chung etal. [27]
✓
✗
✓
✓
✗ ✗ ✗
Elmisery etal. [35]
✓
✓
✓
✓
✓
✓
✗
Baskar etal. [19]✗
✓
✓
✓
✗ ✗ ✗
Patan etal. [76]
✓
✗
✓
✓
✗ ✗ ✗
Qureshi etal. [79]
✓
✗
✓
✗
✓
✗
✓
Sridhar etal. [87]✗
✓
✓
✗
✓
✗
✓

5703
1 3
Towards sustainable smart IoT applications architectural…
consumption. Another study to smart grid growth by power consumption manage-
ment is presented in [38]. In this study, the fault-tolerance features of a software
framework called Resilient Information Architecture Platform for Smart Grid
(RIAPSG) were proposed. This proposed framework presents the fault management
subsystem can improve real-time schedule resources and power depletion rate.
In the commercial IoT applications, with respect to the cross-modified strategy,
the best IoT composite services are dynamically chosen. Huo etal. [46] studied a
new proposed algorithm in order to provide adequate response time and high accu-
racy compared to other techniques such as the genetic algorithms. Regarding the
IP case study, the quality variables are considered to obtain the optimal response
for the proposed approach to service selection and composition. The experimental
results showed that this proposed algorithm has an acceptable response time and
low mean convergence iteration. The drawback of this study is that no attention has
been given to serial optimization of the task nodes to test the proposed process. The
new classification and design of the QoS-aware web service based on a multi-agent
approach in the IoT application layer were studied in [89]. The conceptual approach
to service composition is focused on a context-aware approach to distributed opti-
mization of service architecture. For the prevention of the coordination overheads,
this article proposed a multi-agent method distributed QoS-aware to orchestrate
and composed the execution of multiple services. The authors also used the simple
additive weighting (SAW) approach to perform the QoS factor efficiency in select-
ing and composing existing IoT services. The advantage here is to use the principle
of graphs to demonstrate an acceptable response time and better composite service
relative to the other methods to composition. The downside in this system is that the
scalability is not considered.
In the smart agriculture environment, the use of smart phones to acquire different
types of agricultural information was proposed in [40]. This article was a demonstra-
tion of advanced farming techniques through videos to increase awareness of preci-
sion agriculture adoption. The advantage of this proposed system is improving real-
time performance. The weakness is the energy consumption not considered. Sinha
etal. [86] solved several imperative agricultural issues through using a user-centric
IoT system. The proposed system helped farmers to monitor their agricultural fields
in real time and provided meticulous feedback about good quality crops for harvest.
Moreover, the conceptual design is optimized for the food supply chain in a way
that allows producers to increase their overall profit on the goods sold. The appli-
cability of the proposed architecture is tested using cases involving multiple uses
covering the various aspects of the farming process. The advantage of this system is
to improve the response time and reliability. The complementarity of technologies
between IoT, image processing, and data fusion accelerates business transformation
outcomes. A smart agricultural architecture called "Hydra fusion architecture" using
the data fusion technique was proposed in [91] for overcoming sensor limitations
and maximizing resource utilization. Moreover, two applications are developed to
monitor experimental cultures of precocious-dwarf cashew and coconut trees, aim-
ing at smart water management. Increasing sensor accuracy, accelerating decision
making, and cost-effectiveness are the main advantages of the proposed architecture.

5704
Z.H.Ali, H.A.Ali
1 3
2.11 Evaluation ofgeneral aspects IoT applications
The evaluation process in this section is conducted based on three tables. Table13
summarizes the general aspects of IoT applications by highlighting the upside and
downside of each application. The main context of each proposed framework/
design, specifying the published year, is provided in Table14. The most important
QoS metrics include availability, response time, reliability, scalability, delay time,
power consumption, and system cost shown in Table15 in order to measure the rate
of sustainability in different IoT applications.
3 Sustainable IoT applications: challenges andbenets
Based on the information mentioned in Sect.2, the sustainability in IoT applications
can be defined as maximizing resource utilization while ensuring that the service
quality is achieved. As a result, the challenges of IoT applications can break down
into two perspectives: (i) service performance challenges that included under QoS
challenges and (ii) network architecture challenges that included under ubiquitous
computing challenges. The discussion about the fundamental challenges and differ-
ent IoT design issues will be presented in the next section. Figure 10 depicts the
related challenges of sustainable IoT applications.
3.1 Sustainable IoT applications—from QoS perspective: challenges andbenefits
During the last few years after the diffusion of IoT, a lot of IT services replaced
their traditional behavior with the composition of IoT services to meet transforma-
tive business needs and to extend industrial networks. The IoT composite services
are bringing developed IoT services for satisfying the complexity of user’s require-
ments and without compromising the service quality and service cost [16]. As a
result, studying the QoS metrics that achieve service interoperability and provide a
service-oriented environment has become a crucial point. In this section, we high-
light leverage end-to-end QoS solutions as optimizing uniform framework with col-
laboration, identification, and standard communication protocols, data management
for realizing high interoperability and lower network bandwidth, and lower power
consumption. These levels are included under the goal of enhancing the overall ser-
vice performance.
1. Standardized Design for Collaboration there are three fundamental architectural
approaches to build collaborate and trust IoT model: (i) event-based approach
uses publish/subscribe paradigm, (ii) service-oriented approach targets service
discover such as reliability and reusability, and (iii) application-driven approach
is an application driven by an underlying process [21]. As shown in Fig.11, the
IoT framework consists of three main layers from the top to bottom: the applica-

5705
1 3
Towards sustainable smart IoT applications architectural…
tion layer, network layer, and sensor layer as well as an implicit layer to avoid the
heterogeneous issues called the "middleware" layer. This layer provides a set of
middleware solutions and it butted between the application and network layer.
• Application layer: this layer constructs an interface to promote user interac-
tion with accessible applications such as healthcare and smart transport ser-
vices. The goal of this layer is to ensure delivery of IoT services with low
latency and high throughput (service reliable). To this end, the QoS param-
eters in this layer can be regarded as follows: (i) end-to-end latency, (ii) scal-
ability, (iii) availability, (iv) reliability, (v) throughput, and (vi) cost-effective-
ness.
• Network layer: this layer includes many different communication technologies
and protocols (e.g., IEEE 802.15, Bluetooth Low Energy (BLE), ZigBee, and
RFID). Besides sundry traffic mechanisms are classified into two groups: (i)
throughput and delay tolerant elastic traffic and (ii) the bandwidth and delay
sensitive inelastic (real-time) traffic [51]. Therefore, the QoS parameters in
this layer can be listed as follows: (i) communication cost, (ii) traffic cost (iii)
communication overhead, (iv) latency time, (v) bandwidth, and (vi) through-
put.
• Sensor layer: the layer involves a myriad of the different sensing entities.
The entities may be sensors, RFID tags, and any intelligent devices have
communication capabilities. Sensor layer constructs a straight connec-
tion with IoT infrastructure. To this end, it should strive to perform sev-
eral tasks, for instance resource utilization, the management of the physical
entities, and transmitting/reacting to events that produced by sensors. The
QoS parameters in this layer can be listed as follows: (i) interoperability,
(ii) power consumption, (iii) resources utilization, (iv) averting overload
effort on the connected devices, and (v) redundancy.
The IoT network involves a huge amount of heterogeneous things with vari-
ous resource-constraints. To avoid heterogeneity increasingly over the network
and meet QoS performance metrics, several approaches must be considered as
follows:
I Middleware-based IoT Architecture the IoT is an ultra-large-scale networks with
a myriad of components (e.g., M2M, Sensor Networks, RFID, and SCADA),
and the diversity of applications with sundry demands. All of them need to an
abstracted way can handle this diversity and offer an interoperable communica-
tion. The middleware technique plays an essential role in this context. It can be
defined as a software layer between the network layer and application layer. This
technique has the ability to support interoperability between the diverse things
and applications. Recently, a lot of enterprises have been developed their soft-
ware components to support the IoT middleware solutions.

5706
Z.H.Ali, H.A.Ali
1 3
According to [81], there are several requirements of the IoT middleware,
which can be grouped into two units: (i) middleware service requirements
and (ii) middleware architectural design requirements. The middleware
service requirements itself involve functional requirements which are con-
cerned with the management issues (e.g., resource discovery, resource
management, events management, code management, and data manage-
ment) and non-functional requirements which are concerned with the
service quality issues (e.g., availability, scalability, reliability, popular-
ity, timeliness, and security and privacy). The middleware architectural
design requirements include a number of parameters that are related to the
qualification of the IoT architectural design in order to support the diver-
sity of "things" (e.g., programming abstraction, interoperable, service-
based (flexibility), adaptive, context-aware, autonomous, distributed).
II Semantic-based IoT Architecture the semantic level M2M interoperability is
considered as the recent method to improve the IoT middleware and to realize
a high level of interoperability between the connected-things. To this end, there
is a set of projects use semantic technologies such as HYDRA middleware (also
called LinkSmart) [94], COCOA [69], and TCE [67]. These projects have the
same target is to "the enhancement of the service-oriented architecture within
pervasive computing applications on top of heterogeneous devices and Sensor
Networks (SNs)" [55]. On the other hand, the IoT paradigm is generally used
to publish the information/knowledge about the physical things over the web.
Therefore, it needs to build robust context-aware to represent both the informa-
tion and the IoT resources; actually, this process is done by using the semantic
web knowledge representation technologies like Resource Description Frame-
work (RDF), Web Ontology Language (WOL), and RDF Schema (RDFS).
2. Networking Issues, Identification, and Communication Protocols Anywhere-any
path, anything-any object, and anytime-any context are driving a new digital
transformation of the communication era. Although these concepts facilitate net-
work connectivity and cloud access for a wide range of sensor devices, they may
bring negatively affect routing performance. Thus, the network bandwidth and
power rate are consumed. Three networking solutions are offered to achieve scale
flexible deployment and better rate of power consumption. These solutions can
be addressed as follows:
• The Transmission Control Protocol (TCP)/Internet Protocol (IP) (TCP/IP)
stack is a conceptual model that can provide network address space reach
232
bits (
IPv4
networking).
• The
uIPv6
stack provides network address space
2128
bits (
IPv6
networking),
as well as it supports the routing protocol with a low-power loss such as
IPv6
over Low-Power Wireless Personal Area Networks (6LoWPAN) to allow a
sensor device to send and receive
IPv6
packet over IEEE 802.15.4 standard.
IEEE 802.15.4 is typically designed to enable the communication between

5707
1 3
Towards sustainable smart IoT applications architectural…
Table 13 A list of the literature studies and other information in the general aspects IoT applications
References Design/propose Features Weakness
Zhang etal. [103] An Intelligent monitoring system
A security system
Reliable system
Improving security and privacy issues
Alleviating communication overhead
Lower bandwidth cost
Mitigating the power transmission
Scalability is must be considered
The proposed system is relatively expensive
Li etal. [57] An online IoT monitoring system to control environmental
factors of farm
Easy-to-use
Reliable system
Scalable system
Low response time
High power consumption
Availability is lack
High cast
Vermesan etal. [92] A novel Internet of Robotic Things (IoRT) framework Availability
High interoperability
Improving security and privacy issues
Alleviating communication overhead
Mitigating the power transmission
Delay time is must be considered
Reliability is must be considered
Scalability is must be considered
The proposed system is relatively expensive
Abdullah etal. [1] A QoS sensitivity scheduling algorithm for IoT environ-
ment
Low power consumption
Reducing communication and compu-
tational process over the network
Mitigating the power transmission
Reliable system
Scalable system
Low delay time
Low availability
High response time
High cost
Jin etal. [52] A content-based cross-layer scheduling approach called
"CONCISE" for industrial IoT applications
Cost-effective
Low delay time
Reliable system
Availability is must be considered
Scalability is must be considered
High response time
Ali etal. [16] A QoS-aware service composition
A dynamic QoS provisioning framework (QoPF) for IoT
oriented service
Reliable system
Low delay time
Network bandwidth is reduced
Improving response time
Improving real time decision making
Scalability is must be considered
Power consumption is not considered
Iqbal etal. [49] A new schema for improving communication among intel-
ligent things
Improving response time
Improving energy consumption
Low availability
Costly system

5708
Z.H.Ali, H.A.Ali
1 3
Table 13 (continued)
References Design/propose Features Weakness
Wei etal. [95] An IoT-based energy consumption control
Monitoring and tracking services
Reducing system complexity
Fault tolerance system
Power reduction
Low delay time
Availability is lack
Costly system
Baker etal. [17] A QoS-aware multi-cloud service composition Power reduction
Improving service performance
Reliability is lack
High response time
Zhou etal. [107] An energy-aware service composition Optimizing energy consumption rate
System is relatively low cost
The complexity of system is high
System is not reliable
High response time
High delay time
Not support large scale of network
Qureshi etal. [78] A new distributed Software-Defined Networking (SDN)
approach
Enhancing response time
System is relatively reliable
System is relatively low cost
System is relatively scalable
The complexity of system is low
Optimizing energy consumption rate
The communication overhead is relatively high
Resource utilization is relatively low
Ghosh etal. [38] A Resilient Information Architecture Platform for Smart
Grid (RIAPSG)
Optimizing energy consumption rate
Enhancing response time
System is fault-tolerance
Resource utilization
System is relatively reliable
The complexity of system is high
Not support large scale of network
System is relatively high cost
Huo etal. [46] A QoS-aware service composition Enhancing response time
Low delay time
Cost-effective
Scalability is not supported
Temglit etal. [89] A QoS-aware service composition Improving the response time
Avoiding the single point of failure
Low delay time
High reliability
Scalability is not considered

5709
1 3
Towards sustainable smart IoT applications architectural…
Table 13 (continued)
References Design/propose Features Weakness
Hamad [40] An intelligent environment for precision agriculture Improving real-time performance
Reliable system
Low delay time
High power consumption
Availability is lack
High cast
Sinha etal. [86] A user-centric IoT system to optimize the various agricul-
tural issues
High reliability
Enhancing response time
Low delay time
High power consumption
High cast
Torres etal. [91] A smart agricultural architecture called "Hydra fusion
architecture"
Smart water monitoring application
Improve response time
Increasing resource utilization
High accuracy
Cost-effective
Low delay time
High power consumption
Scalability is not supported

5710
Z.H.Ali, H.A.Ali
1 3
Table 14 The main context of literature studies in the overall aspects IoT applications
Research Context Year
[103] An IoT environmental control and security system 2011
[57] An IoT monitoring system in WSN 2015
[103] A collaborative autonomous robotic system based on the concept of Internet of Things 2020
[1] A QoS-aware message scheduling context 2013
[52] A content-based cross-layer scheduling context 2017
[16] A dynamic QoS-aware IoT service composition 2019
[49] A smart electrical grid based IoT system 2018
[95] An IoT-based energy-aware context 2011
[17] A multi-cloud IoT service composition 2017
[107] A power-aware algorithm in WSN 2018
[78] An distributed Software-Defined Networking (SDN) approach for addressing the scalability and robustness issues 2020
[38] A fault-tolerance features of a software framework called Resilient Information Architecture Platform for Smart Grid
(RIAPSG)
2020
[46] A QoS-aware IoT service composition 2016
[89] A QoS-aware multi-agent IoT service composition 2017
[40] A real-time context-aware system 2018
[86] An agricultural monitoring system 2019
[91] A smart agricultural architecture with water monitoring applications 2020

5711
1 3
Towards sustainable smart IoT applications architectural…
Table 15 The evaluation QoS performance metrics of the overall aspects IoT applications studies
References QoS parameters
Availability Response time Reliability Scalability Delay time Power
consump-
tion
Cost
Zhang etal. [103]
✓
✓
✓
✗
✓
✓
✓
Li etal. [57]✗
✓
✓
✓
✓
✗ ✗
Vermesan etal.
[92]
✓
✓
✗ ✗ ✗ ✗
✓
Abdullah etal. [1]✗ ✗
✓
✓
✓
✓
✗
Jin etal. [52]✗ ✗
✓
✗ ✗ ✗
✓
Zainab H. Ali
etal. [16]
✗
✓
✓
✗
✓
✗ ✗
Iqbal etal. [49]✗
✓
✗ ✗ ✗ ✗ ✗
Wei etal. [95]✗
✓
✗ ✗
✓
✓
✓
Baker etal. [17]
✓
✗ ✗
✓
✓
✓
✗
Zhou [107]
✓
✗ ✗ ✗ ✗
✓
✓
Qureshi etal. [78]✗
✓
✓
✓
✓
✓
✓
Ghosh etal. [38]
✓
✓
✓
✗ ✗
✓
✗
Huo etal. [46]✗
✓
✗ ✗
✓
✗
✓
Temglit etal. [89]
✓
✓
✓
✗
✓
✓
✓
Hamad [40]
✓
✓
✓
✓
✓
✗ ✗
Sinha etal. [86]
✓
✓
✓
✓
✓
✗ ✗
Torres etal. [91]
✓
✓
✓
✗
✓
✗
✓
Fig. 10 The key challenges of sustainable IoT applications

5712
Z.H.Ali, H.A.Ali
1 3
sensor devices on a large-scale with a low rate of power consumption. Unfor-
tunately, IEEE 802.15.4 devices are expensive when compared with
IPv6
stack
[13].
• The Rime stack involves a group of lightweight communication protocols that
are designed to obtain low-power wireless networks [20].
RFID is a lightweight protocol that uses tags in offering a unique addressing
schema for sensing nodes as well as it can reduce power consumption within the
sensory world. Surely, the passive RFID tags don’t have the power supplies, but
they get the energy required for the communication between ID and the RFID
reader through the reader’s interrogation signal. Otherwise, the microchip tech-
nique is a genesis of the RFID manufacture. The amalgamation of the sensing
technology and RFID microchips allowed the emergence of the new applications
such as e-health; so that system can be used to automatically tracking objects in
real-time without the need of being in line-of-sight [13–42].
3 Data Management the rapid growth of data generated from the IoT entities, e.g.,
RFID, sensors, and embedded devices makes the traditional methods of data
management are not suitable enough for certain approaches such as big data.
Therefore, it needs to support a smooth integration between heterogeneous data.
The data management in IoT applies in two directions: (i) online direction (real-
time frontend) is a communication-intensive, which has a direct interconnection
with the connected-things. The process in this direction based on an online query
process and their results, and (ii) offline direction (backend process) is a storage-
intensive, which includes the storage and retrieval data. All these operations are
done according to strict normalization rules in order to reduce data redundancy
[58].
Fig. 11 IoT layers with the main QoS criteria [16]

5713
1 3
Towards sustainable smart IoT applications architectural…
4 Power Consumption the power optimization is a significant factor in the sensor
networks due to its ability to overcome the premature end of network lifetime and
extending preserve the stability period. Although the connectivity process of IoT
allows cooperating with a variety of networking "things" such as physical devices,
smart sensors, and automotive objects in the large space with high communication
of capability, these connections may adversely influence on the amount of data
transfer rate and the transmission power consumption alike.
Several studies on controlling the excessive data transmission via developing
communication protocols and tools have appeared as serious attempt to reduce
the power consumption rate such as [64–72, 74]. In the following, we discuss the
general protocols that are used to reduce power transmission over the Internet.
For instance, IEEE 802.15.4 supports low power rate, cost reduction, and low bit
rate communication. An RFID tag follows the ISO 14443 standard. It not only
offering low power but also eliminating the complexity. Bluetooth Low Energy
(Bluetooth LE) provides a wireless personal area network with a low rate of
energy consumption. ISO 18000- 7DASH7 provides low complexity and allevi-
ates the power consumption rate [93]. Moreover, Table16 depicts the approximate
data transmission rate and some other characteristics of several communication
protocols that may use in different IoT applications.
Energy harvesting mechanisms are used to bridge the energy gap among sensor
nodes, so they can harvest many ambient energy resources to power sensor nodes.
On the other hand, energy harvesting plays key role in the sustainability for IoT
applications due to its ability to use renewable source energy or practically unlim-
ited sources and reduce exhaustible sources of energy. There are several reasons to
deplete the energy of sensor batteries include extreme weather conditions, limited
battery lifetime, and the usage of applications that consumes the battery (i.e., Global
Position System (GPS)) even if the device is not in use [93]. The IoT environment
has many ambient energy sources can be considered as follows:
I Radiant energy or Radio Frequency-based energy harvesting RFID is one of
RF energy harvesting solutions, which is now available in the markets. it con-
verts radio waves (RF) to DC power. The conversion efficiency from RF to DC
is between 50% and 75% over a 100 m range of the input power.
II Thermal-based energy harvesting it converts the heat energy to electrical
energy using the Seebeck effect. A thermoelectric harvester offers several
advantages such as a long lifetime, stationary parts, and highly reliable charac-
teristics. But it lacks of efficiency up to (5-6%) of thermal harvesting.
III Mechanical-based energy harvesting it converts mechanical power such as
temperature, pressure, and vibration to electrical power. Typically, this type of
the harvesting system uses either the electromagnetic, electrostatic, and piezo-
electric mechanisms to harvest energy. For instance, in electromagnetic energy
harvesting, vibrations are required to move a magnet coil and generate an elec-
tric current. In electrostatic energy harvesting, vibrations are used to pull the

5714
Z.H.Ali, H.A.Ali
1 3
plates of a charged capacitor against the electrostatic attraction, resulting in
electrical energy due to the capacitance change. In piezoelectric energy har-
vesting, vibrations are used to produce an electric potential difference that can
be extracted as electrical energy. The successful application of energy harvest-
ing systems in the IoT environment should be addressed three main directions:
energy conversion, energy storage, and power management [93]. As a result,
open research challenges that are needed to be covered in this context can be
listed as the generic harvester, miniaturization, protocol adaptation, energy
storage, energy-efficient reliable systems, and efficient prediction techniques
[85, 94].
5. Security and Privacy The dynamic behavior of sensor nodes without an adequate
monitoring system and sensory infrastructure that has low capabilities, especially
the case of passive components (RFID tags), make the IoT environment vulnerable
to malicious attacks. Hence, providing a reasonable level of security has become an
integral part of sustainable applications. The security scope means offering a protec-
tion system to protect data against external attacks, and there are two main prob-
lems related to this scope: authentication and data integrity. Authentication seeks for
creating an authenticated node over the network to exchange messages safely. Data
integrity means the data protected against internal attacks (e.g., modify and delete)
[13].
The privacy policy aims to protect personal information by supporting anonymity
over the Internet. Cryptographic methods with an encryption technique play a fun-
damental role in this context so that they ensure data confidentiality via the Internet.
In this end, the ubiquitous environment has been posed new privacy challenges such
as preserving location privacy, preserving personal information inference, and use
of soft identities [13].
3.2 Sustainable IoT applications—from ubiquitous computing perspective:
challenges andbenefits
Scalability, limited resources and infrastructure, excessive fees, and insufficient
performance in real-time applications are major challenges that may hold back the
realization of sustainable growth in IoT applications. One solution to providing
better network architecture with averting much of these issues that is the comput-
ing platforms. The amalgamation of cloud computing platforms and IoT not only
enhances networking resource utilization but also provides a large number of on-
demand elastic services in trust and reliable manners [41]. As shown in Figs.12
and13, the term of IoT has undergone a lot of research from Google company, so
that it has issued the statistical significance of the comparison between IoT and
cloud computing during the period from 2004 to 2011 to prove that the M2M
technology (means IoT) is heavily important more than cloud computing, while
the term of cloud computing is more popular until now.
"Things-as-a-Service" is the most beneficial outcome of the CloudIoT inte-
gration paradigm. In a broader sense, cloud computing presents the things as

5715
1 3
Towards sustainable smart IoT applications architectural…
ubiquitous services that can be accessed from multiple users by using the vir-
tualization technology, as well as it allows us handling these things through the
conception of "pay-as-you-go," and therefore, the cost is reduced [18]. The inte-
gration paradigm offers a set of features, which cannot be ignored for example:
1. Virtualization Technique the ability to move from the real world to the digital
world in order to create virtual infrastructure and computational resources. IBM
defined this technique as a transparent way of providing collaborative access to
mainframe computers, where time-sharing and resource share allow multiple
users to concurrently use large-scale and highly expensive hardware. There are
different types of virtualization techniques provided by cloud computing, for
instance, network virtualization, storage virtualization, server virtualization, data
virtualization, and application virtualization. Modern cloud computing systems
may take advantage of virtualization technologies to maximize the utilization of
resources and minimize the cost of computation [75].
2. Scalability the scalable system is a significant pattern cloud architecture to man-
age the growth of on-demand systems and services without damage or failures.
There are different ways to accomplish scaling in the network. One is vertical
scaling and another is horizontal scaling. Vertical scaling/scaling up is essentially
referred to resize the networking area by adding more infrastructure, which is con-
sidered an expensive way. By contrast, horizontal scaling/scaling out is a server’s
ability to scale wider services to deal with the rapidly increasing network needs.
This type of scalability can be implemented instantly, as well as it considered as
cost-effective scaling. Typically, just-in-time scalability one of the major factors
needed to enable extending the IoT services in real-time [98].
3. Time acceleration the cloud computing is an efficient way to address, analyze,
and process a huge amount of data streaming out from IoT sensors; due to its high
computational power and storage capability. However, it is not suitable enough
to deal with the real-time applications that need to make a decision on time, the
delay here is not acceptable. The reason among others, cloud computing is a cen-
tral platform therefore all data need to move to the cloud for processing. Although
Table 16 Approximate range of communication protocols [65]
Protocol Bandwidth Frequency Transmission rate
Cellular
2G 900 MHz 1.8 GHz 64 kbps
3G 900 MHz 1.6–2.0 GHz 144 kbps–2 Mbps
4G 100 MHz 2–8 GHz 100 Mbps–1 Gbps
5G 1000 per unit of area 3–300 GHz 1 Gbps <
IEEE 802.15.4
WiFi 6 Mbps 2.4–5 GHz 10 Mbps
Zigbee 868–868.8 MHz 2.4 250 Kbps
LoRaWAN 125–250 kHz 433–868 MHz 5468 bps
6LoWPAN 20–250 kbps 2.4 GHz 250 Kbps

5716
Z.H.Ali, H.A.Ali
1 3
the speed of data processing on the cloud is fast, the network bandwidth is still a
bottleneck [45]. The need to accelerate decision-making by allowing a stronger
reduction of the response time has become inevitable. Several cloud platforms
are introduced to bridge the gap between the execution time and transferring time
without the need to consume network bandwidth, a platform called fog comput-
ing. For better performance with low delay time, fog computing is introduced
as a local place to compute, collect, store, and process data generated by sensor
devices without the need to transfer this data to the cloud [104].
4. Utility Computing it is the most popular cloud services, which is concerned with
providing on-demand computational resources over the internet. The utility ser-
vices seek for maximizing resource utilization meanwhile minimizing cost. It
plays a key role in supporting the scalability for IoT applications without system
failure [23].
5. New Capabilities and New Services unlimited infrastructure, high storage capac-
ity, high availability for services in a reliable manner, and scalable services all of
them are the new capabilities added by the integration paradigm called CloudIoT.
This integration paradigm allows to bridge the divide between the real world
and the digital world by providing new types of services such as Actuation-as-a-
Service (SAaaS) which enables automatic control logics implemented in cloud
computing through Sensing, Sensor Event-as-a-Service (SEaaS) which dispatches
messaging services triggered by sensor events, DataBase-as-a-Service (DBaaS)
which enables ubiquitous database management, Video Surveillance-as-a-Service
(VSaaS) which provides ubiquitous access to recorded video and implementing
complex analyses in Cloud, and Sensor-as-a-Service (SenaaS) which enables
ubiquitous management of remote sensors [23]. The features of the CloudIoT
paradigm between cloud computing and IoT is introduced in Table17.
4 Future research directions andopens business opportunities
This section demonstrates the future research directions and opens business oppor-
tunities in the field of IoT to create successful sustainability development. As we
explored in Sect.2, many IoT applications include smart cities, smart healthcare,
and intelligent transportation require innovative end-to-end networking solutions to
investigator and application designers that focus on how the QoS can be achieved. In
Sect.3, the challenges that face achieving the sustainability of IoT applications were
classified into two categories. One is about the QoS challenges, which are related to
service behavior and service attitude, and another is about the ubiquitous computing
and the integration between cloud and IoT in order to bring a powerful Things-as-a-
Service context. Thus, improving network architecture and real-time performance.
Table18 discusses the list of hot topics future research directions that ensure sus-
tainable development growth in the field of IoT.

5717
1 3
Towards sustainable smart IoT applications architectural…
5 Discussion andconclusion
The importance of this study comes according to its ability to highlight how the con-
vergence of the academic research directions that published a significant number of
papers in the field of IoT every year and the real-time industry’s needs to fulfill suc-
cessful sustainability. Thus, enhancing quality of life with minimal investment. The
outlines how IoT can act as the catalyst to promote the development rate in develop-
ing countries in order to achieve, and surpass, the sustainable development goals in
a time frame unthinkable even a few years ago.
The CloudIoT’s integration project has shown a massive rise in innovative tech-
nologies that ensure the sustainable growth of the major of IoT applications such
as intelligent transportation, smart cities, smart healthcare, and other smart appli-
cations. Although the convergence of cloud and IoT offers powerful ubiquitous
computing scenarios that can achieve better networking solutions, it may face many
challenges not only in performance but also in architecture. In this study, we used
three big electronic databases, namely Elsevier, Springer, and IEEE Xplore to inves-
tigate the sustainable production development in the field of IoT applications. The
selection process of papers was according to pervasive context-aware of IoT applica-
tions and QoS evaluation factors. All selected papers covered the period between
2000 and 2020. After the statistical analysis conducted in the previous sections, the
smart healthcare applications have the lion’s share by 34% of quotas in the literature.
While the smart city and ITS were at the same level by 28% of the literature reviews,
the third level was to utility services and anther related topic is to power consump-
tion by 11%.
Availability, reliability, scalability, response time, power consumption, resource
utilization, communication overhead, interoperability, and security are the major
QoS performance metrics that are used in the evaluation process in this study.
Among these factors, the response time was taken the highest ranking by 45%.
Fig. 12 The importance of M2M versus cloud computing [20]

5718
Z.H.Ali, H.A.Ali
1 3
While the scalability and power consumption managed to edge out the reliability
by one point for second place with 31% against 30%, the midpoint was availability,
delay time, and cost, which fall to around 25%, 19%, and 15%. The lowest rate was
interoperability, resource utilization, location-awareness, communication overhead,
and security, which stood at 10%, 9%, 7%, 5%, and 4% respectively. Our recommen-
dation is that the greater effort must be made to promote the QoS factors, especially
with the factors that have the lowest places.
Sustainable development is the blueprint toward addressing the networking chal-
lenges and issues face IoT applications. Several challenges with the proposed solu-
tions were discussed in this study. Moreover, this study successfully introduced
Fig. 13 The popularity of cloud computing versus M2M [20]
Table 17 The Features provided
by the CloudIoT paradigm Features CloudIoT Cloud I oT
Scalability A A N/A
Availability A A A
Reliability A A A
Mobility A N/A A
Virtualization A A N/A
Interoperability A A N/A
Easy-to-use A A Hard
Power efficiency A N/A N/A
Elastic Services A A N/A
Cost effectiveness Expensive A Expensive
Location-awareness A N/A A
Service delivery model A A N/A
Cloud acceleration A A N/A
Real-time performance A A A

5719
1 3
Towards sustainable smart IoT applications architectural…
Table 18 The list of future researches and open directions that ensure sustainable development growth in the field of IoT
Dimension Description Future directions
Standardization and architectural design The ability to create a standard framework to enhance interoperability
among nodes
Uniform framework
Edge-Compute solutions
Cloud computing
Virtualization technology
Middleware solutions
Semantic web
Semantic interoperability
Context-aware computing It is a computer operation to manage, organize, and customize user’s
demands according to the availability of machine capacity
Cloud computing
Fog computing
Semantic search
Middleware solutions
Ad-Hoc networks
Adaptive and context-aware architecture
Cluster techniques
Networking issues, identification, and
communication protocols
The capacity to improve network performance through congestion manage-
ment, standard protocols, and unique addressing schema
Ad-Hoc networks and Hyper networking
Self-configuration
Virtualization technology (location transparency)
Self-organization networks
Optimizing live migration mechanism
Mapping between both digital world & real world
Conductivity and communication range
Deployment techniques
Device discovery
Semantic Search
RFID tags
Universal authentications mechanism
Cellular connectivity management

5720
Z.H.Ali, H.A.Ali
1 3
Table 18 (continued)
Dimension Description Future directions
Data quality and management The ability to manage and reduce the huge amount of data generated by
sensors and provide standard specifications for all the information to
avoid the useless data as soon as possible
Data normalization techniques
Data aggregation techniques
Data filtering techniques
Semantic techniques
Data analysis techniques
Cloud storage mechanism
Compression algorithms
Machine learning algorithms support wrapper
method and predictive model
Power consumption The ability to manage and reduce the rate of power consumption Semantic interoperability
Micro battery technologies
Computing solutions
Energy harvesting
Efficient mechanism for routing
Hardware, i.e., sensor, embedded device, and RFID
Security and privacy The ability to protect data from unauthorized users. Generally, there are
three levels of the security issues: confidentiality, trust, and integrity
Security in cloud computing
Security in semantic web
Encryption and cryptographic methods
Privacy policies and trust

5721
1 3
Towards sustainable smart IoT applications architectural…
future research directions in the field of IoT with better guidelines for accelerating
business opportunities.
References
1. Abdullah S, Yang K (2013) A qos aware message scheduling algorithm in internet of things
environment. In: 2013 IEEE Online Conference on Green Communications (OnlineGreen-
Comm), pp 175–180. IEEE
2. Ahmed SH, Rani S (2018) A hybrid approach, smart street use case and future aspects for inter-
net of things in smart cities. Future Gener Comput Syst 79:941–951
3. Alam M, Ferreira J, Fonseca J (2016) Introduction to intelligent transportation systems. In:
Intelligent transportation systems, pp 1–17. Springer
4. Ali ZH, Ali HA, Badawy MM (2015) Internet of things (iot): definitions, challenges and recent
research directions. Int J Comput Appl 975:8887
5. Ali ZH, Ali HA, Badawy MM (2017) A new proposed the internet of things (iot) virtualization
framework based on sensor-as-a-service concept. Wirel Pers Commun 97:1419–1443. https ://
doi.org/10.1007/s1127 7-017-4580-x
6. Ali ZH, Badawy MM, Ali HA (2020) A novel geographically distributed architecture based on
fog technology for improving Vehicular Ad hoc Network (VANET) performance. Peer-to-Peer
Netw Appl 13:1539–1566. https ://doi.org/10.1007/s1208 3-020-00910 -9
7. Alqahtani F, Al-Makhadmeh Z, Tolba A, Said W (2020) Internet of things-based urban waste
management system for smart cities using a cuckoo search algorithm. Clust Comput 1–12
8. Amah TE, Kamat M, Bakar KA, Moreira W, Oliveira A Jr, Batista MA (2020) Preparing oppor-
tunistic networks for smart cities: collecting sensed data with minimal knowledge. J Parallel
Distrib Comput 135:21–55
9. Angin P, Bhargava B, Helal S (2010) A mobile-cloud collaborative traffic lights detector for
blind navigation. In: 2010 Eleventh International Conference on Mobile Data Management, pp
396–401. IEEE
10. Ansari K (2018) Cloud computing on cooperative cars (c4s): an architecture to support naviga-
tion-as-a-service. In: 2018 IEEE 11th International Conference on Cloud Computing (CLOUD),
San Francisco, CA, USA, pp 794-801.https ://doi.org/10.1109/CLOUD .2018.00108
11. Arora S, Bala A (2020) Pap: power aware prediction based framework to reduce disk energy
consumption. Clust Comput 1–18
12. Asghari P, Rahmani AM, Javadi HHS (2018) Service composition approaches in iot: a system-
atic review. J Netw Comput Appl 120:61–77
13. Atzori L, Iera A, Morabito G (2010) The internet of things: a survey. Comput Netw
54(15):2787–2805
14. Atzori L, Iera A, Morabito G, Nitti M (2012) The social internet of things (siot)-when social
networks meet the internet of things: concept, architecture and network characterization. Com-
put Netw 56(16):3594–3608
15. Aziz K, Tarapiah S, Ismail SH, Atalla S (2016) Smart real-time healthcare monitoring and
tracking system using GSM/GPS technologies. In: 2016 3rd MEC International Conference on
Big Data and Smart City (ICBDSC), pp 1–7. IEEE
16. Badawy MM, Ali ZH, Ali HA (2019) Qos provisioning framework for service-oriented internet
of things (IoT). Clust Comput 1–17
17. Baker T, Asim M, Tawfik H, Aldawsari B, Buyya R (2017) An energy-aware service composi-
tion algorithm for multiple cloud-based IoI applications. J Netw Comput Appl 89:96–108
18. Barbarán J, Díaz M, Rubio B (2014) A virtual channel-based framework for the integration of
wireless sensor networks in the cloud. In: 2014 International Conference on Future Internet of
Things and Cloud, pp 334–339. IEEE
19. Baskar S, Shakeel PM, Kumar R, Burhanuddin M, Sampath R (2020) A dynamic and interop-
erable communication framework for controlling the operations of wearable sensors in smart
healthcare applications. Comput Commun 149:17–26
20. Behmann F, Wu K (2015) Collaborative internet of things (C-IoT), for future smart connected
life and business. Wiley, Hoboken

5722
Z.H.Ali, H.A.Ali
1 3
21. Belkeziz R, Jarir Z (2016) A survey on internet of things coordination. In: 2016 Third Interna-
tional Conference on Systems of Collaboration (SysCo), pp 1–6. IEEE
22. Bellavista P, Berrocal J, Corradi A, Das SK, Foschini L, Zanni A (2018) A survey on fog com-
puting for the internet of things. Pervasive Mob Comput
23. Botta A, DeDonato W, Persico V, Pescapé A (2014) On the integration of cloud computing and
internet of things. In: 2014 International Conference on Future Internet of Things and Cloud, pp
23–30. IEEE
24. Botta A, De Donato W, Persico V, Pescapé A (2016) Integration of cloud computing and internet
of things: a survey. Future Gener Comput Syst 56:684–700
25. Boukerche A, Robson E (2018) Vehicular cloud computing: architectures, applications, and mobil-
ity. Comput Netw 135:171–189. https ://doi.org/10.1016/j.comne t.2018.01.004
26. Chen Y, Ardila-Gomez A, Frame G (2017) Achieving energy savings by intelligent transportation
systems investments in the context of smart cities. Transp Res Part D Transport Environ 54:381–
396. https ://doi.org/10.1016/j.trd.2017.06.008
27. Chung K, Park RC (2019) Cloud based u-healthcare network with QoS guarantee for mobile health
service. Clust Comput 22(1):2001–2015
28. Debnath AK, Chin HC, Haque MM, Yuen B (2014) A methodological framework for benchmark-
ing smart transport cities. Cities 37:47–56
29. Deen MJ (2015) Information and communications technologies for elderly ubiquitous healthcare in
a smart home. Personal and Ubiquitous Computing 19(3–4):573–599
30. Della Vecchia G, Gallo L, Esposito M, Coronato A (2012) An infrastructure for smart hospitals.
Multimed Tools Appl 59(1):341–362
31. Ding Y, Gang S, Hong J (2015) The design of home monitoring system by remote mobile medi-
cal. In: 2015 7th International Conference on Information Technology in Medicine and Education
(ITME), pp 278–281. IEEE
32. Dinh HT, Lee C, Niyato D, Wang P (2013) A survey of mobile cloud computing: architecture,
applications, and approaches. Wirel Commun Mob Comput 13(18):1587–1611
33. Distefano S, Longo F, Scarpa M (2015) QoS assessment of mobile crowdsensing services. J Grid
Comput 13(4):629–650
34. Donaldson L, Halligan A, Wall D (2003) Improving the child’s experience of healthcare: the
implementation of clinical governance. Curr Paediatr 13(4):298–302
35. Elmisery AM, Rho S, Aborizka M (2019) A new computing environment for collective pri-
vacy protection from constrained healthcare devices to IoT cloud services. Clust Comput
22(1):1611–1638
36. Fafoutis X, Clare L, Grabham N, Beeby S, Stark B, Piechocki R, Craddock I (2016) Energy neu-
tral activity monitoring: wearables powered by smart inductive charging surfaces. In: 2016 13th
Annual IEEE International Conference on Sensing, Communication, and Networking (SECON),
pp 1–9. IEEE
37. Fernando N, Loke SW, Rahayu W (2013) Mobile cloud computing: a survey. Future Gener Com-
put Syst 29(1):84–106
38. Ghosh P, Eisele S, Dubey A, Metelko M, Madari I, Volgyesi P, Karsai G (2020) Designing a
decentralized fault-tolerant software framework for smart grids and its applications. J Syst Arch
101759
39. Gubbi J, Buyya R, Marusic S, Palaniswami M (2013) Internet of things (IoT): a vision, architec-
tural elements, and future directions. Future Gener Comput Syst 29(7):1645–1660
40. Hamad MA, Eltahir MES, Ali AEM, Hamdan AM (2018) Efficiency of using smart-mobile phones
in accessing agricultural information by smallholder farmers in north Kordofan-Sudan. Available
at SSRN 3240758
41. Hayes B (2008) Cloud computing. Commun ACM 51(7):9–11
42. Hema C, Sankar S, etal (2016) Energy efficient cluster based protocol to extend the RFID network
lifetime using dragonfly algorithm. In: 2016 International Conference on Communication and Sig-
nal Processing (ICCSP), pp 0530–0534. IEEE
43. Hou X, Li Y, Chen M, Wu D, Jin D, Chen S (2016) Vehicular fog computing: a viewpoint of vehi-
cles as the infrastructures. IEEE Trans Veh Technol 65(6):3860–3873
44. Hsu CY, Yang CS, Yu LC, Lin CF, Yao HH, Chen DY, Lai KR, Chang PC (2015) Development of
a cloud-based service framework for energy conservation in a sustainable intelligent transportation
system. Int J Prod Econ 164:454–461

5723
1 3
Towards sustainable smart IoT applications architectural…
45. Hu P, Dhelim S, Ning H, Qiu T (2017) Survey on fog computing: architecture, key technolo-
gies, applications and open issues. J Netw Comput Appl 98:27–42. https ://doi.org/10.1016/j.
jnca.2017.09.002
46. Huo L, Wang Z (2016) Service composition instantiation based on cross-modified artificial bee
colony algorithm. China Commun 13(10):233–244
47. Iot definition. https ://www.rfidj ourna l.com/artic les/view?4986. Accessed 30 April 2019
48. Iotvisionineurope. http://www.inter net-of-thing s-resea rch.eu/. Accessed 30 April 2019
49. Iqbal J, Khan M, Talha M, Farman H, Jan B, Muhammad A, Khattak HA (2018) A generic internet
of things architecture for controlling electrical energy consumption in smart homes. Sustain Cities
Soc 43:443–450
50. Ismagilova E, Hughes L, Dwivedi YK, Raman KR (2019) Smart cities: advances in research-an
information systems perspective. Int J Inf Manag 47:88–100
51. Jin J, Gubbi J, Luo T, Palaniswami M (2012) Network architecture and QoS issues in the internet
of things for a smart city. In: 2012 International Symposium on Communications and Information
Technologies (ISCIT), pp 956–961. IEEE
52. Jin Y, Raza U, Aijaz A, Sooriyabandara M, Gormus S (2017) Content centric cross-layer schedul-
ing for industrial IoT applications using 6tisch. IEEE Access 6:234–244
53. Kavitha K, Perumalraja R (2014) Smart wireless healthcare monitoring for drivers community. In:
2014 International Conference on Communication and Signal Processing, pp 1105–1108. IEEE
54. Khalifa AB, Alouani I, Mahjoub MA, Rivenq A (2020) A novel multi-view pedestrian detec-
tion database for collaborative intelligent transportation systems. Future Gener Comput Syst
113:506–527
55. Kiljander J, D’elia A, Morandi F, Hyttinen P, Takalo-Mattila J, Ylisaukko-Oja A, Soininen JP,
Cinotti TS (2014) Semantic interoperability architecture for pervasive computing and internet of
things. IEEE Access 2:856–873
56. Lai Y, Yang F, Su J, Zhou Q, Wang T, Zhang L, Xu Y (2018) Fog-based two-phase event monitor-
ing and data gathering in vehicular sensor networks. Sensors 18(1):82
57. Li H, Wang H, Yin W, Li Y, Qian Y, Hu F (2015) Development of a remote monitoring system for
henhouse environment based on IoT technology. Future Internet 7(3):329–341
58. Li L, Li S, Zhao S (2014) QoS-aware scheduling of services-oriented internet of things. IEEE
Trans Indu Inform 10(2):1497–1505
59. Li Y, Orgerie AC, Rodero I, Amersho BL, Parashar M, Menaud JM (2018) End-to-end energy
models for edge cloud-based IoT platforms: application to data stream analysis in IoT. Future
Gener Comput Syst 87:667–678
60. Lin CC, Deng DJ, Yao CC (2017) Resource allocation in vehicular cloud computing systems with
heterogeneous vehicles and roadside units. IEEE Internet Things J 5(5):3692–3700
61. Lin N, Wang X, Zhang Y, Hu X, Ruan J (2020) Fertigation management for sustainable precision
agriculture based on internet of things. J Clean Prod 124119
62. Lingling H, Haifeng L, Xu X, Jian L (2011) An intelligent vehicle monitoring system based on
internet of things. In: 2011 Seventh International Conference on Computational Intelligence and
Security, Hainan, 2011, pp. 231-233. https ://doi.org/10.1109/CIS.2011.59
63. Lynn T, Endo PT, Ribeiro AMN, Barbosa GB, Rosati P (2020) The internet of things: definitions,
key concepts, and reference architectures. Cloud Thing Contin 1
64. Lyu Y, Yin P (2020) Internet of things transmission and network reliability in complex environ-
ment. Comput Commun 150:757–763
65. Mahmoud MS, Mohamad AA (2016) A study of efficient power consumption wireless communi-
cation techniques/modules for internet of things (IoT) applications
66. Mao X, Miao X, He Y, Li XY, Liu Y (2012) Citysee: urban CO
2
monitoring with sensors. In: 2012
Proceedings IEEE INFOCOM, Orlando, FL, 2012, pp 1611-1619. https ://doi.org/10.1109/INFCO
M.2012.61955 30
67. Masuoka R, Parsia B, Labrou Y (2003) Task computing—the semantic web meets pervasive com-
puting. In: International Semantic Web Conference, pp 866–881. Springer
68. Mohan P, Padmanabhan VN, Ramjee R (2008) Nericell: rich monitoring of road and traffic condi-
tions using mobile smartphones. In: Proceedings of the 6th ACM Conference on Embedded Net-
work Sensor Systems, pp 323–336. ACM
69. Mokhtar SB, Georgantas N, Issarny V (2007) Cocoa: conversation-based service composition in
pervasive computing environments with QoS support. J Syst Softw 80(12):1941–1955

5724
Z.H.Ali, H.A.Ali
1 3
70. Montori F, Bedogni L, Bononi L (2017) A collaborative internet of things architecture for smart
cities and environmental monitoring. IEEE Internet Things J 5(2):592–605
71. Moon D, Chung Y, Pan SB, Park JW (2009) Integrating fingerprint verification into the smart card-
based healthcare information system. EURASIP J Adv Signal Process 2009:16
72. Mora H, Peral J, Ferrández A, Gil D, Szymanski J (2019) Distributed architectures for intensive
urban computing: a case study on smart lighting for sustainable cities. IEEE Access 7:58449–58465
73. MS Windows NT iot history. https ://www.posts capes .com/inter net-of-thing s-histo ry/. Accessed 30
April 2019
74. Muralidhara S, Hegde N, Rekha P (2020) An internet of things-based smart energy meter for moni-
toring device-level consumption of energy. Comput Electr Eng 87:106772
75. Noshy M, Ibrahim A, Ali HA (2018) Optimization of live virtual machine migration in cloud com-
puting: a survey and future directions. J Netw Comput Appl 110:1–10
76. Patan R, Ghantasala GP, Sekaran R, Gupta D, Ramachandran M (2020) Smart healthcare and qual-
ity of service in IoT using grey filter convolutional based cipher physical system. Sustain Cities Soc
59:102141. https ://doi.org/10.1016/j.scs.2020.10214 1
77. Pereira J, Ricardo L, Luís M, Senna C, Sargento S (2019) Assessing the reliability of fog comput-
ing for smart mobility applications in vanets. Future Gener Comput Syst 94:317–332
78. Qureshi KN, Hussain R, Jeon G (2020) A distributed software defined networking model to
improve the scalability and quality of services for flexible green energy internet for smart grid sys-
tems. Comput Electr Eng 84:106634
79. Qureshi KN, Tayyab MQ, Rehman SU, Jeon G (2020) An interference aware energy efficient data
transmission approach for smart cities healthcare systems. Sustain Cities Soc 62:102392
80. Rahman F, Ahamed SI (2013) Towards improving reliability of computational RFID based smart
healthcare monitoring systems. In: International Conference on Smart Homes and Health Telemat-
ics, pp 211–217. Springer
81. Razzaque MA, Milojevic-Jevric M, Palade A, Clarke S (2016) Middleware for internet of things: a
survey. IEEE Internet things J 3(1):70–95
82. Ryden M, Oh K, Chandra A, Weissman J (2014) Nebula: distributed edge cloud for data intensive
computing. In: 2014 IEEE International Conference on Cloud Engineering, pp 57–66. IEEE
83. Seo D, Jeon YB, Lee SH, Lee KH (2016) Cloud computing for ubiquitous computing on m2m and
IoT environment mobile application. Clust Comput 19(2):1001–1013
84. Shaikh FK, Zeadally S (2016) Energy harvesting in wireless sensor networks: a comprehensive
review. Renew Sustain Energy Rev 55:1041–1054
85. Shaukat U, Ahmed E, Anwar Z, Xia F (2016) Cloudlet deployment in local wireless networks:
motivation, architectures, applications, and open challenges. J Netw Comput Appl 62:18–40
86. Sinha A, Shrivastava G, Kumar P (2019) Architecting user-centric internet of things for smart agri-
culture. Sustain Comput Inform Syst 23:88–102
87. Sridhar M, Pankajavalli P (2020) An optimization of distributed voronoi-based collaboration for
energy-efficient geographic routing in wireless sensor networks. Clust Comput 23:1741–1754.
https ://doi.org/10.1007/s1058 6-020-03122 -1
88. Symonds J (2009) Ubiquitous and pervasive computing: concepts, methodologies, tools, and appli-
cations: concepts, tools, and applications. IGI Global, p 1962
89. Temglit N, Chibani A, Djouani K, Nacer MA (2017) A distributed agent-based approach for opti-
mal QoS selection in web of object choreography. IEEE Syst J 12(2):1655–1666
90. Teng H, Liu Y, Liu A, Xiong NN, Cai Z, Wang T, Liu X (2019) A novel code data dissemination
scheme for internet of things through mobile vehicle of smart cities. Future Gener Comput Syst
94:351–367
91. Torres AB, da Rocha AR, da Silva TLC, de Souza JN, Gondim RS (2020) Multilevel data fusion
for the internet of things in smart agriculture. Comput Electron Agric 171:105309
92. Vermesan O, Bahr R, Ottella M, Serrano M, Karlsen T, Wahlstrøm T, Sand HE, Ashwathnarayan
M, Gamba MT (2020) Internet of robotic things intelligent connectivity and platforms. Front Robot
AI 7:104
93. Vermesan O, Friess P (2013) Internet of things: converging technologies for smart environments
and integrated ecosystems. River Publishers, Gistrup
94. Wang X, Dong JS, Chin C, Hettiarachchi S, Zhang D (2004) Semantic space: an infrastructure for
smart spaces. IEEE Pervasive Comput 3:32–39

5725
1 3
Towards sustainable smart IoT applications architectural…
95. Wei C, Li Y (2011) Design of energy consumption monitoring and energy-saving management
system of intelligent building based on the internet of things. In: 2011 International Conference on
Electronics, Communications and Control (ICECC), pp 3650–3652. IEEE
96. Whaiduzzaman M, Sookhak M, Gani A, Buyya R (2014) A survey on vehicular cloud computing. J
Netw Comput Appl 40:325–344
97. Whitmore A, Agarwal A, Da Xu L (2015) The internet of things–a survey of topics and trends. Inf
Syst Front 17(2):261–274
98. Wilder B (2012) Cloud architecture patterns: using microsoft azure. O’Reilly Media Inc,
Sebastopol
99. Yang CT, Chen ST, Liu JC, Liu RH, Chang CL (2020) On construction of an energy monitoring
service using big data technology for the smart campus. Clust Comput 23(1):265–288
100. Yi S, Li C, Li Q (2015) A survey of fog computing: concepts, applications and issues. In: Proceed-
ings of the 2015 workshop on mobile big data, pp 37–42. ACM
101. Yu L, Nazir B, Wang Y (2020) Intelligent power monitoring of building equipment based on
internet of things technology. Comput Commun 157:76–84. https ://doi.org/10.1016/j.comco
m.2020.04.016
102. Zamanifar A, Nazemi E (2020) Eecasc: an energy efficient communication approach in smart cit-
ies. Wirel Netw 26(2):925–940
103. Zhang L (2011) An iot system for environmental monitoring and protecting with heterogeneous
communication networks. In: 2011 6th International ICST Conference on Communications and
Networking in China (CHINACOM), pp 1026–1031. IEEE
104. Zhang P, Zhou M, Fortino G (2018) Security and trust issues in Fog computing: a survey. Future
Gener Comput Syst 88:16–27. https ://doi.org/10.1016/j.futur e.2018.05.008
105. Zhang Q, Jin M, Wang Y, Lei X (2019) Smart city information acquisition system based on inter-
net of things. Clust Comput 22(4):9013–9025
106. Zheng J, Ha C, Zhang Z (2017) Design and evaluation of a ubiquitous chest-worn cardiopulmonary
monitoring system for healthcare application: a pilot study. Med Biol Eng Comput 55(2):283–294
107. Zhou Z, Zhao D, Liu L, Hung PC (2018) Energy-aware composition for wireless sensor networks
as a service. Future Gener Comput Syst 80:299–310
108. Zhu C, Leung VC, Shu L, Ngai ECH (2015) Green internet of things for smart world. IEEE Access
3:2151–2162
109. Zulkifli FY, Mustika IW etal (2018) Development of monitoring and health service information
system to support smart health on android platform. In: 2018 4th International Conference on Nano
Electronics Research and Education (ICNERE), pp 1–6. IEEE
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published
maps and institutional affiliations.
Aliations
ZainabH.Ali1 · HeshamA.Ali1
* Zainab H. Ali
zainabhassan@ai.kfs.edu.eg
https://www.linkedin.com/in/zainab-hassan-65229232/
https://scholar.google.com.eg/citations?user=Uyr4QvkAAAAJ&hl=en
Hesham A. Ali
h_arafat_ali@mans.edu.eg
https://scholar.google.com.eg/citations?user=_nOkE-IAAAAJ&hl=en&oi=ao
https://www.linkedin.com/in/hesham-arafat-ali-91285071/
1 Network Embedded Systems andTechnology Department, Faculty ofArtificial Intelligence,
Kafrelsheikh University, Kafrelsheikh33511, Egypt