See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/320532203

Internet of Things (IoT): Definitions, Challenges, and Recent Research

Directions

Article in International Journal of Computer Applications · October 2015

CITATIONS READS

146 83,138

3 authors:

Zainab H. Ali Hesham Arafat Ali

Kafrelsheikh University Mansoura University
19 PUBLICATIONS 303 CITATIONS 194 PUBLICATIONS 2,540 CITATIONS

SEE PROFILE SEE PROFILE

Mahmoud M. Badawy
Mansoura University
13 PUBLICATIONS 462 CITATIONS

SEE PROFILE

All content following this page was uploaded by Zainab H. Ali on 20 October 2017.

The user has requested enhancement of the downloaded file.

 International Journal of Computer Applications (0975 – 8887)
Volume 128 – No.1, October 2015

Internet of Things (IoT): Definitions, Challenges and

Recent Research Directions

Zainab H. Ali Hesham A. Ali Mahmoud M. Badawy

Dept. of Computer Eng. & Dept. of Computer Eng. & Dept. of Computer Eng. &
Systems, Mansoura University, Systems, Mansoura University, Systems, Mansoura University,
Egypt. Egypt. Egypt.

ABSTRACT with an intelligent behavior and taking into account the

In this paper, we seek to highlight the concept of Internet of existence of the kind of autonomy and privacy. Meanwhile,
Things (IoT) in general, as well as reviewing the main the IoT environment contains a huge number of the
challenges of the IoT environment by focusing on the recent different objects/things can be classified into two types
research directions in this topic. Recently, IoT has emerged namely; i) Things rechargeable batteries things: the most of
as a new technology that is used to express a modern them are mobiles (e.g. Laptop, tablets and mobile phone),
wireless telecommunication network, and it can be defined and ii) Things are non-rechargeable things: these things are
as an intelligent and interoperability node interconnected in static from the mobility point of view [4]. Generally, IoT
a dynamic global infrastructure network, also it seeks to includes three main demands are: the first, a shared
implement the connectivity concept of anything from understanding of the situation of its users and their
anywhere at anytime. Indeed, the IoT environment applications. Secondly, software architecture and pervasive
possesses a large spectrum of challenges has a broad impact communication networks to cover and process contextual
on their performance, which can be divided into two information, and lastly, the analytics tools in IoT that aims
categories, namely, i) General challenges: such as for autonomous and intelligent behavior [5].
communication, heterogeneity, virtualization and security; Considerably, can be expressed the principle idea of IoT is
and ii) Unique challenges: such as wireless sensor network promoting the communication between anything from
(WSN), Radio Frequency Identification (RFID), and finally anywhere at anytime through context-aware applications.
Quality of service (QoS) that is considered as a common Accordingly, IoT has relied on RFID and sensor network
factor between both general and special challenges. In technologies in the implementations. For instance, IBM
addition, this paper highlights the main applications of the company used IoT in Norwegian Sea oil platforms, by
IoT. deploying sensors at seabed that are used to collect real
information to make decision drill in the sea [3].
Keywords On the other hand, the IoT environment like many networks
IoT; heterogeneity; virtualization; WSN; RFID; QoS. suffering from the set of challenges which significantly
affect their performance some of them are common and
1. INTRODUCTION others, are special; the paper divides these challenges into
Today, we are living in the era of smart technologies which two categories, namely, i) General challenges: which
represents a "ubiquitous computing" or “web 0.3”. Internet include common challenges between IoT and traditional
of Things (IoT) has emerged strongly as a more prosperous network such as communication, heterogeneity, QoS,
area to express this kind of a new technology. It is not the scalability, virtualization, data mining and security; and ii)
first technology in this field, but also the cloud computing Special challenges: such as RFID and WSN.
technology has been used to represent the ubiquitous The main objective of this paper provides an overview
computing world. In the seventh in the series of ITU about IoT, its definition, its architecture, and discusses the
Internet Reports originally it was launched in 1997 under differences between IoT and the traditional Internet; then
the title “Challenges to the Network” [1], and it was first highlighting the challenges of IoT and the recent research
coined by Kevin Ashton in the RFID journal 1999 [2], In directions to solve them. Finally, the rest of this paper is
2005 this name was changed to “Internet of things“. The summarized as follows: section II, introducing an overview
vision of IoT according to Kevin’s vision was to enable about IoT concept, its history and its inception also
networked devices to propagate their information about discussing the differences between IoT and the traditional
physical world objects through the web. In recent years, the Internet; section III, focusing on the challenges and recent
most of the IoT proposed architectures are used, web research directions to address them and section IV,
semantic to publish information through the social reviewing a set of the most popular applications in IoT
networks; for instance, the iPhone has innovated service is
Nike + iPod to record information and published it on the 2. Related Work
social networks and the social network friends [3]. In this section, the paper seeks to offer a brief overview
about IoT, its definition, its history and its inception also
Actually, the definition of IoT varies based on who you highlight the architecture design of IoT that is relied on
talk, but formally, it can be defined as a dynamic global three dimensions called “IoT infrastructure”; and the final
network infrastructure with self-configuration and part in this section discusses the similarities and differences
interoperable communication. Simply, IoT means the between both IoT and traditional Internet.
ability to make 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

37
 International Journal of Computer Applications (0975 – 8887)
Volume 128 – No.1, October 2015

2.1. Definitions and History new space named “infrastructure of IoT”, which provides
In 1991, Mark Weiser has described the vision of the future support systems to serve the special things, which can
Internet under the name of “Ubiquitous Computing”. provide various services such as goods identification,
Through this vision he was focused on how to turn on the location identification and data protection. Fig 2 depicts the
smart livable environment in the presence of mobile phone three dimensions of IoT and relationship between them.
technology this provide a powerful multimedia system [6]. In this end, There are several approaches to build an
Kevin Ashton is a one of the pioneers talk about IoT [2]. architecture of IoT, the paper will focusing on two kinds
According to Atzori A.lera et al [7], classified IoT to three namely, architecture called “EPC global network“ and
paradigms namely, i) internet oriented (Middleware), ii) another called “Unite and ubiquitous IoTs or U2IoTs”, to
things oriented (Sensors), and iii) semantic oriented create an application on IoT, the architectural approach
(Knowledge). In 1999 Neil Gershenfeld was speaking about favored which based on an open architecture the EPC
similar things from the Massachusetts Institute of global network. The system designed by AutoID center for
Technology, MIT Media Lab in his book “When Things conveying the dynamic information about objects/things to
Start to think”. provide a history of the product movement for the
In 1999 Auto- ID labs and MIT sought to develop authorized users, the RFID technology plays a key role to
Electronic Product Code EPC, and use RFID to identify differentiate between these mobile objects, this system is
things on the network. In 2003-2004 the emergence of called “the EPC global network”. The IoT uses the EPC
projects serving IoT idea such as Cooltown, Internet0, and global network as a principle to design the architecture
the Disappearing Computer initiative, also IoT start to framework [10].
appear in book titles for the first time. RFID is deployed The future architecture of IoT seeks to achieve connection
was published on a massive scale by the US Department of between real-world, cyber-world and social world. Unite
Defense. In 2005 IoT entered a new level when published and ubiquitous IoTs or U2IoTs is considered as a different
its first report by International Telecommunication Union kind of IoT architecture, it’s used to integrate the physical
ITU. In 2008 a group of companies such as Cisco, Intel, world with the cyber world. The U2IoTs consists of a set of
SAP and over 50 other members of companies met to create heterogeneous systems, including unit of IoT to resemble
IPSO Alliance, to promote the use of Internet protocol (IP) human neural network that provides solutions to specific
and to activate IoT concept. In 2008-2009 IoT was applications; U2IoTs includes the industrial IoT, local
"Born" by Cisco Cisco Internet Business Solutions Group youth, national IoT, and global IoT which integration of
(IBSG) [8]. From the previous perspectives can be defined multiple Unit IoTs with ubiquitous features, and it is similar
IoT as a set of smart things/objects such as home devices, to the social organization framework. The main
mobile, laptop, etc., addressed by a unique addressing characteristics of U2IoT model are cyber, physical, social
scheme and connected to the Internet through a unified co-existence, connectivity and interactivity, space-time
framework this framework may be cloud computing. Fig 1 consistency and multi-identity status [10].
depicts IoT technology.
2.3 Differences between IoT and
Traditional network
In the beginning, the IoT technology has broken a lot of the
traditional ideas of network and started a new era of
telecommunication technology. Can be considered IoT as
an extension and expansion network based on the Internet;
but it is different from either traditional network or the so-
called Internet of people and WSN although considered as
backbone to build any IoT block.
The major equation to represent the IoT environment is
"IoT environment= Internet + WSN", it is a common
statement that uses to express the IoT environment. To
Fig 1: Internet of Things (IoT) technology analyze and judge the correctness of this statement, must be
determined the similarities and differences between IoT,
2.2 Architecture and Design Internet, and WSN according to table 1.
The best design of the architecture is a foundation stone to
build a privileged IoT system; this architecture helped to From the previous knowledge about the IoT environment
address a lot of issues in the IoT environment such as can be judged on this view, it's a wrong; because there are
scalability, routing, networking, etc.. Typically, the IoT two basic reasons for rejecting this view. First; IoT may not
architecture approach based on three main dimensions are: necessarily use IP in all cases for addressing things, because
i) Information items: it includes all items connected to IoT nature of IoT needs lightweight communication protocols,
environment may be sensing items, identifying items and the complexity of the TCP/IP protocol is not suitable in
control items; ii) Independent network: which includes particular, when works with the smart little things. Second,
several features such as self-configuration, self-protection, the IoT environment is mainly based on the connected
self-adaptation, and self-optimization; and iii) Intelligent smart objects unlike traditional network. That’s what makes
applications: which have intelligent behavior over the them move from only a mere extension of the Internet, also
Internet generally; the intelligent behavior may be the behavior of IoT depends on the creation of the
intelligent control, exchange data methods through network interoperable systems [10], based on these arguments, can
items, data processing, all the applications which are related be corrected the previous statement:
to the IoT can be classified according to these dimensions IoT= Internet + WSN+ Smart Items surrounded by
[9]. The intersection between these dimensions creates a Intelligent environment.

38
 International Journal of Computer Applications (0975 – 8887)
Volume 128 – No.1, October 2015

Finally, IoT supports a set of useful features such as

interoperability, self- configuration, self- adaptive and self-
protection. The intelligent environment is a way to ensure
the existence of a minimum level of the previously
mentioned elements within the network.

Fig 2: The three Dimensions of IoT

Table 1. The similarities and differences between IoT, Internet, and WSN [10]
Characteristics IoT Internet WSN
Comm. Protocol Lightweight Comm. (TCP/IP) Lightweight Comm.
protocols. protocols.
Scale degree of Area Cover wide area Cover wide area Cover local area
Networking Determine backbone Determine Self-organization
Approach backbone
Identify objects Must Can not Can
Type of nodes Active and passive Active Active
WSN+ dynamic smart Set of networks Dynamic smart
things+ Internet surrounded contains set of objects
Network design by intelligent environment Fixed objects

Behavior Dynamically Fixed Dynamically

Networking Time Timing synchronization Unlimited Unlimited

3. CHALLENGES AND RECENT networks (MANET) interconnected to fixed networks by

different gateway. In IoT, can’t be predicted where the
RESEARCH DIRECTIONS object moved, and the object may be needed to transmit
In this section, the paper discusses the bulk of popular from network to another. The biggest problem is in
challenges or general challenges of the IoT environment; it dynamic gateways change and the difficulty of Identifying
also displays the recent research directions for each topic. the location of things. The MANET consists of a number of
Finally, Table 3 reviews the recent research directions and self-organized mobile nodes or objects and it considered as
the proposed solutions for each one of them, in addition to a way to maintain a connection, additionally Multi-homed
table 4 which reviews the summary for the future research ad-hoc is seen as an extension to the existing infrastructure
topics in the IoT. in IoT.
3.1 Networking 3.2 Routing
Generally, the Networking issue has a great relevance in the Routing process means selecting the best path between the
Internet because of it includes some of the important factors source and the destination to complete the communication
which uses to manage networks. First of all, traffic and process successfully. There are various ways to determine
protocols that have a significant impact on the behavior of the best path based on the communication protocol type
the network, these points are mentioned in [11] D. Giusto et such as a number of hops, costs, and bandwidth. Can be
al. Sought to deal with networking challenges via mobile classified routing protocols into two main categories are: i)
Ad-Hoc Network. The authors have used mobile ad hoc Reactive protocols: the path is established after

39
 International Journal of Computer Applications (0975 – 8887)
Volume 128 – No.1, October 2015

transmission request is made, ii) Proactive protocols: initial contracted to delivery time service, this feature provides
path before the request is made. In [12], Sudip Misra et al. QoS.
proposed the protocol under the name of “fault-tolerant
routing protocol” for IoT. This protocol has designed by The middleware layer consists of three main layers are: i)
using learning automate (LA) and cross-layer concept. LA Service composition layer: the common layer on top of
dealing with optimization problems to choose optimal SOA middleware; it provides the functionality for
solutions, the need to cross-layer is saving energy of the composition of single service and builds specific
items of IoT (i.e. FRID). application. This layer concerned about services or
providing services only. Service composition architecture
3.3 Heterogeneity consists of individual architecture of the participating
The IoT environment is the best-known example to services, this architecture published SLA [14].
represent the heterogeneity issue because it contains a ii) Service Management layer: the layer allows
plethora of the different devices in their nature; the main management in IoT. Service management can be classified
objective of IoT is creating a common way to abstract the into two areas are: i) runtime: services that based on time as
heterogeneity of these devices and achieving the optimal a critical factor to implement them directly. ii) Design time:
exploitation of their functionality. In this vein, the services a part of maintenance lifestyle and service
researchers always seek to find an effective method to deal development [15]. Service management layer encompasses
with these devices regardless of their nature.In [13], C. G. is a set of services such as object dynamic discovery, status
Garcia et al. sought to introduce solutions to some of the monitoring, service policy enforcement, service Meta
IoT problems such as interconnection, heterogeneity, and model updates and service configuration, some of
generate an application that allows people to interconnect middleware include additional features that are related to
services over the Internet, these solutions are represented in: Qos management and lock management. It’s notable that
creating a domain specific language (DSL), graphic editor through the service management layer can develop new
and IoT platform Midgar software. For instance, over recent services during run-time [7].
years emerged a lot of applications used to solve the
heterogeneous objects problem over the Internet e.g. iii) Object abstraction: the need to object abstraction layer
WhatsApp, Skype and so on, it’s considered as a simple is summarized in vast and heterogeneous objects which
example to overcome this problem. The authors have scattered through IoT, layer organized harmonizing access
reviewed Midgar software to handle the heterogeneous to different devices with common language and procedure.
smart things through the IoT environment and DSL Object abstraction includes wrapping layer consists of two
software is designed for the specific purpose, the main idea sub-layers, interface sub-layer that management,
of this software is generating a domain which allows the incoming/outcoming messages and it provides an interface
interaction between things easy, regardless their nature. exposing the available method through a standard web
service interface. Second sub-layer is a communication sub-
The authors have reviewed Midgar software to handle the layer that implements logic at web service methods and
heterogeneous smart things through the IoT environment translate these methods at devices to communicate with
and DSL software is designed for the specific purpose is real-world objects [7].
generating a domain to enable the interaction between
things easily regardless their nature. The Midgar Software 3.4. Interoperability
used to avoid the complexities in traditional methods that Interoperability concept can be defined as the ability to
are used to handle this problem. In the future the create systems or devices cooperating with each other in an
connectivity is won’t be limited to the electronic devices, efficient way. In [16] Jussi et al. sought to use the semantic
but it will also include people this is will climb the problem; level interoperability architecture for pervasive the
for this reason can be considered Midgar as a first step in computing and IoT; the architecture is relied on the
this point.Moreover, IoT like any networks uses the service semantic information sharing solutions called “smart-M3”.
oriented architecture (SOA) approach to enhance the
heterogeneous resource behavior (i.e. Sensors and The principle idea of the proposed architecture relies on
Actuators) and it provides the highest level of flexibility dividing IoT environment into small spaces to facilitate
and scalability to the system in both the external integration their management process. A Semantic Information Broker
processes and the exchange processes within the SIB is used to provide methods for agents to share semantic
middleware. information with each other and also provides monitoring
and updating of the physical world in real time. The main
3.3.1. Middleware Layer observation of the architecture, performance after using the
The middleware layer is a software layer or a set of sub- agent interaction operations scale very well also enable
layers interposed between the technological and the interaction with the physical world in real time. The
application layer, it provides a standard way for architecture needs for tools the support development and
representation and communication. In general, the deployment of devices and applications in the future IoT
middleware layer supports the transparency concept that is systems.
used to hide all complex details from end user; actually the
transparency concept is a one of the most distinctive 3.5. Quality of Service (QoS)
features of distributed systems.The Service oriented Ideally, QoS is defined as “the amount of time that is taken
Architecture (SOA) is a common example of the to deliver the message from the sender and the receiver” if
middleware technology that is used to deal with IoT, SOA this time is equal or less than pre-specified time
allows to re-use and utilize real world service in a dynamic requirement the QoS is achieved. ITU re-defined QoS
way [7]. The SOA supports some of the services and one of concept as a degree of conformance of delivering service to
them is called “Service Level Agreement (SLA)”, which the user by the provider with agreement between them [17].
used to make an agreement between the service provider For QoS assurance, must cope with service models to
and service user, the most important feature of SLA is the determine which degree of QoS for each Internet service.

40
 International Journal of Computer Applications (0975 – 8887)
Volume 128 – No.1, October 2015

Moreover, Internet services can be classified according to systems and software like applications or services to run
Internet service models which considered as a supplement upon the same server through creating more than virtual
to provide the following: firstly, the ability to categorize machine inside the physical machine. The vision of this
Internet applications by priority; and secondly, determining concept helps to increase the performance of the network
QoS demands necessary to achieve user satisfaction. A via increasing utilization, maximizing scalability, saving
service models consist of three main models based on three cost, etc. [20]. Actually, there are three areas used to
factors, namely, a delay factor is concerned with time, represent the virtualization technology, namely, i) network
which can be classified into Hard Real Time (HRT), Soft virtualization, storage virtualization, and server
Real Time (SRT), and Non-Real Time (NRT), a critical virtualization.
factor is concerned with the kind of process/application (i.e.
Sensitive application or not) (yes/no), and finally, an In [3], S. Alam et al. have created the framework called “an
interactivity factor depending on the user’s subscription IoT Virtualization Framework based on Sensor as a service
(yes/no). According to table 2, the major types of Internet notion” this frame consists of three layers are: real world
service models are: an open service model, an Supple layer, semantic layer and virtualization layer in addition
service model, and a complete service model, the main separately database to record useful information. Primarily,
function of these categories helping to offer QoS the main challenges in the IoT environment can be
provisioning upon Internet services [18]In [19], Ming determined into three items namely, i) there is no registry
ZHOU et al. sought to find an efficient algorithm more mechanism, the framework possesses database to overcome
suitable with large-scale and real time in IoT, they made a this challenge; ii) the heterogeneity and discovery, the
comparison between the three common algorithms are: proposed framework sought to overcome of this issue
Integrated Linear Programming (ILP), Genetic Algorithm through the semantic approach to deal with the
(GA), and Backtracking Algorithm (BA) to find a suitable heterogeneity by providing a standard language called
algorithm that can deal with this context by efficiently. The Sensor Model Language “SensorML”; and finally iii) the
authors chose BA as a more suitable algorithm to serve this interaction between event and service in the IoT
idea because it’s a suitable to cover the large- scale area in environment are absent, the framework uses the
IoT and gave the good results in real-time compared with Virtualization layer to deal with this challenge by efficient
the counterpart of algorithms. way. The future directions of this framework divided into
two points are: enhancing its performance in real-time
Table 2. Internet Services models domain and improving micro-formats for advertising on
social networks.
IoT Delay Process/App. Interactivity
Models 3.8 Big Data
Big Data is a new expression to describe massive data
Open Not real Not mission Interactive whether structure or unstructured, which is difficult to deal
service time critical with traditional database methods and software techniques.
Supple Soft real Mission Application Simply, Big Data defined as a large volume of data [21].
service time critical dependent Dataset considered as a Big Data when it meets 4 V’s-
value, volume, velocity, and variety. Big Data attracts
Complete SRT/HRT Mission Not almost a new industrial field such as online social networks
service is critical interactive (Twitter, Facebook, and Instagram); the collection of data
depending through the social network is very huge, for example twitter
on app. in 2010 producing up 120 terabytes of data of the day [22].
IoT is considered as a good example of Big Data as the
3.6.Scalability amount of data which was collected from deploying sensors
Scalability is one of the most important challenges of IoT, through IoT environment was very large and
which means how to deal with the sustainable growth of the heterogeneous. The coupling between IoT and Big Data
Internet in an efficient manner. In the other words, was very strong [23].
“Scalability is the ability of a system or network to handle In [22], Chang Liu et al. sought to introduce a software
the growing scale of any environment without an effect on architecture addressed real-life based on the extraction from
performance”. Currently, the Internet comprises around 9 the SMARTCAMPUS project, this architecture supported
billion devices with a next era of the Internet which known the concept of Big Data into the IoT environment to deal
Web 0.3 or ubiquitous computing it is expected to reach 24 with data collected from sensors. This class of architecture
billion devices, the increasing of this number have a broad tackled some of challenges such as data storage, avoiding
impact on the performance of the network. processed bottlenecks and high throughput.
In [5], J. Gubbi et al. sought to apply the cloud computing
technology represented in Aneka software with IoT 3.9 Cloud Computing
environment. Generally, cloud computing provides a set of Cloud Computing and IoT are the most popular example to
features such as high storage resources, scalability, represent the ubiquitous computing field; but IoT is not
visualization platform and client delivery; plus the cost of popular like Cloud Computing, both use the distributed
each service based on pay-per-use. Aneka cloud computing computing concept. Cloud computing is a way to access
software provides a utilize storage and computing resources large amount of computational resources and supports a
of both public and private cloud. large number of users in a reliable and decentralized
manner; it’s also provide software cheaply. Cloud
3.7. Virtualization Computing consists of the three main layers are:
Virtualization is known as the ability to share hardware Infrastructure as a Service (IaaS), Platform as a Service
resources among multiple operating systems. The (PaaS) and Service as a Service (SaaS) each one provide
virtualization technology allows for the multiple operating significant features through the cloud data center. Cloud

41
 International Journal of Computer Applications (0975 – 8887)
Volume 128 – No.1, October 2015

computing is considered as a standard framework to represented the model of the proposed self-organized power
represent IoT, and both IoT and cloud computing possess a consumption approximation (SOPCA) algorithm. The
set of benefits and restrictions. IoT represents real world principle idea of this algorithm is the use of wireless
and small things, but it is limited storage in addition to connectivity between peer devices and servers, the devices
traditional problems in the network such as scalability and discover other devices to propagate energy sniffer agent
privacy; in other side, cloud computing has virtually (ESA).
unlimited capabilities and processing power [24]. The
integration of cloud computing with IoT became a very ESA locates devices and keeps estimating energy
important point of recent researches; to produce system able consumption. Source node also locates the other node by
to overcome the many challenges such as scalability, using GPS, ESA updates their internal variables based on
storage resource and virtualization; can be considered the the observation of local energy consumption then it moves
main objective of this integration is to leverage from cloud to another node. The mechanism of SOPCA algorithm
computing in processing power which need for sensors and avoids a re-routing between devices by using flags on the
other things [25]. individual devices. The authors used agent-based model
(ABM) to test the proposed algorithm over random
Recently, a lot of researches are mentioned integration network.
between cloud computing concept and IoT; for example not
as a limitation in [24], Alessio et al. sought to review the 3.11 Security and privacy
existing integration between IoT and cloud computing in The security rule aims to protect it from threats; these
the CloudIoT paradigm and illustrates the benefits from threats classify into two kinds are: the external threats such
them. First of all, the transparence which come with the as attacks on system form attackers and the internal threats
virtualization technology to hide the complexity of sensors represented in misuse of the system or information. There
from the end user; in addition to some of other features such are three main factors of security are: data confidentiality,
as the storage resources, the cloud computing concept privacy and truth. Data confidentiality guarantees only the
provides high capability to store large amount of data which authorized users to access and modify data, and it includes
collected from sensors; the big data considered as a new two aspects: first, access control mechanism and second, an
vision to restructure huge volumes of data through the IoT object authentication process [27]. Truth is guaranteed to
environment; the computational resources, one of the issues apply security rules into system and common example of
of IoT is limited processing resources Cloud Computing truth is digital certificates. Privacy is defined as a control
overcome this problem to improve scalability. access to personal data; and it allows keeping certain
information and data confidential; the features of privacy
The cloud computing also provides a lot of efficient are secrecy, anonymity and solitude [28]. The most current
solutions for the most issues of IoT and sought to offer the researches seek to increase and develop privacy in the
new visions for these issues such as providing ubiquitous applications, the Privacy Enhancing Technologies (PET)
access to sensor data through Sensing-as-a Service (SaaS); can be oriented to the subject, the object, the transaction or
enabling automatic control logics implemented in the cloud the system; it is used to protect identity over the Internet
computing through Sensing and Actuation-as-a-Service [28]. In the IoT environment the security and the privacy
(SAaaS); dispatches messaging services triggered by sensor are important to guarantee a reliable interaction between the
events through Sensor Event-as-a-Service (SEaaS); physical world and the cyber world [28].
enabling ubiquitous management of remote sensors through
Sensor-as-a-Service (SenaaS); enabling ubiquitous database In [29], Biplob R. Ray et al. proposed a framework
management through DataBase-as-a-Service (DBaaS); dependent on group approach and collaborative approach
providing ubiquitous access to any kind of data through called “a hybrid approach”, and used security check handoff
Data-as-a-Service(DaaS); providing ubiquitous layer-2 (SCH) with RFID. The SCH is a bet flags (0,1) (on/off)
connectivity to remote devices through Ethernet-as-a- help to keep track of the security state of the tag; also the
Service(EaaS); enabling ubiquitous access to policy and SCH allows the tag to take a shortcut to clear the security
identity management functionalities through Identity and check or re-clearance tag. Notable, the most of the existing
Policy Management-as-a-Service(IPMaaS); providing protocols which deal with RFID suffer from threats and
ubiquitous access to recorded video and implementing vulnerabilities such as insecure, inefficient identification,
complex analyses in the Cloud through Video Surveillance- throughput and inadaptability. The proposed protocol
as-a-Service (VSaaS). provides customization to ensure the adaptability of the new
efficient techniques. The development of RFID security
3.10 Power Consumption protocol makes the IoT more robust distributed structure.
The power consumption issue is a critical point in wireless
networks. Typically, the efficiency of the work of sensors 4. THE HOT TOPICS AND RELATED
depends on the lifetime of the battery. Nowadays the most CHALLENGES
of devices are equipped with sensors such as smart mobile The IoT consists of a host of other elements, which are
phone, tablet and laptop to deal with the modern considered as an extension for the general challenges of IoT
applications. For instance, the application of weather or it can be called "unique challenges". The section seeks to
prediction that is relied mainly on GPS to determine explain some of these elements in a nutshell.
location; once the GPS application is turned on during the
whole sensing procedure the battery may be drained very 4.1 Radio frequency Identification
quickly.
(RFID)
In [26], K. Batool et al. have dealt with the power RFID is a breakthrough in embedded communication and
consumption issue by using sniffer agents. The paper was WSN, RFID is used to generate a unique ID for the object
autonomously presented on mobile energy consumption of in WSN. It consists of two parts are: passive RFID: which
networked consumer devices to determine the used to power of the reader’s interrogation signal to
approximation of the current power requirement. The paper communicate the ID to the RFID, and access control

42
 International Journal of Computer Applications (0975 – 8887)
Volume 128 – No.1, October 2015

application as well. Active RFID: readers have their own According to [5] can be divided the IoT applications into
battery supply and instantiate the communication. RFID four categories can be summarized in the following: i)
uses Ultrawide Bandwidth (UWB) technology to enhance Personal and home uses WiFi as a backbone providing
RFID performance in a specific IoT application field. UWB higher bandwidth data transfer as well as higher sampling
is a technology allows the next generation of RFID to rate, the healthcare sector is considered as the most popular
overcome many of the current restrictions in current RFID example of this category, ii) Enterprise, the information in
such as low security, reduce area, and sensitivity to this category may be collected from the networks, the
interferences. The RFID contains three key elements are: environment monitoring such as video surveillance is a first
the RFID tag or transponder that carries object, the RFID common example of this category followed by smart home
tags reader or transceivers that read and write tags and and smart environment generally, iii) Mobile, the sensor
back-end database. information can be obtained from large scale WSN for
online monitoring of travel time the popular example of this
4.2 Wireless Sensor Networks (WSN) category is transportation; lastly, iv) Utilities, the
WSN is an important part of IoT, it’s considered as a core information can be obtained from networks to achieve
to build the IoT block, it consists of a group of specialized service optimization and power consumption,Typically, the
sensor data are shared among sensor nodes with main target of companies which are using this type of
communication infrastructure for monitoring some of applications is reducing cost and maximizing profit. The
events or states of objects such as temperature, sound, best example of this kind is smart grid, smart metering and
pressure, etc. these sensor nodes work autonomously and smart water and quality of drinking water. This section will
can be linked between them by self-organizing. Notable, be reviewed the most popular examples of each type of
WSN support the distribution concept between sensor these mentioned categories earlier, also table 5 summarizes
nodes, and each sensor network includes some of elements the categories of the IoT applications.
such as radio transceiver with an internal antenna or
connection to an external antenna, a microcontroller, an Healthcare: the IoTCloud paradigm [33] has been widely
electronic circuit for interfacing with the sensors and an used in the healthcare sector to diagnosis, treatment and
energy source [30]. The section seek to show the tracking the status of the patients remotely. The paradigm
components that make up WSN in the following: must serve four pivots namely, i) tracking: is a function
aimed to identify the patient in motion, ii) identification and
 WSN hardware contains a sensor interface, processing authentication: identification aimed to reduce mistakes in
units, transceiver units and power supply. diagnosis and authentication used to meet with security
requirements, iii) data collection: usually aims to reduce
 WSN communication stack the nodes deployed in an processing time and it is related to integrating RFID
ad-hoc manner for most applications. technology with other health information, lastly iv) sensing:
 WSN middleware is mechanism to combine cyber used to provide real time information about the patient [32].
infrastructure with service oriented architecture (SOA). The issues in this area are defined as: control, security,
SOA is an architectural style that enables the heterogeneity, interoperability and streaming QoS.
composition of applications by using loosely coupled Smart environment (i.e. Smart city, smart home): The
and interoperable services. key idea to gain the smart environment such as smart city or
 Secure data aggregation; it is a very important to ensure smart home while maintaining the level of service without
reliable data collected from sensors. degradation has relied on the integration between both IoT
and cloud computing; over recent year’s escalation of
4.3 Addressing Schemas and attention toward this integration. Generally, The IoT suffers
from a host of the thorniest issues, first of all, heterogeneity
Communication Protocols of objects; there are many middleware technologies are
The Internet contains a large amount of objects which designed to deal with this type of issues, such as RFID
needing to determine their location to complete the middleware and WSN middleware. On the other hand,the
communication process in an efficient manner, it’s cloud computing provides scalability and hide complexity
considered as a main objective of addressing process. of sensors from end users by using the virtualization
Generally, Internet Protocol (IP) used to identify objects technology. The most popular challenges in this type of
through the Internet; there are two versions of IP, namely, i)
application are limited to security and real-time
IPv4, which uses 32-bit addresses ( addresses) to applications.Video surveillance: is an intelligent video to
identify hosts or objects through the Internet, this process monitor the object behavior and activities, it has become an
considered as limited to some extent; and ii) IPv6 is the important tool of security-related application and it can be
latest version of the Internet protocols uses 128-bit ( considered as an alternative to self-contained management
addresses), It covers a large area more than IPv4 [31]. The systems, and the complex video analytics requires Cloud-
RFID technology is a way to identify things into WSN and based solutions, and it can be used to satisfy the need to
IoT, it generates unique ID to identify the smart objects. increase volume of storage capacity of media [32].
Also “IPv6 over Low-power wireless Personal Area Automotive and smart mobility: the aim of this function
Network (6LoWPAN)” IEFT group aims to make IPv6 concentrated on the improvement of transportation and
compatible with low capacity devices. transport through increasing safety road, reducing
congestion, and control traffic. Can be considered the
5. APPLICATIONS integration between Cloud Computing and IoT as a
Recently, the IoT technology has appeared as a one of the promising opportunity to find effective solution is
basics in our lives because it touches a lot of the important characterized by high performance, security and lowest
sides such healthcare, smart water, transportations, cost. The challenges in this type are: identification,
surveillance and so on. Moreover, there are many heterogeneity of sensors, scalability and dynamic behavior
applications which have emerged to serve this concept. of objects.

43
 International Journal of Computer Applications (0975 – 8887)
Volume 128 – No.1, October 2015

Smart energy and smart grid: the biggest problem in the physical worlds, cyber-world and social world. Lastly, in
sensors world is power consumption, the battery in sensors this section is pointed out to the differences between both
may drain very quickly, and both the IoT and Cloud IoT and the traditional network.Section II, reviewed the
Computing provide intelligent solutions to manage energy main general challenges which had a significant impact on
distribution and consumption in heterogeneous the performance of IoT such as communication,
environments. networking, Qos, scalability, virtualization, big data,
6. CONCLUSIONS AND FUTURE heterogeneity and security; this section sought to illustrate
and provide the recent solutions for each element of these
DIRECTIONS challenges. Also, this section reviewed another kind of
IoT is a one the main techniques that is used to express the challenges that are called "unique challenges" under subtitle
ubiquitous computing approach, but it still not popular like "The Hot Topics and Related Challenges"
the cloud computing technology. This paper has sought to
highlight the IoT concept in general through the three which includes more related challenges to the IoT
sections namely; section I, reviewed an overview about the environment.Finally, section III, reviewed a set of the
IoT concept via highlighting its history and its inception popular applications which are offered by IoT and the
since 1999 at the hand of Kevin Ashton who considered as IoTcloud paradigm such as healthcare, smart city, smart
one of the pioneers who talked about IoT and even Cisco grid, smart transportations, etc.. Based on the above, can be
company now. Then it has reviewed the main idea to design considered the IoT environment as a rich search point, and
the IoT structure that relies on the integration between three flourishing area to the research in particular in the
dimensions are: information items, independent network integration topic with cloud computing, which provides the
and intelligent applications. Accordingly, the future of the new sceneries to handle the smart services and applications.
IoT structure relies on the integration among real or
Table 3. Summary for the recent research directions of IoT
Research area Techniques Solutions Open area
Networking [11] MANET The authors have used the MANET as a  Improving Ad-Hoc network.
way to maintain connection between  Network technologies.
things.  RFID.
 Communication protocols.
Learning automates The authors proposed a protocol called  Improving the proposed
Routing [12] (LA) & Cross-layer. “fault-tolerant routing” protocol to serve algorithm to cover the wide
the IoT environment. range of the application
domains
The authors have been sought to improve  Domain Specific language and
Heterogeneity [13] Midgar Software the interaction between objects or things graphic editor to generate
via the graphic editor that generates model smart objects.
defined by a Domain Specific Language.  Insertion of boxes in the
graphic editor with support for
data analysis and fuzzy logic.
 Scalability of IoT platform.
 Security and privacy.
Semantic The authors have been sought to build the  The architecture needs tools
Interoperability [16] Interoperability semantic interoperability architecture to that support development and
Architecture access the information easily, also the deployment of devices and
paper used the monitoring and updating applications into the future IoT
events of physical world in real time. systems.
QoS [19] BT, IP & GA The authors have been reviewed  Decreasing steps of BT
comparison between three algorithms to algorithm for calculating QoS.
determine Qos metrics for composes
service. BT algorithm is the most
appropriate to IoT environment more than
ILP & GA, because it suitable to serve
both high scale of service and real-time
application.
Aneka Hybrid cloud The authors have used cloud computing  Cloud computing.
Scalability [5] computing (private technology with IoT environment to  Energy efficient sensing.
cloud + public cloud) improve scalability and provide storage  Architecture of IoT.
resources.  Data mining.
 Secure reprogrammable
networks and privacy.

IoT Virtualization The authors have created a new  The development of IoT
Virtualization [3] Framework based on framework based on SenaaS notion named framework services micro-
SenaaS technology. it “IoT Virtualization Framework”, the formats for advertising on
main objective of this frame is reusability social networks.
of the sensor information via web browser.  Improving the performance of
an IoT framework at real time.
The authors have sought to improve the  Cloud Computing.
software architecture named "real-life"  Scalability/elasticity.
Big Data [22] SMARTCAMPUS. based on extracted from the  Computation time.
SMARTCAMPUS project to handle Big  Security.

44
 International Journal of Computer Applications (0975 – 8887)
Volume 128 – No.1, October 2015

Data in the IoT environment.

Cloud Computing [24] CloudIoT Paradigm. The authors have highlighted the  Standardization of framework.
integration between the cloud computing  Power consuming.
and IoT, also reviewed the previous  Fog Cloud.
literatures about them.  Complex data mining.
Power Consumption [26] Self-Organized Power The authors have used a self-organizing  The SOPCA can be further
Consumption for dynamic approximation of power explored by the evaluation of
Approximation consumption to create (SOPCA) flooding as well as by using
(SOPCA) Algorithm algorithm. Also the authors have used artificial intelligence
Agent-based model to test this algorithm. algorithms.
The authors have used SCH to improve  Improving and addressing the
Security [29] SCH and ensure security of RFID system. integration of a tag tamper
resistance mechanism.

Table 4. Summary for The future research topics in the IoT.

Research Directions Info. Open directions
Discovery / identification. The ability to create a standard addressing  Mapping digital & real.
schemes more efficient and reliable,  Device discovery.
additionally, it must be provided  Semantic search.
convergence of IP and RFID.  Universal authentications
mechanism.
Design / architecture The IoT architecture uses an open  Cloud computing.
architecture approach to maximize  Ad-Hoc networks.
interoperability and handle the  Adaptive and context-aware
heterogeneity. architecture.

Networking The Networking issue includes both  Ad-Hoc networks, Hyper

routing and communication protocols. it networking.
seeks to improve the performance of  Self-configuration.
network through congestion management  Virtualization technology
and traffic. (location transparency).
 Self-organization networks.
Standardization The aim of standardization issue is the  IoT standardization.
ability to create a standard interface or  Cloud computing.
standard framework to achieve highest  Semantic web.
level of interoperability between devices.  Semantic interoperability.

Energy Consumption Typically, the efficient of sensors rely on  Semantic interoperability.

the lifetime of battery. Recently, the  Micro battery technologies.
micro power technology used to address  Energy harvesting.
this issue.

Security The aim of security is protecting data  Security for cloud computing.
from unauthorized users. Generally, the  Security for semantic web.
security issue contains three levels are:  Improving encryption methods.
confidentiality, trust, integrity  Privacy policies and trust.

Table 5. The categories of IoT applications

Categories Services Connectivity Challenges
Personal/Home healthcare WiFi, 3G, 4GLTE  Heterogeneity.
 Interoperability.
 Control methods.
 Real-time.
 Security.
 QoS.
Enterprise Smart Cities WiFi, 3G, 4GLTE, Satellite  Scalability.
 Identification and discovery.
Smart Environment WiFi, 3G, 4GLTE, Satellite  Heterogeneity.
 Virtualization.
Video Surveillance WiFi, Satellite  Control methods.
 Real-time.
 Big data.
 Power consumption.
 Security.

45
 International Journal of Computer Applications (0975 – 8887)
Volume 128 – No.1, October 2015

Mobile Smart Transportation / WSN, Satellite  Identification and discovery.

 Control methods.
Smart traffic  Scalability.
 WSN
 IoT architecture and design.
 Real-time.
 Cost.
Utilities Smart Grid WiFi, Satellite & Celular  Power consumption.
 Scheduling (Load balance).
Smart Energy  Real-time.
 Scalability.
Smart Water  Control methods.
 Cost.

2014. Midgar: Generation of heterogeneous objects

7. REFERENCES interconnecting
[1] InternetofThings,2014http://www.itu.int/en/publications/ applications.DomainSpecificLanguageproposal for Intern
gs/pages/publications.aspx?parent=SPOLIR.IT2005&me etof Things scenarios”, Computer Networks, 143-158.
dia=paper#
[14] SOA, 2014 http://en.wikipedia.org/wiki/Service-
[2] InternetofThings,2015http://www.rfidjournal.com/article oriented_architecture.
s/view?4986.
[15] Gopala Krishna Behara, Srikanth Inaganti. 2007
[3] Sarfraz Alam, Mohammad M. R. Chowdhury, Josef Noll, Approach to Service Management In SOA Space,
2010 SenaaS:An Event-driven Sensor Virtualization BPTrends.
Approach for Internet of Things Cloud, Networked
Embedded Systems for Enterprise Applications [16] Jussi Kiljander, Alfredo D’elia , Francesco Morandi ,
(NESEA), 2010 IEEE International Conference on, 1-6. Pasi Hyttila, Janne Takalo-Mattila , Arto Ylisaukko-Oja ,
Juha-Pekka Soininen, Tullio Salmon Cinotti. 2014
[4] Andras Kalmar, Rolland Vida, Markosz Maliosz, 2013 Semantic Interoperability Architecture for Pervasive
Context-aware Addressing in the Internet of Things Computing and Internet of Things, 856 – 873.
using Bloom Filters, Cog l nfoCom 2013· 4th IEEE I
nternational Conference on Cognitive I nfocommun [17] ITU-T E.860. Framework of a service level agreement
ications' (Dec. 2013) 487 - 492. 2002,2014.https://www.itu.int/ITUD/tech/events/2005/N
airobi2005/Presentations/Day2/Nairobi_2_2_4.pdf.
[5] Jayavardhana Gubbi, Rajkumar Buyya, Slaven Marusic,
Marimuthu Palaniswami, 2013 Internet of Things (IoT): [18] Marie-AurelieNef, Leonidas Perlepes, Sophia
A Vision, Architectural Elements, and Future Directions. Karagiorgou, George I.Stamoulis, Panayotis K.Kikiras,
Future Generation Computer Systems, 1645-1660. 2012 Enabling QoS in the Internet of Things, CTRQ:
The Fifth International Conference on Communication
[6] Daoliang Li, Yingyi Chen, Oct. 2010, Computer and Theory, Reliability, and Quality of Service, 33-38.
Computing Technologies in Agriculture. Springer, 24-31.
[19] Ming ZHOU, Yan MA, 2013 QoS-aware computational
[7] L. Atzori, A. lera, G. Morabito, The Internet of Things: method for IoT composite service, The Journal of China
Survey. Computer networks, 2787–2805. Universities of Posts and Telecommunications, 35-39.
[8] Internet of Things, 2014 http://postscapes.com/internet- [20] Lenn-Wei Lin, Chien-Hung Chen, Chi-YiLin, 2013
of-things-history. Integrating QoS awareness with virtualization in cloud
[9] Huansheng Ning, Hong Liu, 2012 Cyber-Physical-Social computing system for delay-sensitive application, Future
Based Security Architecture for Future Internet of Generation Computer Systems, 478–487.
Things, Advances in Internet of Things, 1-7. [21] Bigdata,2014http://www.webopedia.com/TERM/B/big_d
[10] Nihong Wang, Wenjing Wu, 2012 The Architecture ata.html.
Analysis of Internet of Things, Computer and Computing [22] Chang Liu, Chi Yang, Xuyun Zhang, Jinjun Chen, 2014
Technologies in Agriculture V IFIP Advances in External Integrity Verification for OutSourced Big Data
Information and Communication Technology, 193-198. in Cloud and IoT: A Big Picture, Future Generation
[11] D.Giusto, A.lera, G.Morabito, L.Atzori (Eds.), 2010 Computer System.
Objects Communication Behavior on Multihomed [23] Cyril Cecchinel, Matthieu Jimenez, Sebastien Mosser,
Hybrid Ad Hoc Networks, Springer, 3-11. Michel Riveill, 2014 An Architecture to Support the
[12] Sudip Misra, P. Venkata Krishna, Harshit Agarwal, Collection of Big Data in the Internet of Things, Services
Anshima Gupta, Mohammed S.Obaidat, 2012 An (SERVICES), 2014 IEEE World Congress on, 442-449.
Adaptive Learning Approach for Fault-Tolerant Routing [24] Alessio Botta, Walter de Donato, Valerio Persico,
in Internet of Things. IEEE Wireless Communications Antonio Pescape, 2014 On the Integration of Cloud
and Networking Conference: PHY and Fundamentals, Computing and Internet of Things, 2014 International
815 – 819. Conference on Future Internet of Things and Cloud
[13] Cristian González García, B. Cristina Pelayo G-Bustelo, (FiCloud), 23 – 30.
Jordán Pascual Espada, Guillermo Cueva-Fernandez,

46
 International Journal of Computer Applications (0975 – 8887)
Volume 128 – No.1, October 2015

[25] Noura Alhakbani, Mohammed Mehedi Hassan, M. [29] Biplob R. Ray, Jemal Abawajy, Morshed Chowdhury,
Anwar Hossain, and Mohammed Alnuem, 2014 A 2014 Scalable RFID security framework and protocol
Framework of Adaptive Interaction Support in Cloud- supporting Internet of Things , Computer Networks, 89-
Based Internet of Things (IoT) Environment, G.Fortino 103.
et al. (Eds.): IDCS 2014, LNCS 8729, 136-146.
[30] Wirelesssensornetwork,2014http://en.wikipedia.org/wiki/
[26] [26] Komal Batool, Muaz A. Niazi, 2015 Self-Orgnized Wireless_sensor_network.
Power Consumption Approximation in the Internet of
Things, 2015 IEEE International Conference on [31] IP address, 2014 http://en.wikipedia.org/wiki/IP_address.
Consumer Electronics (ICCE). [32] Xiang Sheng, Jian Tang , Xuejie Xiao ,Guoliang Xue,
[27] Rolf H. Weber, Romana Weber, 2010 Internet of Things, 2013 Sensing as a Service: Challenges, Solutions and
Springer, 41-68. Future Directions, IEEE Sensors Journal, 3733 – 3741.

[28] Huansheng Ning, 2013 Unit and Ubiquitous Internet of [33] OpenSourceIoTCloud,2015https://www.sites.google.com
Things, CRC Press. /site/opensourceiotcloud/installation-guide,.

IJCATM : www.ijcaonline.org 47

View publication stats