CSIT (March 2017) 5(1):87–95

DOI 10.1007/s40012-016-0136-6

SPECIAL ISSUE ICAC 2016 OF CSIT

Internet of Things: challenges and research opportunities

Md. Iftekhar Hussain1

Received: 26 November 2016 / Accepted: 28 November 2016 / Published online: 28 December 2016
Ó CSI Publications 2016

Abstract Faster development of sensor and network suitable communication technologies such as Wi-Fi, BLE,
technologies is facilitating immense deployment of Internet ZigBee, NFC, etc. In processing the huge amount of data
of Things (IoT) towards creating a smart world. In IoT, a generated and serving a large number of users in IoT, a cloud
massive number of heterogeneous resource–constraint is essential. The cloud enables the applications to work
devices communicate with each other without any human anytime and from anywhere. The IoT applications are
intervention and generate a huge amount of data. Unique expected to work autonomously which can adapt and react
research challenges posed by IoT are fascinating the intelligently over different situations, and support for easy
research community. This paper presents some of the integration. However, due to the wide heterogeneity and the
critical issues along with state of the art solutions towards resource–constraint nature of participating devices, these
them. In-depth discussion is provided on various key issues expectations remain as a challenge.
like heterogeneity and interoperability, scalability, QoS, The integration of heterogeneous network systems have
and security. Directions for further researches in those become the driving source of network alteration and has
areas are also pointed out. proposed various novel concepts such as Cloud of Things
(CoT), Web of Things (WoT) and Social Internet of Things
Keywords Internet of Things
Scalability
(SIoT) [43]. Thousands of IoT applications can be recog-
Heterogeneity
Interoperability
Quality-of-Service nized in each domain and new ones emerging every day,
requiring a strong Interoperability (IoP) among things. IoP
further concerns various aspects such as security and pri-
1 Introduction vacy, standardization, etc. This brings new challenges
driving research and innovation in industry and academia
Internet of Things (IoT) is the network of physical objects over the last decade. Accommodating a very large number
planted with sensors, actuators, RFIDs, software, and of heterogeneous resource constraint devices in IoT also
connectivity to enable it to interact with people and other invites attention of research fraternity.
connected devices in achieving some common goals This article presents a holistic perspective on the IoT
[37, 40]. The number of devices connected in the IoT is concept and development, enabling technologies, and
increasing very fast. As such, it has been estimated that critical research challenges. Considering the limited
around 50 billion connected devices will be there by 2020 involvement of the networking and communications sci-
[12]. entific communities in such an important development, I
The devices used for sensing, actuating, and monitoring hope this work can boost more and more research in the
purposes are connected to the Internet by highlighted issues and beyond.
The rest of the article is divided into following sections.
Section 2 describes the background of IoT including its
& Md. Iftekhar Hussain
working principle and the protocol stack. Several critical
[email protected]
research scopes and challenges along with their state of art
1
North-Eastern Hill University, Shillong, India are discussed in Sect. 3. Section 4 concludes the article.

123
88 CSIT (March 2017) 5(1):87–95

2 Background of Internet of Things (IoT) types of applications is desirable in order to support co-
building and openness in IoT. Figure 1 shows a typical
As the IoT encompasses a wide variety of concepts, a brief architecture of IoT. The entire architecture is broadly
background of the IoT components is presented in this classified into edge and platform tiers. The components of
section. IoT in edge tier are the end devices (sensors, RFID,
Camera, etc.), gateways, and sensor networks which are
2.1 What is IoT? connected to the core via access networks. The platform
tier contains the middleware, server and storage, and core
The basic idea of IoT is to equip the physical objects around services for device management, data management, real-
us to sense the surrounding information, provide seamless time processing, analytics engines, and so on. A gateway is
communication and contextual services provided by them. a device that has a short range of access link on one side,
The IoT is simply the network of interconnected things and a wide range of access link on the other side. It is a bit
which are embedded with sensors, software, network con- similar to our home router, which has a local network
nectivity, and embedded devices that enable them to collect access towards our home computers and other IP-enabled
and exchange data making them accessible over the Internet. devices, and WAN access toward our Internet service
IoT brings useful applications like home automation, smart provider. The gateway node connects the two dissimilar
health monitoring, security, automated devices monitoring, networks that exist between IoT and Internet devices.
and management of daily tasks. Every sector like energy, Sensors collect real-time data like video in case of
computing, management, security, transportation are going intrusion detection system, audio or text in case of forest
to be benefited with this new paradigm. fire detection system. Router/Relay nodes work as for-
IoT enables physical objects to see, hear, think and per- warding nodes to transmit all the information associated
form tasks by having them talk together, to share information with the different measured parameters. Both sensor and
and to coordinate decisions [3]. Development of sensors, routers are configured to send all the information to the
actuators, smart phones, RFID tags makes it possible to cloud or server. The sensor device can itself process the
materialize IoT which interact and co-operate each other to information and take action or it may send the collected
make the service better and make accessible at any time, information to the cloud through the Internet.
from anywhere using any network. Wireless sensor tech-
nology allows objects to provide real-time environmental 2.3 Protocol stack
information and context. IoT allows objects to become more
intelligent which can think and communicate among them. The protocols required for effective communication in IoT
In the vision of IoT, ’’things’’ are classified with three can be positioned in a multi-layer stack as shown in Fig. 1.
scopes: people, machine (sensor, actuator, etc.) and infor- Primarily, data link, network, and application layer proto-
mation (clothes, food, medicine, books, etc.) [35]. These cols are available in the literature. The protocols or tech-
should be identified at least by one unique way of identi- nologies are mainly characterized by the light weight
fication for the capability of addressing and communicating property but also focus on energy-efficient and scalable
with each other and verifying their identities. The identi- solutions for resource–constraint devices (Table 1 ).
fiable ’’things’’ are known as ’’objects’’ in IoT. The major Data link layer protocols are used to allocate channels or
characteristics of IoT objects [41] are: (a) ability to sense or medium to stations for coordinating data transmission
actuate, (b) small size, (c) limited capability, (d) limited
energy, (e) connected to physical world, (f) intermittent
IoT Nodes Remote
connectivity, (g) mobility, (h) managed by devices, not
people etc.
Da
ta
The scope of objects ranges from small to large. It is Co d
mm man
and Internet Com
difficult to connect to power supply for the moving objects IoT Hub
Data
all the time. So they need to operate with a self-sustaining
Analytics

Command
energy source. Therefore, energy efficient communication Co
and mm
mm and
mechanisms are essential for IoT. Co
ta Da
Da ta

2.2 How it works?

Cloud and Server
IoT comprises of a wide variety of devices connected to
Edge tier Platform tier
each other with reference to their functional organization
and configuration. An uniform standard architecture for all Fig. 1 A typical architecture of IoT

123
CSIT (March 2017) 5(1):87–95 89

among smart devices. ZigBee [20] technology is based on In application layer, several new protocols are intro-
IEEE 802.15.4 standard which is designed to provide a duced to adapt with high volume and vast network of IoT
wireless data solution characterized by secure and reliable devices. Machine-to-machine (M2M) communication
wireless network architectures. ZigBee is one of the most Message Queue Telemetry Transport (MQTT) [15] is
widely used protocol standards of IoT which enables smart designed for IoT devices of small size that have low
objects to work together. The ZigBee devices are low latency bandwidth, high cost, low processing power and unreliable
devices and more responsive compared to Bluetooth devices networks such that they can communicate seamlessly
as it takes only 30 ms to go from passive to active mode among them. It provides a one-to-one classic M2M appli-
whereas Bluetooth takes approximately 3 seconds. BLE [27] cation in IoT. Some other suitable protocols are Constraint
is a short-range communication protocol that can save ten Application Protocol (CoAP) [32], Data-Distribution Ser-
times more energy than classic bluetooth by using a con- vice protocol (DDS) [36], Extensible Messaging and
tention-less MAC with low latency and fast transmission. Presence Protocol (XMPP) [28], and Advanced Message
The traditional IEEE 802.11 (Wi-Fi) standards are not suit- Queuing Protocol (AMQP) [39]. CoAP is a web service
able for IoT application due to its low scalability, frame oriented application which is designed to facilitate com-
overhead, and power consumption. The IEEE 802.11 munication over resource–constraint electronic devices on
working group started a task group to develop 802.11ah [34] the Internet. CoAP provides simplicity, low overhead and
standard which supports low overhead and power friendly multicast support which can easily be translated into HTTP
communication which is suitable for IoT. To collect envi- for simplification.
ronmental information through sensors, IoT uses RFID To provide human-to-human (H2H) communication in
reader in many cases. It is an identification technology that IoT, another XML-based instance message oriented pro-
can automatically identify target devices and can collect data tocol called XMPP [28] is proposed. It is designed to
through its radio frequency signal without human interven- connect people to other peoples by instant messaging. The
tion. It has the ability to uniquely identify the objects which Data-Distribution Service (DDS) protocol [36] offers
are able to give the object location and can manage the interoperability between IoT connected smart devices and
correct communications and information processing. provides scalability, performance enhancements for IoT
The network layer protocols provide an abstract idea of applications. It supports multicast as well as multiple
the smart world devices. In addition to the traditional net- transport protocols such as TCP/IP and UDP/IP for pro-
work layer protocols like IPv4 and IPv6 several new routing viding high throughput and low latency. For its distributed
protocols such as 6LoWPAN [24], RPL [42], CoRPL [2], processing nature it can directly be connected to sensors,
CRB-RPL [44] are proposed for IoT. To apply Internet smart devices without using any centralized server.
protocol in such a huge number of resource–constraint Advanced Message Queuing Protocol (AMQP) [39] is
devices, IPv6 [28] over 6LoWPAN [24] is preferred. Aiming designed for IoT which is mostly suitable for server-based
to provide a suitable routing solution, the Routing Over Low analysis and control plane. It provides interoperability and
power and Lossy Networks (ROLL) working group of the reliability for the queuing messages between servers.
Internet Engineering Task Force (IETF) designed a new IPv6 AMQP also feasible for flexible routing, publish-subscribe
Routing Protocol for Low power and Lossy networks messaging, transaction and security.
(RPLs). It provides efficient routing paths for multi-hop
mesh technologies in low power and lossy networks. RPL
used link costs and node information as routing metric. Node 3 Challenges and research opportunities in IoT
information included available energy resources, workload,
throughput, latency, reliability, and so on. The RPL is further As the IoT is the integration of heterogeneous technologies
modified in CoRPL [2] and CRB-RPL [44] which use Cog- that are used to sense, collect, act, process, manage,
nitive Radio (CR) for decision making. transmit, notify, and store data, there are many research

Table 1 IoT protocol stack

Layer Protocols

Application layer CoAP [32], MQTT [15], SMQTT [33], XMPP [28], DDS [36], AMQP [39], IETF CORE, HTTP, SSH
Network layer 6LowPAN [24], 6TiSCH [10], IPv4, IPv6 [28], RPL [42], CoRPL [2], CRB-RPL [44], IETF ROLL
Data link layer 802.15.4, RFID, 802.11g/ac/ad/ah [34], BLE [27], LTE-A [13], Z-Wave, NFC, ANT? [45], LoRaWAN [21], SigFox [22]

123
90 CSIT (March 2017) 5(1):87–95

issues and challenges which have cropped up spanning 3.2 Interoperability (IoP)
several research areas. In this section, the key research
issues and challenges are explained showing future In IoT, each device would be subject to different conditions
research directions in each area. such as energy constrained, communication bandwidth
requirements, computation and security capabilities [38].
3.1 Heterogeneity Things could be made by various manufacturers that do not
necessarily follow the common standard. Moreover, devi-
Heterogeneity is one of the critical issues of IoT. Device ces may also operate using a variety of communication
heterogeneity can be in terms of different technologies, technologies. These technologies do not necessarily con-
sensing, various software, and processing strategies used in nect IoT devices to the Internet in the same way a typical
IoT. In traditional computer environments, computer computer device usually do. Figure 2 explains a typical IoP
devices are treated equally when connected to the Internet. scenario in the context of IoT. Here, various devices trying
Their functionalities vary depending on how the users use to communicate among them using heterogeneous com-
them. However, when we talk about IoT, each device munication technologies (RFID, Bluetooth, Wi-Fi, ZigBee,
would be subjected to different conditions such as power NFC, etc.). For example, a Bluetooth mouse cannot com-
consumption, processing capabilities, and communication municate with a laptop which is using Wi-Fi technology.
bandwidth requirements [11]. The heterogeneity of IoT To solve these issues we can use a smart gateway or IoT
may be due to- cloud to provide IoP among them.
The heterogeneous devices which are broadly used in
1. Operating conditions The sensor devices operate in IoT need to talk and work together. To achieve this, dif-
different conditions like- temperature, pressure, and ferent IoT systems need to be integrated. However, at the
voltage. technical level, interoperability still represents a significant
2. Functionality The IoT devices may either deliver data barrier. Up to 60% of the values that IoT systems might
periodically or on demand basis. disclose is currently locked by a lack of interoperability
3. Resolutions The objective of IoT devices may be [11]. As several heterogeneous systems need to commu-
tracking, monitoring, actuating, etc. nicate, IoT faces various IoP challenges before being able
4. Hardware platform The hardware platform varies to create real domain services with seamless communica-
according to their architecture and design. Based on tion of devices and data.
this, the supporting operating systems and applications Different categories of IoP issues are appearing in the
are also different. context of IoT. Various types of IoP issues that are needed
5. Service pattern The pattern of IoT services may differs to be addressed in supporting seamless and heterogeneous
in generation rate, packet size, etc. communications in the IoT are shown in Fig. 3.
6. Implementations Different programming languages are
used to develop IoT applications using different 1. Technical Interoperability It is the incompatibility of
operating systems, such as Android, iOS, etc. the communication technologies and protocols that are
7. Interaction modes The interaction between IoT devices used to exchange information. This type of interoper-
and the remote user can be request/response or ability provides only low level integration which
command type. guarantees the correct transmission of bits among
different heterogeneous systems.
When a plethora of IoT applications towards improving
quality of human life are being developed, various new
Smart Gateway / IoT Cloud
devices, protocols, network connectivity methods and
resulting application models are evolving. Sticking to sin- RFID
Interoperability
NFC
gle protocol standard in IoT is not an easy task. Moreover,
even within one standardized protocol suite there are a
number of different application domains and communica-
tion technologies available. Forcing IoT users to support
these different types of diversity is not feasible as they W
i-F Zig
typically lack the proper resources (e.g., know-how, time, Bluetooth
i be
e
In
and processing resources) to handle the specifics con- ter
op
er
strained devices and networks. To provide support for abi
lit
yI
seamless communication among such heterogeneous ss
ue
devices, interoperable solutions are really required to be
incorporated. Fig. 2 A scenario of interoperability issue in IoT

123
CSIT (March 2017) 5(1):87–95 91

that all industries maintain the same pattern of

organization and makes it easier to manage multiple
clients.
Achieving IoP among all the heterogeneous devices across
various communication technologies is indispensable. It
will defeat the very purpose of having billions of sensors,
actuators and tiny smart devices connected to the Internet if
they are actually unable to communicate with each other in
any way. In fact, for the IoT to flourish, things connecting
to the communication networks, which can be heteroge-
Fig. 3 Different levels of interoperability neous, need to be able to communicate with other things or
applications. Some possible approaches to provide IoP are
discussed below-
2. Syntactic Interoperability It can be found whenever
different people or systems use different structures and 1. Protocol Translation: Gateway acts as proxies to
types to represent information and knowledge [1]. convert proprietary protocols to TCP/IP and vice
Syntactic interoperability is usually associated with versa. It has low complexity, low-cost, and can easily
data structures and formats (e.g., JSON or XML). be adopted. But due to lack of unified standards,
Certainly, the messages passed over by communication protocol translation causes isolation among IoT
protocols need to have an explicit encoding and syntax, applications.
even if it is only in the form of bit-tables. [38] 2. IPv6 over WSN: IPv6 can be used in WSN by
3. Semantic Interoperability Semantic interoperability compressing the header and using the stateless auto-
issue refers to having different meaning of the same configuration of IPv6. But, as IoT things are of
content. Semantic Interoperability is mainly concerned different sorts and sizes, further efforts are still
with the human rather than machine-to-machine inter- required to ensure that the proposed stack is adapt-
pretation. It gives the meaning of the information able to devices with different and limited capabilities.
which is being exchanged between them [9]. It is 3. Using Device Ontology: Using ontology, we can share
considered as an important barrier to providing inter- the common understanding of the structure of infor-
operability as the information and knowledge repre- mation among people or software agents. It gives
sented in most of the software have not distinctly meta-information, knowledge and description about
defined semantics to allow unambiguous understand- devices such as- device name, vendor details, hardware
ing of the meaning of content. description and software description, etc. But, all
4. Pragmatic Interoperability It explains the unexpected- device providers will not adopt the same set of
ness about the effect of the exchanged messages and ontologies because of their different contexts.
context between the sender and receiver. Pragmatic 4. Web of Things (WoT): Here, every node in the
IoP provides the compatibility between the intended network run a web server. The client is able to access
against the actual use of exchanged message within an any devices included in the network. The main
appropriate shared context [26]. The approach of limitation of WoT is that it is based on user-centric
pragmatic interoperability is still largely unsettled, as architecture. Here, actions done by devices are always
illustrated by the various proposed definitions ([5, 6], initiated by a user.
[30]) and the lack of a authorized understanding. 5. Service Oriented Middle-ware (SOM): Data generated
5. Organizational Interoperability Organizational Inter- from the sensor devices are treated as services. The
operability is the inability of organizations to effec- interaction between producer and consumer is done by
tively communicate and transfer meaningful the registry which provides discovery, managements,
information among them when they are using a variety and access to the consumers. But, most existing SOMs
of different information systems over widely different are WSNs-centric and their scale is limited to WSNs,
infrastructures, geographic regions and cultures. It which is typically in the range of thousands, much less
depends on successful technical, syntactical, semantic than the ultra large-scale (billions) of IoT.
and pragmatic interoperability. Organizational Inter- 6. Designing a Generic Protocol Stack: We can design a
operability is the ability of an enterprise, a company or protocol stack by integrating all the other low power
other large organization to practical link activities, technologies which can be able to provide interoper-
such as supply chains, product design, manufacturing, ability at different aspects such as- physical integration
in a competitive and efficient way [7]. This makes sure that focuses on interconnecting different devices,

123
92 CSIT (March 2017) 5(1):87–95

1. To deal with the reliability issues, scheduling mech-

anisms need to be developed for maintaining duplicate
services in case failure [17].
2. The services in a cloud should take the responsibility
to enable automated bootstrapping, registration, mon-
itoring, and upgrade.
3. A data processing pipeline is required to be developed
for collecting, cleaning, enrichment, and transforma-
tion on streaming data [23].
4. Three dimensions- X, Y, and Z axes based scaling
approach is very important to handle massive requests,
actions, and data [4]. The X-axis scaling involves in
dividing the requests among multiple servers. Y-axis
approaches divide the workload based on actions. The
Fig. 4 Different entities participating in IoT incoming request data or the response data are carried
out in Z-axis scaling approaches.
application integration executing different applications
5. In order to scale an IoT application, it needs to break
concurrently, etc. It supports all types of low-power
down into multiple independent functional units, each
radio hardwares (NB, HBC, and UWB).
of which performs one dedicated function [23]. Each
of these functional units should be independently
3.3 Scalability deployable and executable.
6. It needs to adopt multiple data storage technologies.
Scalability is the key to handling the explosive growth in The query and retrieval requirements coupled with the
the IoT. With the increasing number of nodes, all other analytics algorithms that run on the selected data,
components (as shown in Fig. 4) are also increasing pro- should determine the choice of data storage or the
portionally. The IoT applications must have the ability to database technology [29].
support an increasing number of connected devices, users,
application features, and analytics capabilities without any
3.3.2 Network and communication protocol
degradation in the QoS. Monitoring, securing, and
managing an increasing number of devices require a pro-
In the context of IoT communications, there may be
portionate increase in the resources.
thousands of devices which try to transmit simultaneously.
As a massive number of connections are needed to be
For example, hundreds of Philips HUE lamps in a dense
maintained by a cloud server with the deployed devices.
ZigBee network was shown at the 2014 Light and Building
Provisioning scalability in such architecture is a primary
Fair in Frankfurt [8]. A large number of devices need to
concern. Further, networking and communicating among
connect to a single network for a special objective. As
such huge number of devices is another challenge of IoT.
such, networking and communication with limited channel
The scalable approaches to handle these challenges pri-
bands is challenging. LTE-A [13], IEEE 802.11ah [18], and
marily lie in i) Cloud and server platform and ii) Network
6LowPAN [31] are some of the existing scalable solutions
and communication protocol.
for MAC, PHY, and network layers. Different strategies for
providing scalability are:
3.3.1 Cloud and server platform
1. The physical layer techniques have greatly improved
When the number of devices grows, numbers of connec- the link capacity in wireless networks. New modula-
tions and data produced by these devices also grows. As tion and coding schemes (MCSs) are required for
such handling massive connections, data volumes, and improving network capacity.
request/response a big challenge. Some of the cloud solu- 2. With increasing number of active things in IoT, the
tions for IoT are openIoT1 , Compose2, ClouT3, and Kaa4. Medium Access Control (MAC) protocols should be
Different strategies adopted to enhance scalability are: able to control contention and collisions over the shared
wireless medium to deliver stable performances.
1
http://openiot.eu/. 3. To uniquely identify the devices, the addressing
2
http://www.compose-project.eu/. schemes IPv4 and IPv6 are the solutions. The limited
3
http://clout-project.eu/home2/. address space of IPv4 has driven the transition to IPv6.
4
http://www.kaaproject.org/. The IPv6 can give available unique and globally

123
CSIT (March 2017) 5(1):87–95 93

routable addresses. However, higher overhead in use of

IPv6 is a great concern for such resource constraint
networks.
4. Another important aspect of provisioning scalability is
to minimize the protocol overhead as the network size
and the physical layer capacity increases.
5. The devices are constraint in nature and it is very
important that the protocols are optimized to consume
very low power.
6. Data aggregation techniques are used to deal with the
redundant data. It combines all the redundant and
correlated data into valid high-quality information
which is in turn transmitted to the sink through the
intermediate nodes. This process can reduce the Fig. 5 An example of smart health-care system
number of redundant packets transmitted.
3. Identifying individuals in a system (such as a country,
3.4 Security and privacy a network, or an enterprise) and accordingly control
their access to resources within that system by
Several new security and privacy challenges are introduced associating user rights and restrictions are very
in IoT. Due to the resource-constrained nature of IoT important.
devices, it is very difficult to protect the information. On 4. Provisioning data availability by serving and protect-
the other hand, IoT requires global connectivity and ing them in a cloud is one of the major issues of the
accessibility; which means that anyone can access IoT future IoT. Some of the data protection solutions
devices at anytime and anyway. For that, the number of available in the literature are Commvault5, Net-
attacks available to malicious attackers might become Backup6, Rubrik7, and Veeam8.
overwhelming. Again, the communication between con- 5. Privacy preserving services are also necessary in the
strained devices and other entities depends on radio wave, context of IoT.
which is susceptible to many attacks [16]. The indispens-
able complexity of the IoT, where multiple heterogeneous 3.5 Quality of service (QoS)
devices located in different contexts can exchange infor-
mation with each other, further complicates the design and QoS provisioning in IoT applications has become a hot and
deployment of efficient, interoperable, and scalable secu- unexpanded research topic in todays era. There are so
rity mechanisms. Cloud computing and the ubiquity also many QoS challenges in IoT to be addressed like avail-
increases the urgency of the privacy leakage problem. As a ability, reliability, mobility, performance, scalability,
result, there is an increasing demand for the development interoperability, security, management, and trust [19].
of novel security and privacy mechanisms to guarantee the In a smart health-care monitoring system (as shown in
security, privacy, integrity, and availability of resources in Fig. 5), it is necessary to constantly monitor the patients
IoT. physiological parameters. For example, a pregnant woman
Traditional security approaches cannot be directly used parameters such as blood pressure and heart rate of the
in IoT because of the different standards and communica- woman and heart rate and movements of fetal to control
tion stacks involved. Therefore, novel and new security and their health condition. This monitoring systems not only
privacy methods and solutions are needed to deal with monitor vital signs but also detect abnormalities and
security threats in IoT. Some of the future research chal- transmit the data in real-time to health-care providers. In
lenges are listed as follows- such a scenario, reliable delivery of those sensitive data to
1. Designing lightweight security solutions for resource– the service provider without compromising delay and
constraint networks and devices is an important task. accuracy is important. Hence, requirement of QoS for the
2. Universal authentication is a network identity verifi- IoT application is very much necessary to provide the
cation method that allows users to move securely from
one place to another without entering identifying 5
http://www.commvault.com.
information multiple times. The same can be applied 6
http://www.veritas.com/product/backup-and-recovery/netbackup.
7
in IoT. http://www.rubrik.com.
8
http://www.veeam.com.

123
94 CSIT (March 2017) 5(1):87–95

Table 2 QoS requirements of various IoT applications

Application Type of data flow Delay sensitiveness Priority Availability Reliability Scalability

Health-care Continuous High High High High High

Environment monitoring Event Medium Moderate High High High
Smart vehicle Continuous High High High High High
Smart home Event, Query Medium Moderate High High Moderate
Children monitoring Continuous Medium Moderate High High Moderate
Social networking Query Low Low Low Moderate Moderate
Smart traffic Control Continuous High High High High High
Smart business Continuous Medium Moderate High High High

services in some better and smart way. Many protocols are 8. Real time requirements: Multimedia applications like
developed to provide QoS to various applications video chatting, texting, audio chatting shows different
depending upon their types of services or their needs. types delay sensitivities, which is very difficult to
Table 2 discusses several IoT applications and their achieve as a whole.
required QoS parameters for seamless and better services. Some of the possible solutions to provide QoS to several
Due to the architectural differences and a wide range of IoT applications are mentioned below:
devices and networks involved in it, provisioning QoS in
IoT is a very challenging task. The main challenges in 1. Application specific QoS architecture.
provisioning QoS in IoT are as follows - 2. Designing efficient MAC protocol to deal with energy
efficiency, throughput, and delay.
1. Resource–constraint devices: The sensor devices in
IoT are placed in remote areas, so power constraints is
a big issue in this case [25]. Besides this power
constraints there are some other issues like bandwidth, 4 Conclusion
buffer size, memory available, processing capability of
the nodes. Over the last few years, paradigm of Internet has changed
2. Traffic load: The traffic load in IoT is unbalanced as drastically. Due to the increased availability and the
the sink node is getting data from a large amount of advancement of the sensor and cheap hardware technolo-
sensors scattered in the environment. This unbalanced gies, it has been possible to attach sensors to all the
traffic load also effects in the overall QoS of a physical devices around us enabling them to communicate
network. with each other without human intervention. This article
3. Data redundancy: As already mentioned, the sink node provided a brief understanding to the main concepts, pro-
receives data from a large number of sensor nodes. tocol stacks and features of IoT. It analyzed the objectives
Among all these data, there will be some redundant and challenges of IoT technology by indicating some of the
data [14]. Sending this redundant data will cause probable solutions. This work will help in providing the
wastage some extra amount of energy which will effect theoretical foundation for developing large-scale practical
in QoS. Using some data fusion or data aggregation IoT systems.
techniques the redundant data can be eliminated.
4. Scalability: Scalability is also an open issue for IoT
applications. An application should work consistently References
even when some more users are added to the network.
5. Fault tolerance: Node failure or link failure is a big 1. Ahsan M, Talib MR, Sarwar MU, Khan MI, Sarwar MB (2016)
Ensuring interoperability among heterogeneous devices through
issue towards provisioning QoS in wireless networks. IoT middleware. Int J Comput Sci Inf Secur 14(4):251–255
6. Heterogeneity: Heterogeneity in IoT effects in QoS 2. Aijaz A, Su H, Aghvami AH (2015) CORPL: A routing protocol
provisioning. for cognitive radio enabled AMI networks. IEEE Trans Smart
7. Multiple receiver and traffic types: Application wise Grid 6(1):477–485
3. Al-Fuqaha A, Guizani M, Mohammadi M, Aledhari M, Ayyash
there will be multiple receiver which needs different M (2015) Internet of things: a survey on enabling technologies,
types of traffic model. This is also affect in the QoS in protocols, and applications. IEEE Commun Surveys Tutor
IoT. 17(4):2347–2376

123
CSIT (March 2017) 5(1):87–95 95

4. Aljawarneh S (2012) Cloud computing advancements in design, the 5th international conference on communication, theory, reli-
implementation, and technologies. IGI Global, Hershey ability, and quality of service (CTRQ 2012), pp 33–38
5. Asuncion CH, Van Sinderen M (2011) Towards pragmatic 26. Neiva FW, David JMN, Braga R, Campos F (2016) Towards
interoperability in the new enterprise—a survey of approaches. pragmatic interoperability to support collaboration: a systematic
Springer, Berlin review and mapping of the literature. Inf Softw Technol
6. Bravo M, Velazquez J (2008) Discovering Pragmatic Similarity 72:137–150
Relations between Agent Interaction Protocols. Springer, Berlin 27. Nieminen J, Savolainen T, Isomaki M, Patil B (2015) IPv6 over
7. Chen D, Daclin N et al (2006) Framework for enterprise inter- bluetooth(r) low energy. RFC 7668, RFC Editor. https://tools.ietf.
operability. In: Proceedings of IFAC workshop enterprise inte- org/html/rfc7668.txt
gration, interoperability and networking (EI2N). Norway, 28. Saint-Andre P (2004) Transmission of IPv6 Packets over IEEE
pp 77–88 802.15.4 Networks. RFC 3920, RFC Editor. https://www.ietf.org/
8. Dandelski C, Wenning BL, Perez DV, Pesch D et al (2015) rfc/rfc3920.txt
Scalability of dense wireless lighting control networks. IEEE 29. Sarkar S, Kundu A (2016) An indexed approach for multiple data
Commun Mag 53(1):157–165 storage in cloud. In: Information systems design and intelligent
9. Desai P, Sheth A, Anantharam P (2015) Semantic gateway as a applications, pp 639–646. Springer
service architecture for IoT interoperability. In: IEEE interna- 30. Schoop M, Moor Ad, Dietz JL (2006) The pragmatic web: a
tional conference on mobile services, 2015, pp 313–319 manifesto. Commun ACM 49(5):75–76
10. Dujovne D, Watteyne T, Vilajosana X, Thubert P (2014) 6tisch: 31. Shelby Z, Bormann C (2011) 6LoWPAN: the wireless embedded
deterministic ip-enabled industrial internet (of things). IEEE internet, vol 43. Wiley, Hoboken
Commun Mag 52(12):36–41 32. Shelby Z, Hartke K, Bormann C (2014) The constrained appli-
11. Elkhodr M, Shahrestani S, Cheung H (2016) The internet of cation protocol (CoAP). RFC 7252, RFC Editor. https://tools.ietf.
things: New interoperability, management and security chal- org/html/rfc7252.txt
lenges. arXiv preprint arXiv:1604.04824 33. Singh M, Rajan MA, Shivraj VL, Balamuralidhar P (2015)
12. Evans D (2011) The internet of things. how the next evolution of Secure MQTT for internet of things (IoT). In: 5th International
the internet is changing everything, whitepaper. Cisco Internet conference on communication systems and network technologies
Business Solutions Group (IBSG) (CSNT), pp 746–751
13. Ghosh A, Ratasuk R, Mondal B, Mangalvedhe N, Thomas T 34. Sun W, Choi M, Choi S (2013) IEEE 802.11 ah: a long range
(2010) LTE-advanced: next-generation wireless broadband 802.11 WLAN at sub 1 GHz. J ICT Stand 1(1):83–108
technology [Invited Paper]. IEEE Wirel Commun 17(3):10–22 35. The internet of things—concept and problem statement(draft).
14. Gubbi J, Buyya R, Marusic S, Palaniswami M (2013) Internet of https://tools.ietf.org/html/draft-lee-iot-problem-statement-00
things (IoT): a vision, architectural elements, and future direc- 36. Ungurean I, Gaitan NC (2015) Data distribution service for real-
tions. Future Gener Comput Syst 29(7):1645–1660 time systems-a solution for the internet of things environments.
15. Gunner B (2013) MQTT will enable the internet of things. https:// Ann Univ Dunarea de Jos of Galati: Fascicle II Math Phys Theor
www.ibm.com/developerworks/community/blogs/c565c720- Mech 38(1):72–76
fe84-4f63-873f-607d87787327/entry/tc_overview?lan=en 37. Vasco Lopes N, Pinto F, Furtado P, Silva J (2014) IoT archi-
16. Han G, Shu L, Chan S, Hu J (2016) Security and privacy in tecture proposal for disabled people. In: IEEE 10th international
internet of things: methods, architectures, and solutions. Secur conference on wireless and mobile computing, networking and
Commun Netw 9(15):2641–2642 communications (WiMob), pp. 152–158. doi:10.1109/WiMOB.
17. Hao YFSJJ (2015) A scalable cloud for internet of things in smart 2014.6962164
cities. J Comput 26(3):1–13 38. Van der Veer H, Wiles A (2008) Achieving technical interoper-
18. IEEE draft standard for information technology-telecommunica- ability. Eur Telecommun Stand Inst 3:1–30
tions and information exchange between systems-local and 39. Vinoski S (2006) Advanced message queuing protocol. IEEE
metropolitan area networks-specific requirements-part 11. IEEE Internet Comput 10(6):87–89
P802.11ah/D6.0, (Amendment to IEEE Std 802.11REVmc/D5.0) 40. Wang C, Daneshmand M, Dohler M, Mao X, Hu R, Wang H
pp 1–645 (2016) (2013) Guest editorial—special issue on internet of things (IoT):
19. Kaur S, Mir RN (2015) Quality of service in WSN-a review. Int J architecture, protocols and services. Sensors J IEEE
Comput Appl 113(18) 13(10):3505–3510
20. Lin MS, Leu JS, Li KH, Wu JLC (2013) Zigbee-based internet of 41. Whitmore A, Agarwal A, Da Xu, Xu L (2015) The internet of
things in 3D terrains. Comput Electr Eng 39(6):1667–1683 things—a survey of topics and trends. Inf Syst Front
21. Lorawan specification. https://www.lora-alliance.org/portals/0/ 17(2):261–274
specs/LoRaWAN%20Specification%201R0.pdf 42. Winter T (2012) Rpl: IPv6 routing protocol for low-power and
22. Mikhaylov K, Petäjäjärvi J, Haenninen T (2016) Analysis of lossy networks. RFC 7252
capacity and scalability of the lora low power wide area network 43. Xu K, Qu Y, Yang K (2016) A tutorial on the internet of things:
technology. In: The proceedings of 22nd European wireless from a heterogeneous network integration perspective. IEEE
conference. VDE VERLAG GmbH, pp 1–6 Network 30(2):102–108
23. Misra P (2016) Build a Scalable Platform for High-Performance 44. Yang Z, Ping S, Sun H, Aghvami H (2016) CRB-RPL: a receiver-
IoT Applications. Technical report, TCS Experience certainty based routing protocol for communications in cognitive radio
24. Montenegro G, K N, Hui J (2007) Transmission of IPv6 Packets enabled smart grid. IEEE Trans Veh Technol 65:1–10
over IEEE 802.15.4 Networks. RFC 4944, RFC Editor. https:// 45. Zaloker Joseph (2014) ANT/ANT?. Technical report, Arrow
tools.ietf.org/rfc/rfc4944.txt M2M representative
25. Nef MA, Perlepes L, Karagiorgou S, Stamoulis GI, Kikiras PK
(2012) Enabling QoS in the internet of things. In: Proceedings of

123