IoT as a Service

Full text

IoT as a Service
T.Reuban Gnana Asir
Principal Product Manager, Video Business Unit,
Nokia, Chennai, India
Dr. Hansa Lysander Manohar
Associate Professor, College of Engineering Guindy,
Anna University, Chennai, India
Wilson Anandaraj
Director, IMS Business Unit,
Nokia, Chennai, India
K.Naga Sivaranjani
Senior Engineer, Video Business Unit
Nokia, Chennai, India
Abstract— The Internet of Things (IoT) is transforming the
everyday physical objects that surround us into an ecosystem
of information that will enrich our lives. From refrigerators to
parking spaces to houses, the IoT is bringing more and more
things into the digital fold every day, which will likely make
the IoT a multi-trillion dollar industry in the near future.
While the IoT represents the convergence of advances in
miniaturization, wireless connectivity, increased data storage
capacity and batteries, the IoT wouldn’t be possible without
sensors. Sensors detect and measure changes in position,
temperature, light, etc. and they are necessary to turn billions
of objects into data-generating “things” that can report on
their status, and in some cases, interact with their
environment. Because sensor endpoints fundamentally enable
the IoT, sensor investments are an early indicator of the IoT’s
progress. One day we will see “IoT as a Service” technology
offered and used the same way we use other “as a service”
technologies.
Keywords— IoT, Cloud Systems, IaaS
I. INTRODUCTION
The Internet of Things is emerging as the third wave in
the development of internet. In 1990s, fixed internet wave
connected 1 billion computers while 2000s mobile wave
connected another 2 billion users, Now the IoT has the
potential to connect ten times, as many as 28 billion ‘things’
to the internet by year 2020 [1], ranging from wearables,
cars, home equipments, Industrial equipments, etc.
According to industry analyst firm IDC, the installed
base for the Internet of Things will grow to approximately
212 billion devices by 2020, a number that includes 30
billion connected devices. IDC sees the growth driven
largely by intelligent systems that will be installed and
collecting data across both consumer and enterprise
applications [2]
These type of applications can involve the electric
vehicle and the smart house, in which appliances and
services that provide notifications, security, energy saving,
automation, telecommunications, computers and
entertainment will be integrated into a single eco system
with a shared user interface. IoT is providing access to
information, media and services, through wired and wireless
broadband connections. The Internet of Things makes use of
synergies that are generated by the convergence of
Customer, Business and Industrial Internet Consumer,
Business and Industrial Internet. The convergence creates
the open, global network connecting people, data and things.
This convergence leverages the cloud to connect the
intelligent things that sense and transmit a broad array of
data, helping creating services that would not be obvious
without this level of connectivity and artificial intelligence.
Including India, huge investments are ongoing across the
globe into IoT research, implementation in various forms as
smart projects [3]
The use of platforms is being driven by transformative
technologies such as cloud, things and mobile. The Internet
of Things and Services makes it possible to create networks
incorporating the entire manufacturing process that converts
factories into a smart environment. The cloud enables a
global infrastructure to create new services, allowing
anyone to create content and applications for global users.
Network of things connect things globally and maintain
their identity online. Mobile allows the connection to this
global infrastructure anytime, anywhere. The result is a
globally accessible network of things, users, and consumers
who are available to create businesses, contribute content,
generate and purchase new services [4]
II. IOT ON THE CLOUD
Cloud computing technologies have been intensively
exploited in development and management of the large-
scale IoT systems, because theoretically, cloud offers
unlimited storage, compute and network capabilities to
integrate diverse types of IoT devices and provide an elastic
runtime infrastructure for IoT systems. Self-service, utility
oriented model of cloud computing can potentially offer fine
grained IoT resources in a pay-as-you-go manner, reducing
upfront costs and possibly creating cross-domain application
opportunities and enabling new business and usage models
of the IoT cloud systems. However, most of the
contemporary approaches dealing with IoT cloud systems
largely focus on data and device integration by utilizing
cloud computing techniques to virtualize physical sensors
,QWHUQDWLRQDO&RQIHUHQFHRQ,QQRYDWLRQVLQLQIRUPDWLRQ(PEHGGHGDQG&RPPXQLFDWLRQ6\VWHPV,&,,(&6
978-1-4673-8207-6
1093

and actuators. Although, there are approaches providing
support for provisioning and management of the virtual IoT
infrastructure the convergence of IoT and cloud computing
is still at an early stage.
System designers and operations managers face
numerous challenges to realize large-scale IoT cloud
systems in practice, mainly because these systems impose
diverse requirements in terms of granularity and flexibility
of IoT resources consumption, custom provisioning of IoT
capabilities such as communication protocols, elasticity
concerns, and runtime governance. For example, modern
large-scale IoT cloud systems heavily rely on the cloud and
virtualized IoT resources and capabilities (e.g., to support
complex, computationally expensive analytics), thus these
resources need to be accessed, configured and operated in a
unified manner, with a central point of management.
Further, the IoT systems are envisioned to run
continuously, but they can be elastically scaled in/down in
off-peek times, e.g., when a demand for certain data sources
reduces. Due to the multiplicity of the involved stakeholders
with diverse requirements and business models, the modern
IoT cloud systems increasingly need to support different and
customizable usage experiences. Therefore, to utilize the
benefits of cloud computing, IoT cloud systems need to
support virtualization of IoT resources and IoT capabilities
(e.g., gateways, sensors, data streams and communication
protocols) [5].
III. IOT CLOUD PLATFORM
In addition to the DataCenters, the IoT platform
comprises various other IoT Infrastructure, that includes
Sensors, Actuators, Communication Brokers, Gateways.
Most of these IoT Infrastructure are getting more software
defined and software controlled ones, which fuels the IAAS
Services.
Figure.1. Software Defined IoT Cloud Systems
Figure 1 gives high-level graphical overview of the main
building blocks and enabling techniques, needed to support
the main principles of Software Defined IoT Systems.
Software-defined IoT systems comprise a set of
resource components, hosted in the cloud, which can be
provisioned and controlled at runtime. The IoT resources
(e.g., sensory data streams), their runtime environments
(e.g., gateways) and capabilities (e.g., communication
protocols, analytics and data point controllers) are described
as software-defined IoT units. Software-defined IoT units
are software-defined entities that are hosted in an IoT cloud
platform and abstract accessing and operating underlying
IoT resources and lower level functionality.
Generally, software-defined IoT units are used to
encapsulate the IoT resources and lower level functionality
in the IoT cloud and abstract their provisioning and
governance, at runtime. To this end, the software-defined
IoT units expose well-defined API and they can be
composed at different levels, creating virtual runtime
topologies on which we can deploy and execute IoT cloud
systems.
Therefore, main principles of software-defined IoT
systems include:
• API Encapsulation – IoT resources and IoT
capabilities are encapsulated in well-defined APIs, to
provide a unified view on accessing functionality and
configurations of IoT cloud systems.
• Fine-grained consumption – The IoT resources
and capabilities need to be accessible at different granularity
levels to support agile utilization and self-service
consumption.
• Policy-based specification and configuration –
The units are specified declaratively and their functionality
is defined programmatically in software, using the
welldefined API and available, familiar software libraries.
• Automated provisioning – Main provisioning
processes need to be automated in order to enable dynamic,
ondemand configuring and operating software-defined IoT
systems, on a large-scale (e.g, hundreds gateways).
• Cost awareness – We need to be able to assign
and control costs of delivered IoT resources and capabilities
in order to enable their utility-oriented consumption.
• Elasticity support – They should support
elasticity governance, by exposing runtime control of elastic
capabilities through well-defined API..
• Multi-layer management support – The multi
layer management layers include: the application layer, the
control layer and the resource layer [6]. The functions
include support fault handling, configuration, accounting,
performance and security management as described in ITU-
T M.3400
These principles are enabled by the software
defined IoT units and support for centrally managed
configuration.
978-1-4673-8207-6
1094

IV. IAAS PROTOTYPE
Various aspects needs to be considered for IaaS
deployment, which includes
- Number of input devices (sensors and gateways)
- Number of Output devices
- Amount of data consumption
- Frequency of data consumption
- Processing speed of middleware, backend servers
In this paper, we will see the prototype of the IaaS, with
an example, leaving the calculations for deployment in later
stage, as future work. Hence the prototype will deal with 4
main components namely, sensors, output device,
middleware and back end servers, as shown in Figure 2.
From customer premises perspective, we see Sensors &
Gateways as input components and say, end user smart
phones as output devices.
Figure.2 IaaS Prototype
- Sensors and Gateways:
Wireless Sensor Networks (WSN) can realize the
short distance communication among the objects by
constructing wireless networks in ad-hoc manners.
However, it’s difficult to connect the WSN and
mobile communication networks or the Internet
with each other because it lacks of uniform
standardization in communication protocols and
sensing technologies and the data from WSN cannot
be transmitted in long distance with the limitation of
WSN’s transmission protocols. Therefore, with the
development of the Internet of Things, a new type
of network equipment called the Internet of Things
Gateway is invented, whose goal is to settle with the
heterogeneity between various sensor networks and
mobile communication networks or Internet,
strengthen the management of the WSN and
terminal nodes, and bridge traditional
communication networks with sensor networks to
make network communication easier and manage
the devices of sensor networks.[7]
- Middleware:
Global Sensor Network (GSN) is a middleware that
enables users to more easily integrate with various
different types of sensors. Primarily, GSN handles
the thread management and data storage aspects of
dealing with sensors. The users must create an XML
file and a wrapper for each different sensor that they
want to connect to the network [8]. The XML file
tells GSN about the basics of the sensor: what kind
of data (numerics, spatial) it will be giving to the
system, any parameters (such as how often to ping
the sensor to get data), and which wrapper or virtual
sensor to use with the physical sensor. The wrapper
tells GSN how to connect (e.g. which network
protocol to use) to the sensor when it is first
initialized, what to do in order to get data from the
sensor, and what to do with the data when it is
received from the sensor. These two things must be
created for every new sensor that one wishes to
connect to GSN. Currently, there is no easy way for
users to create new XML files or wrappers, besides
taking the existing ones and modifying them or
writing them from scratch. Furthermore, while GSN
stores the data from the sensors in an SQL database,
it doesn’t do anything to the data other than display
it on a local web application, and the only data
displayed is the raw sensor data - with no
interpretation
- Backend Servers :
The backend servers are a combination of storage
server, data interpreters, UI interfaces and few 3rd
party applications enhancing the communication
flow.
Firebase is a cloud storage service that can be
leveraged in order to enable the middleware to be
decoupled and interface with 3rd party applications
[9]. Since Firebase stores data in a JSON format,
many applications can already interpret that format,
but Firebase also allows developers to create
listeners for specific sections of the JSON document
that will fire when data is changed, added, removed
or moved. This means that creating a mobile or web
application that will interface is incredibly simple.
Furthermore, since Firebase enables users to add
authentication to their own personal Firebase, users
can rest assured that their data is protected from
malicious attackers.
978-1-4673-8207-6
1095

The Data Interpreter pulls all of the sensor data
down from Firebase, and then using the devices that
were listed as being monitored by each sensor in the
XML file, determines which sensors need to
conclude that each device is on or off before the
interpreter can actually determine that the device is
on or off. It creates Firebase entries for each one of
the devices that are listed as being monitored and
uploads their initial state. Following that, no
changes will be made unless the state of the device
changes - whether the change is the device going
from on to off, off to on, or simply changing
location.
Figure.3 User-friendly output using web interface
- Output interface :
This includes the web interfaces that display the
data provided about the devices in a format that is
easy for users to interpret in a short period of time.
Furthermore, 3rd party applications like Jess
Applications posts data to social forums like twitter
will feed the output to end user. Hence the backend
server upon coupling with such Apps, will act as
output interface.
V. FUTUREWORK
With additional efforts, the prototype needs to get
converted into product, get demonstrated to meet the
expectations of IoT. Each of these building blocks, also
have quite some challenges and limitations, by its own
unique strengths and restrictions, which needs to get tackled
appropriately, based on testing and optimization. IaaS, upon
leveraging the NFV and SDN technologies opens mammoth
opportunities towards performance improvements.
VI. CONCLUSION
The building blocks of IaaS are stretchable, expandable
and can be located anywhere based on the scope, volume
and geographical situations. Thus we see, IoT can be served
as a Service and can be operated as OPEX, rather than a
capital investment, Based on the expansion requirements,
the building blocks can aswell get expanded proportionally,
leading to successful deployment of IaaS. It is also a good
model, for service providers to enter into the IoT
commercial market.
Abbreviations and Acronyms
GSN – Global Sensor Network
IaaS – IoT as a Service
IoT – Internet of Things
JSON – Java Script Object Notation
OPEX - Operational Expenditure
XML – Extensible Markup Language
WSN – Wireless Sensor Network.
ACKNOWLEDGMENT 
We would like to thank Nokia, College of Engineering
Guindy, for giving us such an opportunity to carry out this
research work and also for providing us the requisite
resources and infrastructure for carrying out the research.
REFERENCES
[1] “The Internet of Things: Making sense of the next mega-trend”
Goldman Sachs Global Investment Research, IoT Primer, Sept 2,
2014
[2] IDC, Worldwide Internet of Things (IoT) 2013-2020 Forecast:
Billions of Things, Trillions of Dollars, Doc # 243661, October 2013.
[3] T.Reuban Gnana Asir, Wilson Anandaraj, K.Naga Sivaranjani,
“Internet of things and India’s Readiness”, International Conference
on Computing Paradigms (ICCP2015) 24, 25 July, 2015, 274-279
[4] Ovidiu Vermesan, Peter Friess, “Internet of Things – From Research
and Innovation to Market Deployment.
[5] Stefan Nastic, Sanjin Sehic, D, et al, “Provisioning Software-defined
IoT Cloud Systems”, International Conference on Future Internet of
Things and Cloud, (2014) 288-295.
[6] “Y.3300 : Framework for Software Defined Networking”,
https://www.itu.int/rec/T-REC-Y.3300/en
[7] Qian Zhu, Ruicong Wang, et al, “IOT Gateway: Bridging Wireless
Sensor Networks into Internet of Things”, IEEE/IFIP International
Conference on Embedded and Ubiquitous Computing, (2010) 347-
352
[8] GSN Team, Global Sensors Networks, 2009.
[9] “Firebase documents,” https://www.firebase.com/docs/
978-1-4673-8207-6
1096