ISSN (ONLINE): 2395-695X

ISSN (PRINT): 2395-695X

International Journal of Advanced Research in Basic Engineering Sciences and Technology (IJARBEST)
Volume 2, Special Issue 19, October 2016

FOG COMPUTING AND ITS APPLICATIONS

Mrs. R. Waheetha, M.C.A. M.Phil.,
Associate Professor
Department of Computer Science
Holy Cross Home Science College
Thoothukudi, Tamil Nadu, India.

Mrs.Sowmya Fernandez, M.C.A.

Associate Professor
Department of Computer Science
APC Mahalaxmi College for Women
Thoothukudi, Tamil Nadu, India.

Abstract applications, g) Heterogeneity. In this paper

we argue that the above characteristics make
Cloud computing is used as a delivery the Fog the appropriate platform for a
platform which is a promising way for number of critical Internet of Things (IoT)
storing user data and provides a secure services and applications, namely,
access to personal and business information. Connected Vehicle, Smart Grid , Smart
The users are provided with on-demand Cities, and in general, Wireless Sensors and
services through the Internet. Fog Actuators Networks (WSANs).
Computing extends the Cloud Computing
paradigm to the edge of the network, thus Keywords: Fog Computing, Sensors,
enabling a new breed of applications and Actuators, Internet Of Things, Latency
services. The fog computing provides the
Introduction
services like data, compute, storage and
application to end user. Fog Computing is Fog computing also known as
not a replacement of cloud, it just extends fogging is a distributed computing
the cloud computing by providing security infrastructure in which some application
in the cloud environment. With Fog services services are handled at the network edge in a
we are able to enhance the cloud experience smart device. Fog computing is a paradigm
which monitors the data and helps in
by isolating users’ data that need to live on
detecting an unauthorized access. According
the edge. The main aim of the fog to Cisco, due to its wide geographical
computing is to place the data close to the distribution the Fog computing is well suited
end user. Some characteristics of the Fog for real time analytics and big data. Fog
are: a) Low latency and location awareness; computing involves a dense geographical
b) Wide-spread geographical distribution; distribution of network and provides a
c) Mobility; d) Very large number of nodes, feature of location access. With this any
unauthorized activity in the cloud network
e) Predominant role of wireless access, f)
can be detected. To get the benefit of this
Strong presence of streaming and real time

Special Issue 19 56 © IJARBEST PUBLICATIONS

 technique a user need to get registered with capabilities at the edge and support the local
the fog. Once the user is ready by filling up ingestion of data and the quick turn around
the sign up form he will get the message or of results.
email that he is ready to take the services
from fog computing. Fog networking A study by IDC estimates that by 2020, 10
consists of a control plane and a data plane. percent of the world’s data will be produced
For example, on the data plane, fog by edge devices. This will further drive the
computing enables computing services to need for more efficient fog computing
reside at the edge of the network as opposed solutions that provide low latency and
to servers in a data-center. Fog nodes holistic intelligence simultaneously.
provide localization, therefore enabling low
latency and context awareness, but the Fog computing has its own supporting body,
Cloud provides global centralization. The the Open Fog Consortium, founded in
main task of fog is to deliver data and place November 2015, whose mission is to drive
it closer to the user who is positioned at a industry and academic leadership in fog
location which at the edge of the network. computing architecture. The consortium
Here the term edge refers to different nodes offers reference architectures, guides,
to which the end user is connected and it is samples and SDKs that help developers and
also called edge computing. If we look IT teams understand the true value of fog
according to architecture, fog is situated computing.
below the cloud at the ground level. Fog
computing is a new technology in which Analytics software companies are also
mobile devices interact with one another and scaling products and developing new tools
support the data communication within the for edge computing. Apache Spark is an
Internet of Things. These characteristics example of a data processing framework
make the Fog the appropriate platform for a based on the Hadoop ecosystem that is
number of critical Internet of Things (IoT) suitable for real-time processing of edge-
services and applications, namely, generated data.
Connected Vehicle, Smart Grid , Smart
Cities, and, in general, Wireless Sensors and Insights obtained by the cloud can help
Actuators Networks (WSANs). update and tweak policies and functionality
at the fog layer.
The fog placed at the perfect
position Characteristics of Fog Computing
IoT nodes are closer to the action, but for the Fog Computing is a highly virtualized
moment, they do not have the computing platform that provides compute, storage, and
and storage resources to perform analytics networking services between end devices
and machine learning tasks. Cloud servers, and traditional Cloud Computing Data
on the other hand, have the horsepower, but
Centers located at the edge of network.
are too far away to process data and respond
in time.
Edge location, location awareness, and
low latency: Fog computing supports
The fog layer is the perfect junction where
there are enough compute, storage and endpoints with finest services at the edge of
networking resources to mimic cloud the network.

Special Issue 19 57 © IJARBEST PUBLICATIONS

 Geographical distribution: In sharp different providers. Hence, Fog components
contrast to the more centralized Cloud, the must be able to interoperate, and services
services and applications targeted by the Fog must be federated across domains.
demand widely distributed deployments.
The Fog, for instance, will play an active Cisco's IOx Fog Computing Platform and
role in delivering high quality streaming to Edge Computing Competition
moving vehicles, through proxies and access
The Fog computing vision has taken form in
points positioned along highways and tracks.
Cisco's IOx platform, which brings
Large-scale sensor: Networks to monitor distributed computing capabilities to the
edge of the network by running applications
the environment and the Smart Grid are
directly on Cisco network devices, such as
other examples of inherently distributed ruggedized switches, routers and IP video
systems, requiring distributed computing cameras.
and storage resources.
While Cisco may have coined the term fog
Very large number of nodes: As a computing, it has competition in the nascent
consequence of the wide geo-distribution, as edge computing market from companies like
evidenced in sensor networks in general and EMC, VMware, Intel and IBM, all of which
the Smart Grid in particular. are rolling out — or are in the process of
rolling out — products that deliver edge
Support for mobility: It is essential for computing capabilities.
many Fog applications to communicate
directly with mobile devices, and therefore Why Fog and Why Now?
support mobility techniques, such as the
There are four main reasons:
LISP protocol, that decouple host identity
from location identity, and require a 1. Time: Real time processing and cyber-
distributed directory system. physical system control. Edge data analytics
as well as the actions it enables through
Real-time interactions: Important Fog
control loops, often have stringent time
applications involve real-time interactions
requirement and can only be carried out on
rather than batch processing.
the edge, “here and now.” This is
Heterogeneity: Fog nodes come in different particularly essential for internet: the vision
form factors, and will be deployed in a wide of millisecond reaction time on networks
variety of environments that enable virtual-reality-type interfaces
between humans and devices.
Real time interactions: Fog computing
requires real time interactions for speedy 2. Cognition: Awareness of Client-centric
service. objectives. Following the end-to-end
principle, some of the applications can be
Interoperability and federation: Seamless best enabled by knowing the requirements of
support of certain services (streaming is a the clients. This is especially true when
good example) requires the cooperation of privacy and reliability cannot be trusted in

Special Issue 19 58 © IJARBEST PUBLICATIONS

 the Cloud, or when security is enhanced by In the bottommost layer lie the end devices
shortening the extent over which (sensors), as well as edge devices and
communication is carried out. gateways. This layer also includes apps that
can be installed in the end devices to
3. Efficiency: Pooling of local resources. enhance their functionality. Elements from
There are typically hundreds of gigabytes this layer use the next layer, the network, for
sitting idle on tablets, laptops and set-top communicating between themselves, and
boxes in a household every evening, or between them and the cloud. The next layer
across a table in a conference room, or contains the cloud services and resources
among the passengers of a public transit that support resource management and
system. Similarly, idle processing power, processing of IoT tasks that reach the cloud.
sensing ability and wireless connectivity On top of the cloud layer lays the resource
within the edge may be pooled within a Fog management software that manage the
network. whole infrastructure and enable quality of
4. Agility: Rapid innovation and affordable Service to Fog Computing applications.
scaling. It is usually much faster and cheaper Finally, the topmost layer contains the
to experiment with client and edge devices. applications that leverage fog computing to
Rather than waiting for vendors of large deliver innovative and intelligent
boxes inside the network to adopt an applications to end users.
innovation, in the Fog world a small team Looking inside the Software-Defined
may take advantages of smart phone API Resource Management layer, it implements
and SDK, proliferation of mobile apps, and many middleware-like services to optimize
offer a networking service through its own the use of the cloud and Fog resources on
API. behalf of the applications. The goal of these
services is to reduce the cost of using the
Architecture of Fog Computing
cloud at the same time that performance of
applications reach acceptable levels of
latency by pushing task execution to Fog
nodes.

Application of Fog Computing

Smart Traffic Lights: Video camera that
senses an ambulance flashing lights can
automatically change street lights to open
lanes for the vehicle to pass through traffic.
Smart street lights interact locally with
sensors and detect presence of pedestrian
and bikers, and measure the distance and
speed of approaching vehicles. Intelligent

Special Issue 19 59 © IJARBEST PUBLICATIONS

 lighting turns on once a sensor identifies highest tier and momentary storage at the
movement and switches off as traffic passes. lowest tier.
Neighboring smart lights serving as Fog Self Maintaining Train: Another
devices coordinate to create green traffic application of fog computing is self
wave and send warning signals to maintaining trains. A train ball-bearing
approaching vehicles. Wireless access points monitoring sensor will sense the changes in
like Wi-Fi, 3G, road-side units and smart the temperature level and any disorder will
traffic lights are deployed along the roads. automatically alert the train operator and
Vehicle-to-Vehicle, vehicle to access points, make maintenance according to. Thus we
and access points to access points can avoid major disasters.
interactions enrich the application of Fog
computing.
Wireless Sensor and Actuator Networks
Connected car: Autonomous vehicle is the (WSAN): The real Wireless Sensor Nodes
new trend taking place on the road. A (WSNs), were designed to extend battery
software is used to add automatic steering, life by operating at predominantly low
enabling literal "hands free" operations of power .Actuators serves as Fog devices
the vehicle. Starting out with testing and which control the measurement process
releasing self-parking features that don't itself, the consistency and the oscillatory
require a person behind the wheel. Fog behaviors by creating a closed-loop system.
computing will be the best option for all For example, in the lifesaving air vents
internet connected vehicles why because fog sensors on vents monitor air conditions
computing gives real time interaction. Cars, flowing in and out of mines and
access point and traffic lights will be able to automatically change air-flow if conditions
interact with each other and so it makes safe become dangerous to miners. Most of these
for all. At some point in time, the connected WSNs entail less bandwidth, less energy,
car will start saving lives by reducing very low processing power, operating as a
automobile accidents. sink in a unidirectional fashion.

Smart Grids: Smart grid is another Smart Building Control: In decentralized

application where fog computing is been smart building control wireless sensors are
used. Based on demand for energy, its installed to measure temperature, humidity,
obtained ability and low cost, these smart or levels of various gaseous components in
devices can switch to other energies like the building atmosphere. Thus information
solar and winds. The edge processes the data can be exchanged among all sensors in the
collected by fog collectors and generate floor and the reading can be combined to
control command to the actuators. The form reliable measurements. Using
filtered data are consumed locally and the distributed decision making the fog devices
balance to the higher tiers for visualization, react to data. The system gears up to work
real-time reports and transactional analytics. together to lower the temperature, input
Fog supports semi-permanent storage at the fresh air and output moisture from the air or

Special Issue 19 60 © IJARBEST PUBLICATIONS

 increase humidity. Sensors respond to the Benefits of Fog Computing
movements by switching on or off the lights.
Observance of the outlook the fog Extending the cloud closer to the things that
computing are applied for smart buildings generate and act on data benefits the
which can maintain basic needs of business in the following ways:
conserving external and internal energy.
● Greater business agility: With the right
tools, developers can quickly develop fog
IoT and Cyber-Physical Systems (CPSs): applications and deploy them where needed.
Fog computing has a major role in IoT and Machine manufacturers can offer MaaS to
CPSs. IoT is a network that can interconnect their customers. Fog applications program
ordinary physical objects with identified the machine to operate in the way each
address using internet and customer needs.
telecommunication. The characteristic of
● Better security: Protect your fog nodes
CPSs is the combination of system’s
using the same policy, controls, and
computational and physical elements. The
procedures you use in other parts of your IT
association of CPSs and IoT will transform
environment. Use the same physical security
the world with computer based control and
and cyber security solutions.
communication systems, engineered systems
and physical reality. The object is to ● Deeper insights, with privacy control:
integrate the concept and precision of Analyze sensitive data locally instead of
software and networking with the vibrant sending it to the cloud for analysis. Your IT
and uncertain environment. With the team can monitor and control the devices
growing cyber physical systems we will be that collect, analyze, and store data.
able to develop intelligent medical devices,
smart buildings, agricultural and robotic .● Lower operating expense: Conserve
systems. network bandwidth by processing selected
data locally instead of sending it to the cloud
Software Defined Networks (SDN): SDN for analysis
is a growing computing and networking
concept. SDN concept together with fog Conclusion
computing will resolve the main issues in
Fog computing, the new concept of the
vehicular networks irregular connectivity,
cloud at the edge of the network, is
collisions and high packet loss rate.SDN
considered the appropriate platform for
supports vehicle to-vehicle with vehicle-to-
many Internet of Things services and
infrastructure communications and main
applications along with virtualization of
control. It splits control and communication
WSN/WSAN. Fog computing gives the
layer, control is done by central server and
cloud a companion to handle the extra bytes
server decides the communication path for
of data generated daily from the Internet of
nodes.
Things. Processing data closer to where it is
produced and needed to solve the challenges

Special Issue 19 61 © IJARBEST PUBLICATIONS

 of exploding data volume, variety, and
velocity. It avoids the need for costly
bandwidth additions by offloading gigabytes
of network traffic from the core network. It
also protects sensitive IoT data by analyzing
it inside company walls. Ultimately,
organizations that adopt fog computing gain
deeper and faster insights, leading to
increased business agility, higher service
levels, and improved safety. Hence, we can
come to the conclusion that fog computing
and cloud computing will complement each
other while having their own advantages and
disadvantages. Fog computing will grow in
helping the emerging network paradigms
that require faster processing with less delay
and delay jitter, cloud computing would
serve the business community meeting their
high end computing demands lowering the
cost based on a utility pricing mode.

REFERENCES
[1] Ivan StojmenovicSIT, Deakin University,
Burwood, Australia and SEECS, University of
Ottawa, Canada,”The Fog Computing Paradigm:
Scenarios and Security Issues”
[2] F. Bonomi. Connected vehicles, the internet of
things, and fog computing. VANET 2011, 2011
[3] www.google.com/fog computing-
ThoughtsOnCloud
[4] Cisco, “Cisco delivers vision of fog computing to
accelerate valuefrom billions of connected devices,”
Cisco, Tech. Rep., Jan. 2014.
[5] I. Stojmenovic and S. Wen, “The Fog Computing
Paradigm:Scenarios and Security Issues,” in
Proceedings of FedCSIS, 2014.

Special Issue 19 62 © IJARBEST PUBLICATIONS