Volume 6, Issue 3, March – 2021 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

Fog Computing: Helping Realize the Power of the

Internet of Things
Abhishek Sonar Prajwal Dubey Siddharth Shukla
Department of Computer Department of IT Department of IT
Vidyalankar Institute of Technology, Vidyalankar Institute of Technology, Vidyalankar Institute of Technology,
Mumbai, India Mumbai, India Mumbai, India

Abstract:- Cloud is very important because IoT devices, example, if it is a surveillance application and because of
sensors, and different types of devices process a latency that is involved maybe the intruder might have
tremendous amount of information and finally that already intruded into the territory. If it is a medical
information has been processed and that is the reason emergency scenario for the time it takes to send it to the
that cloud came into existence. All this information will cloud, cloud processing, getting a response back, etc. The
be sent to the cloud for further refinement and so on. patient might even die. This situation has arisen because of
The main issue with the cloud in the IoT environment is the high latency. Thus, there is a need to reduce latency for
the information transmission speed. Let us say there is effectiveness which can be done through fog computing. The
a cloud and there are different IoT devices deployed. whole idea is to bring the cloud facilities, the attractiveness
Each of these devices sends all these data to the virtual of the cloud closer to the IoT device layer, to solve the
cloud for further refining and storage. Now, that problems that are faced by cloud computing, to reduce the
becomes a problem because these IoT environments have latency. This entire time can be reduced by the use of fog
constraints concerning bandwidth, degree of processing, computing. In IoT, there is a device layer where all these IoT
memory utilization, energy consumption, etc. This devices, the physical devices operate, there is a cloud where
processing can be handled with the help of the cloud but all the data is sent for processing and storage. Fog acts like
what about the bandwidth what about the energy middleware where some of the mathematical computation,
consumption because what is going to happen in this sort data processing, information storage is going to take place
of scenario of the use of cloud in the IoT context is a lot before the data is collected by these devices and sent to the
of data is going to float all around. Over the network, a cloud. Before it is dispatched to the cloud for some
lot of information is going to be dispatched to the cloud intermediate processing, the repository for making quicker
and that will unnecessarily consume the bandwidth and decisions is the primary aim of fog computing. Fog
the limited energy that is resident in all these devices. Computing takes the model of Cloud Computing to another
IoT devices have to sense the physical phenomena level in the field of networking. Various applications and
occurring around them and send the data to the cloud services are also permitted because of fog computing. Some
and get an action or comment back. There is a need for trivial features of the Fog Computing are as follows: a)
fog computing to reduce latency. cloud servers might be Decreased latency and knowledge of that particular place; b)
physically located in different continents, cities. This Extensive Geographical allotment; c) mobility; d) More
physical limitation also introduces a large latency in numbers of clients; Nodes, e) Wireless access central
communication. position, f) Solid Presence of applications for live streaming;
g) Non-uniformity. We emphasize in this journal that the
Keywords:- Cloud Computing, Fog Computing, IoT. above features make fog computing an integral part of
networking.
I. INTRODUCTION
There are a lot of clients that face Web applications
The fog was introduced by cisco and it was like an and batch processing in real-time. For them the Cloud
idea to bring the facilities that are provided by the cloud Computing model is an important substitute for possessing
close to the operating IoT devices because all these data sent and operating private data centers. Several factors bestow to
to the cloud will not only take the limited bandwidth that is the mega DC scale economy: greater predictability of
there in this kind of environment but also that is going to massive aggregation, which makes for greater usage of
take a lot of time. In this particular case, the time that is power, storage, and various components of the network.
required will be the time from when that piece of data is sent
to the cloud. let t1, t2, t3 be the time for processing and Cloud computing makes the business and the clients
finally, that response will be sent back with time t4 for free from the specifications of IoT. It becomes a concern for
maybe for activation or any other reason. The total time applications that are sensitive to latency, nodes in the
finally becomes t1+t2+t3+t4. This is the total time that it vicinity. The evolving wave of deployments of the Internet,
takes until when the receiving device gets a signal about The Internet of Things (IoT) needs improvement for
what to do and by this time in most of the real-life mobility, location sensitivity, low geo-distribution, and
applications of IoT most of the unwanted events can take latency of services.
place as there is a lot of latency involved in this process. For

IJISRT21MAR278 www.ijisrt.com 315

Volume 6, Issue 3, March – 2021 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
There is a need for a new platform to reach These 3.1 Enabling real-time analytics
criteria; a network that is called Fog Computing. Fog In fog environments, to reduce latency and optimize
Computing, has given rise to new applications and services. throughput,resource management services are intended to
The interaction between the fog and the cloud is fruitful, determine dynamically which analytical activities are being
especially when it is related to Data analytics and transferred to which cloud or edge-based asset.. These
processing. systems must also take into account other requirements, such
as the data protection laws of different countries, including,
II. UTILITIES for instance, medical and financial data.

Fog Computing is a highly dematerialized plan of 3.2 Programming models and architecture
action that, usually but not exclusively situated at the For fog and IoT environments, most stream and data
extremity of the net work, provides data computing, processing systems, including Apache Storm and S, do not
processing, storage, and networking assistance between user provide adequate scalability and versatility because their
devices and t raditional cloud computing data centers. design is focused on static configurations. Fog environments
need the potential to increase and decrease assets
2.1 Healthcare and monitoring spontaneously because transforming nodes are usually
In healthcare, where processing and event response is mobile appliances that constantly enter and exit the web.
important in real-time, Fog computing may be useful. One
suggested device uses fog computing to track, predict, and 3.3 Security, Reliability, Fault Tolerance
avert stroke victims from dropping. The learning algorithms It is a key challenge to implement safety in fog
for fall-detection are spontaneously distributed over edge atmosphere because they have many service suppliers and
devices and cloud assets.Observations have shown that this end-user, as well as dispersed resources. It is difficult to
system has a shorter responsiveness and less power than develop and enforce authentication and authorization
cloud-alone outlooks.. A smart-healthcare system based on techniques that can operate with several fog nodes with
proposed fog computing allows low latency, support for discrete computational capabilities. Possible remedies are
mobility, and awareness of geographic location and data public-key infrastructures and trusted implementation
privacy. environments. Fog deployment clients often have to prepare
for individual sensors, networks, service platforms, and
2.2 Intelligent Services applications to fail. To assist with this, protocols, such as the
Fog computing can be preowned for smart utility SCTP i.e., Stream Control Transmission Protocol, could be
systems, whi ch concentrate on raising energy production, introduced that trade-in with packet and event authenticity in
distribution, and billing. In such settings, edge systems can wireless sensor networks.
report more fine-looking en ergy usage information e.g.,
hourly and regular readings instea d of monthly readings) to 3.4 Power Consumption
users' mobile tools than conventiona l smart resources. Fog ecosystems are made up of multiple nodes.
Therefore, the measurement is distributed and can be less
2.3 Augmented reality economically viable than that of a centralized cloud
In latency-sensitive augmented-reality systems, fog structure. Efficacious filtering and sampling methods, and
computing plays a vital role. For instance, an online brain- joint computation and network asset management will
computer-relation game, the EEG Tractor Beam augmented reduce energy utilization in fog atmosphere using powerful
multiplayer, based on electroencephalogram readings communication protocols such as CoAP.
gathered by sensors. On fog modules, it carries out
“continuous real-time brain-state classification and then IV. CONCLUSION
tunes cloud servers, classification models”[1]. A wearable
method of cognitive assistance that uses Google Glass The seamless integration of edge and cloud resources
technology assists individuals with impaired mental acuity is enabled by Fog computing. It supports the decentralized
perform multiple activities, including asking them the names and smart processing of unparalleled IoT sensor-generated
of individuals they encounter but don't remember. For data volumes deployed for smooth physical and cyber-
Delay-tolerant tasks like reporting errors and logging, environment integration. This will bring many advantages to
devices interact with the cloud in this application. For “time- society by, for instance, allowing smart applications for
sensitive activities, the computer streams video from the healthcare. In this way, the further enhancement of fog
Glass Camera to the fog processing units. The scheme computing could assist the IoT to achieve its enormous
shows how” [1] end-to-end latency is significantly reduced potential.
by using nearby fog instruments.
REFERENCES
III. CHALLENGES
[1]. M. A. Bouras, F. Farha and H. Ning, "Convergence of
To realize the full potential of fog computing, there are computing, communication, and caching in Internet of
several threats, including Managing load distribution among Things," in Intelligent and Converged Networks, vol. 1,
edge and cloud services, monitoring and sharing of APIs and no. 1, pp. 18-36, June 2020, doi:
services, and SDN communications. 10.23919/ICN.2020.0001.

IJISRT21MAR278 www.ijisrt.com 316