Review Paper On Edge Computing

Akash Ranjan
Undergraduate
(Lovely Professional University)

Abhishek Bhattacherjee
Professor
(Lovely Professional University)
Abstract—With the rapid expansion of the Internet of which will lead to delays in data transmission. In certain
Everything (IoE), the number of smart devices connected to the situations where immediate feedback is needed, such as
internet is escalating rapidly, generating vast amounts of data. traffic monitoring, cloud computing will be unable to satisfy
However, this phenomenon has resulted in several problems, real-time business requirements.
such as bandwidth overload, slow response times, and
inadequate security and privacy in traditional cloud computing Security: To give an illustration, while using different apps
models. In today's intelligent society, traditional cloud on a smartphone, the apps may request access to the user's
computing is no longer sufficient to meet the diverse needs of data, including private information. When this data is
data processing. Consequently, edge computing technologies uploaded to the cloud center, there is a significant likelihood
have emerged as a new computing paradigm that performs of it being vulnerable to privacy breaches or cyber-attacks.
calculations at the edge of the network, emphasizing proximity
to the user and data source. At the network's edge, edge Energy : As the count of smart devices surges, the energy
computing is lightweight and suitable for local, small-scale data consumption of data centers in China has risen considerably.
storage and processing. This article reviews the related Despite attempts to enhance the efficacy of energy
research and findings of edge computing. It begins by consumption in cloud computing, the mounting need for data
summarizing the concept of edge computing and comparing it energy consumption surpasses it. With the advent of an ever-
to cloud computing. Next, it outlines the architecture of edge evolving intelligent society, the energy consumption
computing, keywords, technology, security, and privacy demands of cloud computing will inevitably escalate.
protection. Finally, it concludes by summarizing the
applications of edge computing.
a) Cloud Computing And Edge Computing: In the
Keywords—Edge computing, cloud computing, Internet of
Things.
past, prior to the inception of edge computing, traditional
cloud computing utilized the network to transfer all data to
Introduction the cloud computing center, and addressed computing and
As society becomes more intelligent and people's needs storage issues in a centralized manner. The evolution of
evolve, intelligence has infiltrated various industries and cloud computing is described in literature [15], where in
everyday life. Edge devices have become ubiquitous in all August 2006, at the search engine conference (Sessane Jose
aspects of society, from smart homes and autonomous 2006), the CEO of Google introduced the idea of cloud
vehicles in transportation to intelligent production robots computing for the first time, officially unveiling the concept.
and cameras in manufacturing. As a result, the number of As Google and other search engines continued to advance,
devices connected to the internet has surged. According to cloud computing grew robustly. At present, cloud computing
the Global Cloud Index by Cisco, in 2016, there were 17.1 has become a powerful network service platform with a
billion devices connected to the internet. By 2019, the total range of technologies including distributed computing, load
amount of data traffic in global data centers is expected to balancing, parallel computing, network storage, and
reach 10.4 Zettabyte (ZB), with 45% of the data being virtualization.
stored, processed, and analyzed on the edge of the network. However, with the increasing popularity and development of
Additionally, the number of wireless devices connected to the Internet of Things (IoT) in people's lives, the number of
the network is projected to exceed 50 billion by 2020. The IoT-connected devices has surged, and an enormous amount
volume of data generated by devices worldwide has also of data is being generated. The network bandwidth of cloud
increased significantly, from 218ZB in 2016 to 847 ZB in computing is no longer sufficient to meet the needs of real-
2021. The International Data Corporation (IDC) predicts time systems [16] and time-sensitive operations.
that by 2020, the number of terminals and devices Consequently, cloud computing models are flawed in terms
connected to the network will exceed 50 billion, and the of load management, real-time performance [17]–[19],
total global data in 2020 will surpass 40 ZB. transmission bandwidth, energy consumption, and data
Cloud-based big data processing has presented numerous security and privacy protection [20].
drawbacks considering the persistent and significant increase
in data volume and the various demands for data processing.
Real Time: If numerous edge devices are incorporated, a
b) Edge Computing:As the Internet of Things (IoT)
significant quantity of data from the terminals will still need
continues to gain traction and expand in people's daily lives,
to be sent to the cloud for processing. This will result in a
substantial increase in intermediate data transmission, the number of devices connected to it has surged, resulting
causing a reduction in data transmission performance and in a vast amount of data generation. However, the current
placing a heavy load on network transmission bandwidth, network bandwidth of cloud computing falls short of

XXX-X-XXXX-XXXX-X/XX/$XX.00 ©20XX IEEE

meeting the real-time demands of systems and time- The architecture of edge computing involves creating a
sensitive operations [16]. This has exposed the drawbacks of network of edge devices that connect the cloud services to
cloud computing models in terms of load management, real- the edge of the network, which is closer to the terminal
time performance [17]–[19], transmission bandwidth, devices. This is done by introducing intermediary edge
energy consumption, and data security and privacy devices between the terminal devices and the cloud
protection [20]. computing services. The collaboration between the cloud
and edge layers is typically divided into three layers: the
As a solution, edge computing aims to offer services and terminal layer, the edge layer, and the cloud computing
perform computations at the network's edge and the site of layer. Each layer performs specific functions that contribute
data generation. Edge computing migrates cloud to the overall edge computing architecture.
computing's network, computing, storage capabilities, and
resources to the network's edge and provides intelligent 1. TERMINAL LAYER
services to meet the IT industry's critical needs for agile
linking, real-time business, data optimization, application The terminal layer of the edge computing architecture
intelligence, security and privacy, and to meet the low comprises various devices that are connected to the
latency and high bandwidth requirements of the network. network, including smartphones, sensors, cameras,
Nowadays, edge computing has become a research focus smart cars, and other Internet of Thing’s devices. These
[10]–[14]. devices not only consume data but also provide data. To
reduce service delay, the focus is on the devices' data
perception rather than computing power. Therefore,
millions of devices in the terminal layer collect raw data
Similarity and Difference Between Cloud And Edge and upload it to the upper layer, where it is stored and
The emergence of edge computing does not mean that processed.
cloud computing will be replaced. Instead, both technologies
should coexist and complement each other in network, 2. BOUDARY LAYER
business, application, and intelligence aspects to facilitate the
digital transformation of the industry. Although data can be The edge layer is the central component of the three-tier
processed at the edge, it still needs to be summarized in the architecture and is located at the network's edge. It
cloud for in-depth analysis and meaningful results. In the comprises widely distributed edge nodes situated between
context of the Internet of Things, if all the data generated by terminal devices and clouds, such as base stations, routers,
connected devices are sent to the cloud, it will cause a huge
access points, switches, and gateways. The edge layer
load. Edge computing is necessary to share the cloud's
burden and handle tasks within its scope. Cloud computing enables the access of terminal devices to the network, stores
and edge computing work together in many real-life and processes data uploaded by these devices, and transmits
applications, including intelligent manufacturing, energy, the processed data to the cloud. Due to its proximity to the
security, privacy protection, and smart homes, to promote the user, the data transmission to the edge layer is ideal for real-
progress of the Internet of Things. time data analysis and intelligent processing, making it more
efficient and secure than cloud computing.
Cloud computing has its advantages in processing large-scale
data and conducting in-depth analysis for non-real-time data 3. CLOUD LAYER
processing, such as business decision-making. In contrast, Cloud computing remains the most dominant data
edge computing focuses on small-scale real-time intelligent
processing center among the various federated services
analysis, which can better meet local business needs. Edge
offered in cloud-edge computing. The cloud computing
computing can reduce the burden on network resources by
storing and processing data locally, thus improving the layer is composed of a set of high-performance servers
utilization efficiency of network bandwidth. Both cloud and storage devices that possess immense computing
computing and edge computing play a crucial role in the and storage capabilities. This layer is best suited for
future development of intelligent Internet of Things. areas that demand significant amounts of data analysis,
such as regular maintenance and business decision
Text heads organize the topics on a relational, support. The cloud computing center can serve as a
hierarchical basis. For example, the paper title is the primary permanent data repository for the edge computing layer,
text head because all subsequent material relates and and it can perform analysis and processing tasks that are
elaborates on this one topic. If there are two or more sub- beyond the capabilities of the edge computing layer, as
topics, the next level head (uppercase Roman numerals)
well as integrate global information. Moreover, the
should be used and, conversely, if there are not at least two
cloud module can dynamically adjust the deployment
sub-topics, then no subheads should be introduced. Styles
named “Heading 1”, “Heading 2”, “Heading 3”, and strategy and algorithm of the edge computing layer
“Heading 4” are prescribed. based on the control policy.

TABLE I. DIFFRENCE BETWEEN CLOUD AND EDGE COMPUTING APPLICABLE NETWORK REAL CALCULATION
SITUATION BANDWIDTH TIME MODE
PRESSURE

c) Architecture Of Edge Computing

CLOUD GLOBAL MORE HIGH LARGE SCALE
CENTRALIZED

EDGE LOCAL LESS LOW SMALL SCALE

INTELLIGENT
 REFERENCES
The template will number citations consecutively within
brackets [1]. The sentence punctuation follows the bracket
[2]. Refer simply to the reference number, as in [3]—do not
use “Ref. [3]” or “reference [3]” except at the beginning of a
sentence: “Reference [3] was the first ...”

Figure : Edge Computing Reference Number footnotes separately in superscripts. Place the
actual footnote at the bottom of the column in which it was
EDGE COMPUTING APPLICATION cited. Do not put footnotes in the abstract or reference list.
Use letters for table footnotes.
Every new technology that emerges is eventually applied
to different scenarios. The primary factor that determines the Unless there are six authors or more give all authors’
viability of a new technology is whether it can efficiently names; do not use “et al.”. Papers that have not been
solve real-world problems. In this context, we present the published, even if they have been submitted for publication,
challenges and opportunities that edge computing is likely to should be cited as “unpublished” [4]. Papers that have been
encounter during its application. As edge computing accepted for publication should be cited as “in press” [5].
continues to improve in theory, more and more applications Capitalize only the first word in a paper title, except for
based on edge computing are becoming a reality. This proper nouns and element symbols.
section seeks to provide an in-depth understanding of edge
For papers published in translation journals, please give
computing through its application in five different scenarios,
the English citation first, followed by the original foreign-
namely the edge computing video cache scene, 5G
language citation [6].
communication scene, edge computing network video live
broadcast scene, predictive maintenance scene, and security
monitoring scene. [1] G. Eason, B. Noble, and I. N. Sneddon, “On certain integrals of
Lipschitz-Hankel type involving products of Bessel functions,” Phil.
A) Edge Computing Video Cache Trans. Roy. Soc. London, vol. A247, pp. 529–551, April 1955.
(references)
Every new technology that emerges is eventually [2] J. Clerk Maxwell, A Treatise on Electricity and Magnetism, 3rd ed.,
applied to different scenarios. The primary factor that vol. 2. Oxford: Clarendon, 1892, pp.68–73.
determines the viability of a new technology is [3] I. S. Jacobs and C. P. Bean, “Fine particles, thin films and exchange
whether it can efficiently solve real-world problems. anisotropy,” in Magnetism, vol. III, G. T. Rado and H. Suhl, Eds.
New York: Academic, 1963, pp. 271–350.
In this context, we present the challenges and [4] K. Elissa, “Title of paper if known,” unpublished.
opportunities that edge computing is likely to [5] R. Nicole, “Title of paper with only first word capitalized,” J. Name
encounter during its application. As edge computing Stand. Abbrev., in press.
continues to improve in theory, more and more [6] Y. Yorozu, M. Hirano, K. Oka, and Y. Tagawa, “Electron
applications based on edge computing are becoming spectroscopy studies on magneto-optical media and plastic substrate
interface,” IEEE Transl. J. Magn. Japan, vol. 2, pp. 740–741, August
a reality. This section seeks to provide an in-depth 1987 [Digests 9th Annual Conf. Magnetics Japan, p. 301, 1982].
understanding of edge computing through its [7] M. Young, The Technical Writer’s Handbook. Mill Valley, CA:
application in five different scenarios, namely the University Science, 1989.
edge computing video cache scene, 5G
communication scene, edge computing network IEEE conference templates contain guidance text for
video live broadcast scene, predictive composing and formatting conference papers. Please
ensure that all template text is removed from your
maintenance scene, and security monitoring
conference paper prior to submission to the
scene. conference. Failure to remove template text from
your paper may result in your paper not being
published.