International Journal of Computer Networks and Applications (IJCNA)

DOI: 10.22247/ijcna/2020/203914 Volume 7, Issue 5, September – October (2020)

SURVEY ARTICLE

Concepts and Contributions of Edge Computing in

Internet of Things (IoT): A Survey
S. Magesh
Maruthi Technocrat E Services, Chennai, Tamil Nadu, India.
[email protected]

J. Indumathi
Department of Information Science and Technology, Anna University, CEG, Chennai, Tamil Nadu, India.
[email protected]

Radha RamMohan. S
Department of Computer Applications, Dr. M.G.R Educational and Research Institute, Chennai, Tamil Nadu, India.
[email protected]

Niveditha V. R
Department of Computer Science and Engineering, Dr M.G.R Educational and Research Institute, Chennai,
Tamil Nadu, India.
[email protected]

P. Shanmuga Prabha
Department of Computer Science and Engineering, Saveetha School of Engineering, Saveetha Institute of Medical
and Technical Sciences, Chennai, Tamil Nadu, India.
[email protected]

Published online: 23 October 2020 1. INTRODUCTION

Abstract – Edge has become a growing trend in recent years. Edge computing fetches memory and computes better than
Bringing computing and analytics remarkably close to the data traditional data centre by bringing them significantly closer to
where it originated is the leading cause of edge computing. As the location whenever they are needed frequently in the form
the data is growing day by day, there arises the bottleneck in
of local devices or any physical units across different areas. It
computation and network layers. Due to the enormous growth of
Internet of Things (IoT) devices with its recent applications, the is a disclosed platform integrating network, computation,
need for real-time computation has readily driven edge storage, and other application entities on its edge of the
computing. Today data processing is an excellent paradigm for network which is physically close to the data [1]. As the
real-time data. In the integration of various IoT devices to solve number of smart devices increases, day by day, the
the computing perplexities, created the emergence of the Edge consumption of power in data centres also increases rapidly.
computing. This paper clarifies concepts and contributions of In this case, the cloud cannot improve computing efficiency
edge computing associated with IoT devices. The proposed work [2] to meet the increasing demand for the source of power.
produces a thumbnail survey on edge computing and its Cloud computing has a strategic form of data centres in its
performance management towards IoT devices. The
hand [3]. Of course, with substantial resource capacities with
characteristics and architecture of Edge computing over IoT
devices are furnished. The state-of-the-art on edge computing itself. As cloud servers handle more user applications, the
applications in the real-time scenario is discussed in this article. cloud faces significant challenges to meet the energy
The proposed work explores the key benefits of Edge computing consumption demand in the data centre. Now, improving the
towards IoT devices, along with the comparative principles of efficiency of energy consumption is existing research going
edge computing over the Cloud, are represented. The existing on [4]. But the need for various tactile internet applications
challenges of edge computing are also discussed in this work. that include augmented reality and virtual reality, the demand
Index Terms – Edge Computing, IoT devices, Data Processing, for bandwidth dragged the enterprises towards the edge, and it
Performance Computing. became essential to have the servers very close to the edge for
proper response. This response made edge computing come

ISSN: 2395-0455 ©EverScience Publications 146

International Journal of Computer Networks and Applications (IJCNA)
DOI: 10.22247/ijcna/2020/203914 Volume 7, Issue 5, September – October (2020)
SURVEY ARTICLE
into existence, focusing the distribution of infrastructure computing technologies solving many real-time issues with
towards the edge devices and edge nodes [5, 6]. IoT devices.
Next coming to the Internet of Things (IoT) applications, the Shufen Wange [14] discusses on development of smart homes
cloud computing architecture faces severe problems. Mobile using IoT devices. These IoT devices monitor the internal
devices that are interconnected with cloud servers are placed status of the home, regulates the home environment providing
at a certain distance, trying to obtain better services that home safety and proper security. As everything cannot be
increase the load on Radio Access Network (RANs), which uploaded in the cloud especially some video data, edge
results in very high latency [7]. IoT is drawing attention computing monopolized for processing home data through the
towards the digital transformation as the end-user demands gateway reducing the overflow of household data, prevent the
are increasing [8]. Cisco reported that more devices similar to leakage of data improving the privacy of the edge node.
50 billion are connected to IoT [9].The new challenges faced
M. Chen et al. [15] devised Edge and cognitive computing
by IoT, including latency, constraint in capacity, resources,
(ECC) that endows smart healthcare systems when the
devices, uninterrupted connectivity and reliability, which
patients are in an emergency. The edge devices assigned
cannot be promised by rationalized cloud computing[10]. IoT
according to the patient's health status. On assigned edge
generates a massive quantity of data and exchanges those data
devices that process robustly with low latency, cognitive
relevant to real-world objects [11]. These IoT devices require
computation is carried out. The cognitive data engine collects
a large scale of the network, heterogeneity in-network level
the patient's data, and the cognitive resource engine collects
and devices, considerable number of operating system events
the information relevant to edge, Cloud, and other network
and applications. These features are a challenging task for
resources. The data is sent to these allocated resources
real-time data, and it becomes very tough to be implemented
according to the risk level and requirement of the patient.
by IoT and cloud environment [12]. Sending every data to the
cloud needs high bandwidth which is not affordable with Zhang et al. [16] proposed video analysis in edge computing.
cloud framework. Recent research is investigating that only This video analysis is named as Edge video Analysis for
edge nodes can support these IoT devices and its requirements Public Safety framework (EVAPS). With this edge computing
[13]. The proposed work discussed the concepts of edge framework, the video analysis workload is distributed
devices in terms of architecture, potentialities, benefits, state- optimally on both the edge nodes and the cloud Centre. The
of-the-art on real-time applications, use cases, complementary edge node prevents unnecessary data transmissions to the
concepts over cloud, challenges, and future directions. The cloud.
different aspects of edge services depicted in Figure 1.
L.U. Khan et al. [17] Proposed a smart city scenario using
edge computing. In this system, the data is taken from the
environment. Vehicles run applications and transmit data to
the edge computing servers via roadside units (RSU) for
further analysis and processing. The RSUs and edge servers
are interrelated and finally connected to the cloud. For
seamless connectivity and security, RSUs are deployed. The
main objective of the proposed work is to reduce the cost of
network deployment, fulfilling the quality of service.
B. R. Stojkoska and K. Trivodaliev [18] proposed an IoT and
edge computing based smart home with a three-tier
framework. The unit comprises smart homes tier, nano grid
tier and microgrid tier. The first tier consists of a household
item that enables wireless communication. The data is
collected at the sink where it is locally processed and stored.
In the second grid, sinks of smart home communicate with
one another. This communication is enabled through mesh,
Figure 1 Edge Computing Progress in Different Aspects
cluster, or star. Sinks of various smart homes interact via IoT
2. LITERATURE SURVEY ON EDGE COMPUTING gateways enabled by the edge computing for reliable
AND IOT computation.

This section presents a comprehensive survey on the edge Richard Olaniyan et al. [19] defines the Opportunistic Edge
computing applications and potentialities embedded with IoT computing (OEC) paradigm. It also furnishes the management
devices in diverse areas. The section also discusses edge framework to construct, organize and monitor the
infrastructure of the scalable edge that also involves the

ISSN: 2395-0455 ©EverScience Publications 147

International Journal of Computer Networks and Applications (IJCNA)
DOI: 10.22247/ijcna/2020/203914 Volume 7, Issue 5, September – October (2020)
SURVEY ARTICLE
stakeholders. OEC utilizes the broker to launch pools of first phase, the network is divided into small sections where
resources with the contributions of the end-user at the edge. the candidate node is determined. In the next phase, the utility
Thus, OEC pools are used to bring the cloud nearby for a metric evaluates the performance of every candidate node by
predefined duration. The author also compares the OEC validating the link quality and correlation. To achieve
model with that of the existing models. The experimental maximum throughput, the best node is deployed in the
results depict the merits and demerits of OEC over cloud network. The proposed work describes certain concepts of
computing and fog computing. The paper also summarizes the edge computing and provides at most contributions in terms
resource management factors in OEC environments. of real time applications and use cases in the IoT
environment.
Yunkon Kim and Eui-Nam Huh [20] explores data caching
concepts in edge computing. The state-of-the-art relays on 3. CONCEPTS OF EDGE COMPUTING
machine learning, reactive and proactive caching. A
The smartness of Edge computing is that edge devices collect
lightweight caching algorithm, namely EDCrammer (Efficient
and share data without transporting to another end of the
Data Crammer), is implemented to manage computing
server. It brings the data close to the point of interaction. The
resources. This algorithm uses an enhanced PID (Proportional
concept of edge computing is stated as a computing mode that
Integral-Differential) controller for data streaming.
executes at the edge of the network. The cloud server is
EDCrammer helps in streaming data traffic, reduce the uplink
considered as downlink data, and IoT is justified as uplink
load in the cloud by providing high-quality video services.
data and edge to edge computing is deliberated as arbitrary
This model cache data efficiently, and the caching rate is
computing [24-26].
controlled amidst the cloud and the edge nodes. The algorithm
produces a hit ratio of about 96% with a cache capacity of
around 1.5MB. Thus, with the edge node, the user can receive
better Quality of Service (QoS) by decreasing the load
congestion on the data centre.
Y. Liu et al. [21] explored the computation offloading using
edge computing. The main objective is to provide a better
quality of service. The system framed a novel mechanism
consisting of some set of interactions carried over amidst the
cloud and local edge for the Stackelberg game. In this game,
cloud server operator (CSO) serves as a leader that produces
payments to end-server owners (ESOs), and this ESO is
considered as a follower determining the computation offload
for CSO. This model promises efficient computation
offloading and extends its cooperation among ESOs.
J. Yang et al. [22] proposed a new model of smart IoT toys. Figure 2 Concepts of Edge Computing Framework
The author deployed a new prototype for a data exchange Edge computing can store and processing data significantly
accounting system based on edge computing. This prototype faster when compared to other devices. It is implemented for
was designed with Hyper ledger Fabric-based blockchain real-time applications. In edge computing, algorithms run on
system. Exchange and payments of data are made amidst an edge server, or it increases the capacity of the smartphones
demanders and suppliers through blockchain consortium. by running algorithms via smartphones. The storage resources
Smart contracts are deployed for validating the data exchange include cloudlets, data centre and fog nodes. And it operates
records and billing between the peers. The proposed system at the edge of the network devices that remain nearing to the
provides secure and reliable interactions between smart toys sensors or mobile devices [27]. The main objective of edge is
and IoT devices. This edge computing-based data exchange to transmit the network of the cloud, storage, resources to the
(EDEC) comprises five phases, namely registration, the respective network edges providing brilliant services at the
release of data products, order generation, the transmission of edge of the network devices to meet the emergency needs of
data, accounting details and finally payment. The transactions the industries and factories. The edge computing is designed
are all recorded and stored on edge computing. to meet the requirements of low latency with high bandwidth
Z. Zhao et al. [23] deployed a three-step implementation on the network edge. Now, edge computing is considered for
approach to reduce the number of edge servers for improving many research purposes [28- 32]. The proposed work
the throughput between Edge nodes and IoT devices. The discusses the architecture, potentialities, benefits, use cases,
proposed model consists of three concepts such as state-of-the-art applications, and challenges of edge
discretization, utility metric and deployment algorithm. In the computing in IoT devices shown in the above Figure 2.

ISSN: 2395-0455 ©EverScience Publications 148

International Journal of Computer Networks and Applications (IJCNA)
DOI: 10.22247/ijcna/2020/203914 Volume 7, Issue 5, September – October (2020)
SURVEY ARTICLE
3.1. Edge Computing Architecture Edge computing can also be defined in terms of computing
power and latency. It allows us to perform necessary
Edge computing brings the resources of computation, storage,
computations with low latency with high computing-intensive
and networking significantly closer to the devices, users, and
calculations. Cloud, edge node, edge gateway and edge
applications. It is viewed as one of the significant
devices are the most essential layers in handling application
technologies associated with the Internet of Things and
and transmitting data at the edge devices. Figure 4 illustrates
artificial intelligence in the next generation of networks [33].
the layers of edge computing architecture [34].
It enables the data to travel at remarkably high speed without
transferring data to the cloud or data centre. Edge computing
is multilayered distributed architectures that balance the
workload amidst edge layer, edge cloud, edge network and
enterprise layer. Edge computing comprises three nodes:
Device edge, Local edge, and Cloud. Figure 3 illustrates the
overview of an edge architecture with the relevant
components within each node. Industry solutions/Apps can be
considered as duplicate components that exist both in the
device edge and local edge. Specific workloads are integrated
either with the device edge and local edge, and they can be
dynamically migrated amid of the nodes either manually or
automatically. Local edge node manages multi-cloud
management, network services.
 The device edge is the place where the device is occupied.
These devices include cameras, sensors and any physical
devices gathering data and interact with edge data. Some
applications such as AI models with deep learning, video Figure 4 The Architecture of Edge Computing
analytics are managed on these edge devices. The purpose of the cloud is to serve as long term storage and
 Local edge encloses both applications and the network. coordinates the resources in the case of irregular tasks. The
The applications of the device edge reside in this node. edge node by default acts as downstream. These nodes
The non-runnable applications on the device edge are possess high computation power, 2q manages the network
executed at local edge. Such as IoT processing. Some traffic. On the other hand, edge gateway manages
virtualized and containerized network layers will run on computations, and it serves as a platform for lower-level
this local edge. devices such as mobile phones, cars, and sensors. In contrast,
edge devices are composed of limited resources that include
 Cloud is an environment where everything is brought single sensors and traffic lights.
together. It can either run at premises or as a public cloud.
The applications that are not able to run in either of the 3.2. Potentialities of Edge Computing
nodes are handled by this cloud using orchestration layers. 3.2.1. Edge Computing in Business
Nowadays, autonomous vehicles and medical sensors have
become more common. Business and organization can extend
their network services with the edge computing framework.
Edge computing also identifies the machinery issues and fault
tolerance level that provides safety for industrial
manufacturers.
3.2.2. Edge Computing and IoT
Gadgets that are connected to the internetwork generate
massive data which remains a challenge in managing and
analyzing those collected data. This paradigm can be solved
by edge computing by pushing the data from a centralized
system to local edge centres that are close to the source. These
data are analyzed in relative positions with low cost and better
efficiency. The edge devices also solve low connectivity and
Figure 3 Edge Computing – An Overview transferring the cost of data. In the case of industrial IoT, edge

ISSN: 2395-0455 ©EverScience Publications 149

International Journal of Computer Networks and Applications (IJCNA)
DOI: 10.22247/ijcna/2020/203914 Volume 7, Issue 5, September – October (2020)
SURVEY ARTICLE
devices are self-regulated and dynamically respond to changes 3.3.1. Cost-Effective
where the data are collected, analyzed, and implemented
Using IoT devices are very costly because of their high
robustly.
bandwidth and storage capacity. With the use of edge
3.2.3. 5G Networks computing for IoT devices reduces the bandwidth and data
storage requirement. Edge computing also replaces the
The accumulation of 5G networks increases bandwidth and
datacentres. Thus, edge computing significantly reduces the
transmits large volumes of cellular data. Integrating 5G
cost in IoT implementation. Infrastructure cost is also reduced
networks with edge computing helps to keep more data on the
by edge computing as only the relevant data alone are
edge node, and this reduces the latency issue by which data
transmitted to the cloud, thus reducing the bandwidth.
are transmitted via suboptimal bridges before they reach the
concerned users. 3.3.2. Interoperability
3.2.4. Location Optimization Edge computing permits interoperability between the
connected devices and legacy devices. It can convert
As per datacenters specification, it translates 80-90 ms for
communication protocols to suitable IoT connected devices.
information from coast to coast in the United States. But with
edge data centres, it takes just 5 – 10 milliseconds when the 3.3.3. Faster Response Time
processing is close to the source. The data processing takes
Edge computing response at a remarkably high speed by
place in microseconds when the processing is done inside the
reducing the latency, thereby increasing the network speed. It
edge device.
processes the data about the source at which the information
3.3. Benefits of Edge Computing in IoT is gathered, thus reducing the distance that it should travel.
Therefore, the result is processed in microseconds. Edge
In Edge computing, data is processed at the location where it
computing is significantly faster in response to quality
is generated. Every process is done instantaneously. Edge
service.
computing stores and analyzes the data at a remarkably high
speed than the cloud. Another vital aspect of edge computing 3.3.4. Reliability
is data security whose primary need is to establish the
Edge computing provides uninterruptable services as it does
integrity and security of data. Many types of research are
not depend on an internet connection or any other servers.
focused on cloud computing [35], mobile cloud computing
Data storage and processing are done locally using micro data
[36]. Thus, primary research is based on data security in edge
centres. Thus, edge computing assures reliable connection for
computing which is done by migrating a secure data manner
the IoT devices where it can be used in remote locations.
in a distributed environment with limited terminal resources.
Some of the potential benefits of an edge are discussed in this 3.3.5. Security
paper. The key benefits are illustrated in Figure 5.
Edge computing provides security as it is not dependent on
network connections. Edge computing transmits only relevant
information to the cloud. And it distributes processed work
across different data centers and various devices. Thus,
Distributed Denial of Service (DDoS) attack is prevented. As
the data is stored and analyzed locally, it becomes easy to
monitor the security team.
3.4. Edge Computing Vs. Cloud Computing
Edge computing is a group of local processing micro data
centres and reduces the burden of a cloud. It is one regional
office that takes over the computing tasks of the cloud instead
of sending it to the data centre, which is an extremely far
distance. Edge computing is contemplated as the complement
of the cloud, and it does not mean to replace the cloud
services [37]. They both are different that cannot be
interchangeable. And these two technologies cannot be
replaced where edge computing processes the time-sensitive
data, but the cloud is not so. Nevertheless, edge computing is
Figure 5 Key Benefits of Edge Computing preferred in remote locations than the cloud, especially if
there is limited or no connectivity to the central location. The

ISSN: 2395-0455 ©EverScience Publications 150

International Journal of Computer Networks and Applications (IJCNA)
DOI: 10.22247/ijcna/2020/203914 Volume 7, Issue 5, September – October (2020)
SURVEY ARTICLE
complementary aspects of edge and cloud are furnished in possible. This system reduces the parking queues and finds
Table 1. the space very quickly by identifying the free parking slots.
The concept of Image detection systems with surveillance
Complementary Cloud Edge Computing cameras is utilized, which is done by convolution neural
Aspects Computing network (CNN) techniques. Edge computing also provides a
Data Processing Data collected, Data collected, promising solution in offering services to video streaming in
analyzed, and analyzed, and smart cities [38]
processed at the processed 3.5.2. Use Case in Connected Cars
other end. It is a locally. It is a
centralized distributed based In this system, data is collected from the various sensors and
computing relevant information is sent to the base station, which is then
transmitted to the edge node. In the case of video, it is
Data Transfer Any data No need for transmitted to the data centre of the manufacturer. Thus,
processed is transferring. It incorporating connected car data with the sensors improves
transferred to a reduces the traffic management such as traffic flow, traffic light controls
cloud workload on and public transport. Human lives are saved by these
network and connected cars. However, internal factors like airbag
server triggering, speed, brake, etc., provide immediate response to
Data Management Cloud dependent It handles with put the human lives out of danger.
on the data purpose-built 3.5.3. Use Case in Monitoring Machines
centre systems and frees
up the cloud Monitoring machines are possible with IoT sensors. It
computing monitors the metrics of machines. These IoT sensors leverage
edge computing stabilities and analyze the gathered data on
the local edge. From the edge, it is transmitted to the cloud.
Minimum Cloud uses a data Filters the Edge computing performs accurate measures and maintains
connectivity cost centre to filter sensitive data at the process before any interruptions occur. In connected
the sensitive data the source itself parking meters, a resilient network is affixed for monitoring
locally. systems. This enables the data to transmit to the hub in the
Uninterrupted At cloud centres Uses micro data event of fault occurrence. As IoT sensors are embedded with
connection with the connectivity centres data are the traffic lights and all smart devices in the environment, the
Reliability is limited with stored locally edge computing monitors these data locally. The data are
some data loss. and processed. communicated and transmitted centrally only on fault reports.
Consume less 3.5.4. Use Case in Healthcare
bandwidth.
IoT devices play a significant role in providing solutions for
Table 1 Benefits of Edge Computing Vs. Cloud Computing the patient's health status. It receives real-time data of the
patient from the cloud and offers robust analyzes in real-time.
3.5. Edge Computing Use Cases
It provides a better solution when the patient is under critical
Edge computing ascertains better performance and security condition. Edge computing is capable of processing and
for IoT applications. IoT has already entered market in analyzing data and helps physicians to provide the correct
numerous ways. Now, enterprises and industries using IoT are diagnosis. Edge computing permits us to manage the
in thirst of how edge computing applications can be converted connectivity, and data processing is done significantly closer
into immense beneficial use cases. Some of the use cases of where the data is gathered. For example, the Patient's data is
edge computing are described in this section. collected through IoT devices where the patient is monitored
and then transferred to an authenticated electronic health
3.5.1. Use Case in Video Surveillance
record (EHR) with proxy cards. Medical equipment may
Edge computing nodes can recognize entities of interest themselves assemble and process information during
which are then transmitted to the local edge. During fire diagnosis and treatment.
accidents, edge computing can quickly determine whether a
3.5.5. Use Case in Train-Transport
human is caught by fire as the cameras are located close to the
event. And immediately it starts to transmit the video to the A moving train is a fair use case when the train is connected
local edge. This is one of the significant use cases. With this to the internet. Nevertheless, edge computing launches the
surveillance cameras, a smart parking solution is also

ISSN: 2395-0455 ©EverScience Publications 151

International Journal of Computer Networks and Applications (IJCNA)
DOI: 10.22247/ijcna/2020/203914 Volume 7, Issue 5, September – October (2020)
SURVEY ARTICLE
network on the train. A wireless daisy chain is designed with information and vital signs are analyzed at the edge for proper
two edge servers, thus providing coverage inside every diagnosis.; 3. Edge renders augmented reality glasses for
carriage, which in turn linked to the next carriage. This acts as information display of patient medical history; 4. Remote
a standalone network without connecting to the cloud that is ultrasounds that enable remote diagnostics by medical
deployed at the local edge. Edge computing even favours specialists.
when the passengers board the train with their luggage or
3.6.3. On-Premises Patient Monitoring
vehicle. Passengers can have a smart ticket on their phone.
When they reach the vehicle compartment, QR code or any Edge computing enables to process the data locally in the on-
barcode needs to be scanned so that the passenger can get a premises edge on the hospital site. The data privacy is
direct link to an IP camera to view the length of the train on maintained. The information is compiled from various sources
the mobile. within the hospital, and relevant information is extracted.
Abnormal patient's behavior and status are monitored through
3.6. State of the Art on Edge Computing Applications
Artificial intelligence(AI)by the edge and enable right-time
In this section, the state- of- the- art on edge computing is set notifications to the clinicians 360-degree view are made
forth from the aspects of real-time applications. Edge available for complete visibility of the patient dashboards to
computing has extended its applications in a variety of fields send corresponding data which is to be stored in a cloud
such as the gaming industry, retail, fleet management, health securely. Thus, edge computing provides a significant
care, provides services for various sectors such as enterprises, resource for practitioners by reducing the cost per patient and
industries. Few real-time scenarios are discussed in this increase productivity.
section.
3.6.4. Edge Computing in Smart Cities
3.6.1. Autonomous Cars
Autonomous vehicles are rising nowadays, and with the
Today, many industries are launching self-driving cars, as per growing needs of the Internet of Things, smart cities will have
the Gartner report, by 2025, it is said that autonomous the ability to respond instantaneously to the changing
vehicles may upload 1TB of sensor data to the cloud per environment with the help of edge computing. Nowadays, IoT
month. In this context, Edge computing enables driverless devices act as building blocks of a smart city. These devices
cars or self-driving vehicles. This system enhances safety for are embedded in infrastructure of the city that helps to
both drivers and passengers alike. The autonomous vehicle is monitor performance of devices and periodical information
inbuilt with sensors capable of sensing the environment and about these assets. Edge computing transmits necessary
road safety rules without human interaction. Accessing a car technical information that includes traffic, flood, safety, and
with a virtual car key brings a digital solution. This enables monitoring infrastructure. It relays more on-device computing
the drivers to utilize their smartphone or connected gadgets to for decision making in real-time by themselves instead of
access the car and start. The user can access vehicle transmitting data to another node for processing. This mode of
information remotely in a secure manner where the physical instantaneous response is an essential requirement for high
key is not needed. Fleet control, e-autonomy and multimodal bandwidth technologies like connected cars.
autonomy are other features important to this autonomous
3.6.5. Augmented Reality (AR) Gadgets
vehicle. The total care of the driver, the speed of the vehicle,
power, health, and safety are included in fleet management. There exists some wearable AR gadgets like head-mounted
E-Mobility offers one hybrid vehicle with immediate display (HMD) device. This device is applied recently in
communication facilities. On the other hand, multimodal various fields like engineering, logistics medicine apart from
mobility enables users to save time and efficiently travel from defense applications. It consists of two tracking methods that
one place to another. Edge computing enables these vehicles include inside-out tracking and outside-in tracking. In the
to gather data and make brilliant decisions with high-speed former tracking method, external sensor observes the users’
data transmission and provides reliability at low latency. head. Here in this type of tracking, camera is utilized as
external sensor. In the second tracking method, sensor is
3.6.2. Connected Ambulance
computing from the observations. In this type sensor is
Generally, Diagnostic procedures are carried out only after the mounted on the device.
arrival at the hospital even during emergency. In this concern,
3.7. Challenges of Edge Computing
connected ambulance plays a vital role in emergency services.
Edge computing with its low latency, robust data processing Today more and more industries and IT sectors have started
capabilities and mobility at the edge node enables better launching the edge computing devices due to its reliability
treatment by paramedics onsite which is achieved by the and time management. But still, reliability is a huge challenge
following factors: 1. Live streaming of incoming patients' data as edge computing fails to report low fidelity data under some
and status from the sensors enabled through 5G.; 2. Patient

ISSN: 2395-0455 ©EverScience Publications 152

International Journal of Computer Networks and Applications (IJCNA)
DOI: 10.22247/ijcna/2020/203914 Volume 7, Issue 5, September – October (2020)
SURVEY ARTICLE
unpredictable circumstances due to the low level of battery the status of the human interaction.
[39]. Some communication protocol is proposed for IoT data entire edge
which supports many sensor nodes and provides dynamic environment.
network connection [40]. Here some computing challenges of Performance Monitoring the By providing
edge computing are discussed below: Table 2 furnishes some performance of the technology to view
challenges of edge computing and provides some solutions to edge device in end-to-end
meet the challenges. real-time in the monitoring needs
consumer point of to be deployed,
3.7.1. Computation with Hardware Restrictions
view from edge to which helps in
Edge computing is constrained in the technical aspect. It is cloud becomes transactional
challenging to fit more hardware on a data centre in a full difficult. performance from
scale. edge to cloud.
Controlling It sometimes slows A low technology
3.7.2. Accessibility with Operation Issues
Operation down the activity on motion
Edge computing faces some logical difficulties in deploying in managing detection that
IT resources. It does not allow operator costs. On the other physical access, provides a holistic
hand, the company cannot monitor every service and the edge controlling IT view to control
locations. There also occurs operator limitations due to equipment, edge environment
distance, device volume and other geographical issues. managing the when there is some
environment, lagging on power
3.7.3. Remote Paradigm
tracking solving and network.
Skilled operators are needed to operate plug and play network issues
operations. In this scenario, edge computing should provide when
sophisticated facilities in terms of features such as data disconnected.
caching during interrupted connections. Data filtering, Security Control access that Solutions
management of devices, fault tolerance and bandwidth usage. too on all areas at regarding unified
all local edges view at all
3.7.4. Connection-Oriented Issues
seems difficult, physical, network
There is a need for right technology providers to work with all which results in and access of data
latency issues, which plays a vital role in edge computing. threat. with automated
responses need to
3.7.5. Air – Spaced Deployments
be provided
Edge computing faces challenges in managing remote and air- without client
spaced devices. It is necessary to compute a constrained interruption.
location without manual intervention. Table 2 Edge Computing Challenges and Solutions
3.7.6. Security 3.8. Edge Computing: Future Outlook
Another primary consideration is security. This means
communication need to be secure from the data centre to the
edge device. It should also ensure security both during stable
and during transmission, especially sensitive data stored at the
edge device.
Characteristics Challenges Addressing the
Challenges
Scalability Edge computing Intelligent
needs many software needs to
monitors to be deployed to
understand the manage distributed
status and it needs heterogeneous
servers and environments.
network devices. Thus, monitoring
Finally, it becomes the status can be
difficult to view made ease without Figure 6 Future Directions of Edge Computing for IoT

ISSN: 2395-0455 ©EverScience Publications 153

International Journal of Computer Networks and Applications (IJCNA)
DOI: 10.22247/ijcna/2020/203914 Volume 7, Issue 5, September – October (2020)
SURVEY ARTICLE
Edge computing associated with 5G emerges with great data transmission process. This edge computing plays a vital
opportunities in all industries. Its primary motivation is to role in meeting the demands of IoT devices in real-time
bring the data storage and computation process around the characteristics by improving production efficiency besides
data where it is generated. It enables proper data control, specific challenges.
reduces the cost, brings deep insight, and provides continuous
5. CONCLUSION
operations. By 2025, it is assumed that 75% of business data
will be processed at the edge computing when compared This paper discusses some applications and potentialities of
today. Figure 6 depicts the capabilities of Edge computing in Edge Computing in IoT devices. To have a deep insight into
the future. edge computing, this article surveys some additional concepts
of edge computing such as its architecture, state-of-the-art of
There arises a strong prediction for the next few years edge
edge computing applications in real-time scenarios. The paper
market will be headed very stubbornly. It has been estimated
enumerates some of the edge computing use cases in various
that the global revenue will reach $827.6 billion by 2025 for
domains. Nevertheless, the paper also provides some
edge devices and edge networks. The main reason the edge is
complementary aspects of edge computing with that of cloud
popularly growing is that it computes at local and edge
computing. A comparison table has been provided concerning
locations than transmitting to the cloud. It filters the sensitive
the benefits of edge overcloud. And the challenges of edge
data during processing at the edge, and only the selected data
computing and addressing techniques have clearly been
is sent to the cloud. So, edge computing market will be in
described in the proposed work. The paper offers systematic
progress as the amount of data proliferates rapidly. Some
concepts of edge computing from the architecture to the
future predictions of edge computing can be discussed here:
challenges. The in-depth description includes data storage,
 The development of custom form factors will be driven for latency and time-management as edge computing has become
each use case from the processor level. A good example is a research-oriented paradigm. The paper also sketches about
that Hewlett Packard Enterprise (HPE) has invested 145 the challenges of edge computing. Applications of Edge
million US dollars for launching Pensando Systems to computing in IoT devices are furnished in this article. The
market a custom programmable processor. paper investigates the outlook of edge computing. Though it
faces some difficulty edge computing will be the future
 Many providers such as public mega cloud, network in the real world which provides continuous
telecommunication industries, software providers, and data development in the internet, AI, and also plays a vital role in
centre providers are integrating with cloud programming human society and the geological environment.
services with edge computing infrastructure.
REFERENCES
 The multi-vendor packaged solution is chosen by the
companies. A survey by Forrester has revealed that the [1] Wang, R., Yan, J., Wu, D., Wang, H. and Yang, Q., 2018. Knowledge-
centric edge computing based on virtualized D2D communication
main motivation of the organization that needs from edge systems. IEEE Communications Magazine, 56(5), pp.32-38.
computing includes the capacity to handle the future [2] Gao, Y., Guan, H., Qi, Z., Song, T., Huan, F. and Liu, L., 2014. Service
Artificial Intelligence as edge computing minimizes level agreement based energy-efficient resource management in cloud
network latency and affords fast responses. data centers. Computers & Electrical Engineering, 40(5), pp.1621-
1633.
4. DISCUSSIONS [3] Coady, Y., Hohlfeld, O., Kempf, J., McGeer, R. and Schmid, S., 2015.
Distributed cloud computing: Applications, status quo, and
In the technological perspective, edge computing is a challenges. ACM SIGCOMM Computer Communication Review, 45(2),
pp.38-43.
providential framework enriching smart applications with [4] Satyanarayanan, M., 2017. The emergence of edge
expeditious computing and storage resources. Cloud is computing. Computer, 50(1), pp.30-39.
steadily deployed for smart devices with better computation [5] Satyanarayanan, M., Bahl, P., Caceres, R. and Davies, N., 2009. The
and resource management. Due to its latency issues, case for vm-based cloudlets in mobile computing. IEEE pervasive
Computing, 8(4), pp.14-23.
computing and storage resources are transmitted and [6] Elkhatib, Y., Porter, B., Ribeiro, H.B., Zhani, M.F., Qadir, J. and
centralized node to the edge node, thereby reducing the Rivière, E., 2017. On using micro-clouds to deliver the fog. arXiv
burden of cloud computing. Thus, most of the services are preprint arXiv:1703.00375.
pushed towards the edge, as edge computing ensures shorter [7] Peng, M. and Zhang, K., 2016. Recent advances in fog radio access
networks: Performance analysis and radio resource allocation. IEEE
response time with reliability. Edge computing has the Access, 4, pp.5003-5009.
potential to provide better intelligent services with a robust [8] Al-Fuqaha, A., Guizani, M., Mohammadi, M., Aledhari, M. and
response in vehicle automation, video monitoring and other Ayyash, M., 2015. Internet of things: A survey on enabling
real-time applications. As every edge node is placed technologies, protocols, and applications. IEEE communications surveys
& tutorials, 17(4), pp.2347-2376.
remarkably close to the source, data storage and processing [9] Evans, D., 2011. The internet of things: How the next evolution of the
take place in the edge computing node, thus minimizing the internet is changing everything. CISCO white paper, 1(2011), pp.1-11.

ISSN: 2395-0455 ©EverScience Publications 154

International Journal of Computer Networks and Applications (IJCNA)
DOI: 10.22247/ijcna/2020/203914 Volume 7, Issue 5, September – October (2020)
SURVEY ARTICLE
[10] Chiang, M. and Zhang, T., 2016. Fog and IoT: An overview of research [32] Mouradian, C., Naboulsi, D., Yangui, S., Glitho, R.H., Morrow, M.J.
opportunities. IEEE Internet of Things Journal, 3(6), pp.854-864. and Polakos, P.A., 2017. A comprehensive survey on fog computing:
[11] Ganz, F., Puschmann, D., Barnaghi, P. and Carrez, F., 2015. A practical State-of-the-art and research challenges. IEEE Communications Surveys
evaluation of information processing and abstraction techniques for the & Tutorials, 20(1), pp.416-464.
internet of things. IEEE Internet of Things journal, 2(4), pp.340-354. [33] Jararweh, Y., Doulat, A., AlQudah, O., Ahmed, E., Al-Ayyoub, M. and
[12] Razzaque, M.A., Milojevic-Jevric, M., Palade, A. and Clarke, S., 2015. Benkhelifa, E., 2016, May. The future of mobile cloud computing:
Middleware for internet of things: a survey. IEEE Internet Things integrating cloudlets and mobile edge computing. In 2016 23rd
Journal. 3(1), 70–95 (2016). International conference on telecommunications (ICT) (pp. 1-5). IEEE.
[13] Brogi, A. and Forti, S., 2017. QoS-aware deployment of IoT [34] Li, H., Ota, K. and Dong, M., 2018. Learning IoT in edge: Deep
applications through the fog. IEEE Internet of Things Journal, 4(5), learning for the Internet of Things with edge computing. IEEE
pp.1185-1192. network, 32(1), pp.96-101.
[14] Chen, M., Li, W., Hao, Y., Qian, Y. and Humar, I., 2018. Edge [35] Zissis, D. and Lekkas, D., 2012. Addressing cloud computing security
cognitive computing based smart healthcare system. Future Generation issues. Future Generation computer systems, 28(3), pp.583-592.
Computer Systems, 86, pp.403-411. [36] Liang, H., Huang, D., Cai, L.X., Shen, X. and Peng, D., 2011, April.
[15] Zhang, Q., Yu, Z., Shi, W. and Zhong, H., 2016, October. Demo Resource allocation for security services in mobile cloud computing.
abstract: Evaps: Edge video analysis for public safety. In 2016 In 2011 IEEE Conference on Computer Communications Workshops
IEEE/ACM Symposium on Edge Computing (SEC) (pp. 121-122). IEEE. (INFOCOM WKSHPS) (pp. 191-195). IEEE.
[16] 16. Khan, L.U., Yaqoob, I., Tran, N.H., Kazmi, S.A., Dang, T.N. and [37] Wu, Y., Liu, Y., Ahmed, S.H., Peng, J., and Abd El-Latif, A.A., 2019.
Hong, C.S., 2020. Edge computing enabled smart cities: A Dominant Data Set Selection Algorithms for Electricity Consumption
comprehensive survey. IEEE Internet of Things Journal.DOI: Time-Series Data Analysis Based on Affine Transformation. IEEE
10.1109/JIOT.2020.2987070 Internet of Things Journal, 7(5), pp.4347-4360.
[17] Stojkoska, B.R. and Trivodaliev, K., 2017, November. Enabling internet [38] Taleb, T., Dutta, S., Ksentini, A., Iqbal, M. and Flinck, H., 2017.
of things for smart homes through fog computing. In 2017 25th Mobile edge computing potential in making cities smarter. IEEE
Telecommunication Forum (TELFOR) (pp. 1-4). IEEE. Communications Magazine, 55(3), pp.38-43.
[18] Olaniyan, R., Fadahunsi, O., Maheswaran, M. and Zhani, M.F., 2018. [39] Cao, J., Ren, L., Shi, W. and Yu, Z., 2014, October. A framework for
Opportunistic Edge Computing: Concepts, opportunities, and research component selection in collaborative sensing application development.
challenges. Future Generation Computer Systems, 89, pp.633-645. In 10th IEEE International Conference on Collaborative Computing:
[19] Kim, Y. and Huh, E.N., 2019. EDCrammer: An Efficient Caching Rate- Networking, Applications and Worksharing (pp. 104-113). IEEE.
Control Algorithm for Streaming Data on Resource-Limited Edge [40] Debauche, O., Mahmoudi, S., Mahmoudi, S.A., Manneback, P. and
Nodes. Applied Sciences, 9(12), p.2560. Lebeau, F., 2020. A new edge architecture for ai-iot services
[20] Liu, Y., Xu, C., Zhan, Y., Liu, Z., Guan, J. and Zhang, H., 2017. deployment. Procedia Computer Science, 175, pp.10-19.
Incentive mechanism for computation offloading using edge computing:
A Stackelberg game approach. Computer Networks, 129(2), pp.399- Authors
409. Prof. S. Magesh is a notable academician and a
[21] Yang, J., Lu, Z. and Wu, J., 2018. Smart-toy-edge-computing-oriented passionate entrepreneur. He commenced his
data exchange based on blockchain. Journal of Systems academic career as Lecturer and after that elevated
Architecture, 87, pp.36-48. to the level of Assistant Professor, Associate
[22] Zhao, Z., Min, G., Gao, W., Wu, Y., Duan, H. and Ni, Q., 2018. Professor, and Head in the Department of Computer
Deploying edge computing nodes for large-scale IoT: A diversity aware Science and Engineering with his distinguished
approach. IEEE Internet of Things Journal, 5(5), pp.3606-3614. career spanning in engineering institutions over a
[23] Weisong, S., Xingzhou, Z., Yifan, W. and Qingyang, Z., 2019. Edge period of 15 years and 9 years of Corporate
computing: state-of-the-art and future directions. Journal of Computer Experience. He has published 15 refereed
Research and Development, 56(1), p.69. International Indexed journals which include ESCI,
[24] Weisong, S., Hui, S., Jie, C., Quan, Z. and Wei, L., 2017. Edge SCOPUS, WoS, Compendex (Elsevier Engineering Index) to his credit,
computing—An emerging computing model for the Internet of published 2 Patents and a Book with ISBN. He received the Distinguished
everything era. Journal of Computer Research and Development, 54(5), Innovator & Edupreneur Award in the year 2017 and Lifetime Achievement
p.907-924 Award in the year 2019. Presently he serves as the Chairman & Director of
[25] Wang, Y., Liu, M., Zheng, P., Yang, H. and Zou, J., 2020. A smart Magestic Technology Solutions (P) Ltd, CEO of Maruthi Technocrat E-
surface inspection system using faster R-CNN in cloud-edge computing Services and Chief Editor & Director, Jupiter Publications Consortium,
environment. Advanced Engineering Informatics, 43, p.101037. Chennai, Tamil Nadu, India.
[26] Shi, W., Cao, J., Zhang, Q., Li, Y. and Xu, L., 2016. Edge computing:
Vision and challenges. IEEE internet of things journal, 3(5), pp.637- Dr. J. Indumathi, B.E, M.E, Ph.D., D.Sc., D.Litt.,
646. is a Professor from Anna University, Chennai.
[27] Satyanarayanan, M., 2017. The emergence of edge Being a child prodigy, she also, attained her
computing. Computer, 50(1), pp.30-39. B.Sc.(Zoology), M.A(Sociology), M.B.A (Financial
[28] Green, J., 2014. The internet of things reference model. In Internet of Management), Certified Privacy Lead Assessor,
Things World Forum (pp. 1-12). Chartered and Approved Valuer in Computer
[29] Sun, X. and Ansari, N., 2016. EdgeIoT: Mobile edge computing for the Technology, Amateur Wireless Telegraph Station
Internet of Things. IEEE Communications Magazine, 54(12), pp.22-29. License holder (Restricted), Certified Six Sigma
[30] Alrawais, A., Alhothaily, A., Hu, C. and Cheng, X., 2017. Fog Green Belt Holder and Certified Mainframe
computing for the internet of things: Security and privacy issues. IEEE Professional. With a flabbergasting teaching and
Internet Computing, 21(2), pp.34-42. research experience for more than three decades; her interdisciplinary,
[31] Kang, J., Yu, R., Huang, X. and Zhang, Y., 2017. Privacy-preserved expertise is outstanding, spanning diverse subject areas including the latest
pseudonym scheme for fog computing supported internet of cutting-edge technologies. She has a zeal for investigation, research and has
vehicles. IEEE Transactions on Intelligent Transportation innovated and improvised expansively projects which have made a milestone
Systems, 19(8), pp.2627-2637. in the perfection of lifestyle of the humankind. Being a privacy researcher
and activist, she has received many awards and accolades from the various

ISSN: 2395-0455 ©EverScience Publications 155

International Journal of Computer Networks and Applications (IJCNA)
DOI: 10.22247/ijcna/2020/203914 Volume 7, Issue 5, September – October (2020)
SURVEY ARTICLE
organizations, varsities and institutions quantifying to more than 65 awards. Ms. V. R. Niveditha has completed B.E in
Dr. J. Indumathi has published more than 200 scientific papers published in Computer Science Engineering in PB college of
reputed publications, filed 8 Copyrights, 3 patents and Authored 7 books. Engineering , M.Tech Information Security and
Most of her journal publications are indexed in Scopus and Web of Science. Cyber forensics and pursuing doctoral research in
She has delivered several technical invited talks, keynote/plenary speeches the domain of Mobile Security in the Department of
and has chaired several sessions in conferences worldwide. She is the first Computer Science and Engineering of ABET
Lady Secretary for Tamil Nadu Engineering Admissions & first Lady accredited Dr. M.G.R. Educational and Research
Secretary for Tamil Nadu Common Admissions, and successfully Institute, Deemed to be University with Graded
accomplished the counselling process to go online. Autonomous Status, Chennai, India. She published
6 articles in the National conferences, 15 articles in
Dr. S. Radha Rammohan, M.Sc., M.Tech., Ph.D., refereed International Indexed journals which include ESCI, WoS, SCOPUS,
MIE(India),CE(IEI),B.Ed. is a Professor in the Compendex (Elsevier Engineering Index) to her credit, published 2 Patents
Department of Computer Applications, Dr. M.G.R. and 4 Books with ISBN. She has submitted several funded proposals to
Educational and Research Institute, Deemed to be Government funding Agencies.
University with Graded Autonomy Status, Chennai-
600095,Tamilnadu, India. He has 20 years of Ms. P. Shanmuga Prabha, is an Assistant
Experience in Teaching in India and Abroad, 7 Professor in the Department of Computer Science
Years in Industry. He has CCNA certification and and Engineering, Saveetha School of Engineering,
was Cisco Certified Authorized Instructor. He has 4 Saveetha Institute of Medical and Technical
Patent published and 2 Book Published. His area of Sciences, Chennai, Tamil Nadu, India with 2 year 4
research is Mobile Adhoc networks, IoT, Cloud Computing, Blockchain, months of teaching experience. She graduated her
Deep Learning, Wireless Sensor Networks, Big Data. Member of Editorial B.E. from J.N.N Institute of Engineering (Anna
Board for International Journal of Computer Applications, USA. University), M.E. from Sri Sairam Engineering
College (Anna University) awarded 5th Rank in all
among the Universities and pursuing Ph.D. from
Saveetha School of Engineering. She published 5 research papers in Indexed
International Journals. Her research interests are IOT, Big Data Analytics and
Cyber Security.

ISSN: 2395-0455 ©EverScience Publications 156