Content uploaded by Madhusanka Liyanage
Author content
All content in this area was uploaded by Madhusanka Liyanage on Mar 02, 2022
Content may be subject to copyright.
1
Realizing Contact-less Applications with
Multi-Access Edge Computing
Pasika Ranaweera, Chamitha de Alwis, Anca D. Jurcut, and Madhusanka Liyanage
Abstract—The entire world progression has ceased with the
unexpected outbreak of the COVID-19 pandemic, and urges
the requirement for contact-less and autonomous services and
applications. Realizing these predominantly Internet of Things
(IoT) based applications demands a holistic pervasive computing
infrastructure. In this paper, we conduct a survey to determine
the possible pervasive approaches for utilizing the Multi-Access
Edge Computing (MEC) infrastructure in realizing the require-
ments of emerging IoT applications. We have formalized specific
architectural layouts for the considered IoT applications, while
specifying network-level requirements to realize such approaches;
and conducted a simulation to test the feasibility of proposed
MEC approaches.
I. INTRODUCTION
The COVID-19 pandemic has led us to realize that pre-
vailing knowledge, methodologies, technologies, or facilities
are not adequate to mitigate the impact it caused. Though
there are few promising vaccines been introduced by reputable
pharmaceutical institutions, their effectiveness is still in a
questionable state in light of the new mutated strains of
the pathogen. The level of contagiousness characterized by
the stealthy transmission of COVID-19 has led to develop
novel means of human interaction methods to mitigate its
spread. For this, Internet of Things (IoT) concept plays a
key role in accomplishing such strategies. Already established
IoT devices, protocols, and frameworks along with the stan-
dardization institutions are paving the most feasible digital
infrastructure to find solutions for the current dilemma. In
addition, the ’Smart City’ concept aligned with IoT guarantees
the coverage of all essential aspects of urban development for
such digital solutions. Though, a digital infrastructure alone
is not sufficient to meet the current demands. A pervasive
or ubiquitous computing capability is an intrinsic aspect;
where edge computing paradigms are capable of facilitating.
In the pandemics’ context, Internet of Medical Things (IoMT)
directives are the most essential to deploy medical services
featured with contact-less patient inspection, sample extrac-
tion, treatment, and monitoring methods. Infact, the digital
IoMT infrastructure that interconnects all the medical appli-
ances within a medical vicinity are improving the integration
capability of novel requirements for contact-less practices [1].
Pasika Ranaweera and Anca D. Jurcut are with School of Com-
puter Science, University College Dublin (UCD), Ireland. e-mails:
pasika.ranaweera@ucdconnect.ie, anca.jurcut@ucd.ie
Chamitha de Alwis is with the Department of Electrical and Elec-
tronic Engineering, University of Sri Jayewardeneprua, Sri Lanka, email:
chamitha@sjp.ac.lk
Madhusanka Liyanage is with the School of Computer Science, UCD,
Ireland and the Centre for Wireless Communications, University of Oulu,
Finland. e-mail:madhusanka@ucd.ie, madhusanka.liyanage@oulu.fi
It is obvious that accustoming into a contact-less and
confined daily routine that circumvents the tensity and anxiety
require the usage of technology from the consumer perspec-
tive. Such adaptability is not only limited to healthcare, but
for entertainment, occupation, trading, fashion, and education.
Eventhough the current technology offer remote access to
resources, online video conferencing facilities, online col-
laborative platforms, support for smart sensory and actuator
devices are sufficient for the time being, would not scale
with the demand for the digital services in the future. As
IoT based services are employing devices that are limited on
resources, their advancement is depended on the offloading
capability, where resource consuming processing phase of their
operation is performed in an outsourced manner at a third-
party infrastructure. Further, most emerging services are not
designed to operate stand-alone; require real-time and all-
time connectivity to their respective servers or control stations
to maintain their pervasive features along with monitoring
capability. Thus, the most lacking aspect for contact-less
and pervasive digital solutions is the existing networking
infrastructure, where every emerging application or service
demands enormous bandwidths with extremely low latency
to maintain their operation online without interruption. Con-
versely, industries and critical infrastructure are heavily relying
on the emerging fifth generation (5G) related technological
advancements for enabling the contact-less operating services
for improving the current networking infrastructure.
Emerging directives of Augmented Reality (AR), Virtual
Reality (VR), Ultra Reliable Low Latency Communication
(URLLC), enhanced Mobile BroadBand (eMBB), and mas-
sive Machine Type Communication (mMTC) are essential to
contrive the digital infrastructure intrinsic to enable envisaged
use cases [2]. Though, service requisites demanded by these
novel concepts, specially the ubiquitous computing capability;
are inconceivable with the prevailing networking infrastructure
that complemented through cloud computing based services.
The emergent edge computing paradigm Multi-Access Edge
Computing (MEC) attributes the capability to overcome the
stated pitfalls of cloud computing. MEC platforms established
at mobile base stations are capable of provisioning dynamic
services at a proximate locality. With these decentralized
server placements, location and context awareness are possible
while ultra-low-latency, higher bandwidth, security, and pri-
vacy aspects are enhanced to cater novel service specifications.
The edge infrastructure of the MEC is structured with
virtualized entities of Mobile Edge Hosts (MEHs) that operate
as Virtual Machines (VMs), while Mobile Edge Platform
Manager (MEPM) and Virtualization Infrastructure Manager
2
(VIM) are acting as the orchestrator and the hypervisor for the
MEC platform as explicated in [3]. Further, cloud deployments
are enduring unintended latency due to its centralized nature
with globally dispersed server placements. This service model
is not optimal for Internet of Things (IoT) based services
that attribute higher heterogeneity. Thus, an alternative service
model is intrinsic to improve the pragmatic feasibility of the
nascent technologies.
In this article, we investigate the possibility for launch-
ing various IoT technologies, utilizing the edge computing
capabilities of the MEC paradigm through a state-of-the-art
review of the existing work; that serves as a position paper
for contactless pervasive strategies. In fact, findings of this
research aids to realize the use cases specified in Section
II. Section III states various driving technologies enabled by
MEC, while Section IV discusses the realization of the stated
use cases in the context of MEC standardization. Validating the
MEC capability in contrast to cloud computing deployments
is presented in Section V. Section VI concludes the article.
II. EMERGING IOT CONTACT-LE SS USE CASES
Smart City paradigm and IoT play a vital role in enabling
new services of contact-less nature. Though, the coverage of
Smart City is wider, and the context of this paper narrows
to contact-less possibilities to mitigate the concerns with the
pandemic. This section explains some of such important IoT
based cases that falls under the umbrella of Smart City. The
selected use cases are direct contributors for overcoming the
issues of COVID-19, through aiding to develop effective and
efficient contact-less practices.
A. Smart Hospitals
The hospitals and other health care facilities are the most
lacking and unprepared vicinity’s to face this pandemic. Means
of implementing contact-less practices in general can be
achieved with social distancing and circumspect handling of
materials or items. Though, transforming patient inspection,
consulting, laboratory sample collection, conducting experi-
ments, surgeries, and patient caring activities to be contact-
less is proving to be arduous with existing technologies, and
require IoMT directives to succeed. Robotics are engaged to
revitalize these requirements by means of clinical care, logis-
tics, and reconnaissance. Utilization of AR has been envisaged
for Robot Assisted Surgeries (RASs). Further, remote patient
caring can be accomplished with robotic engagement while
patients can be monitored remotely via visual and audible
means with embedding IoT devices in to their medical beds.
The vital medical data aggregated during monitoring, perform-
ing clinical tests, and laboratory experiments are managed
by information systems catered for big data requirements in
state-of-the-art hospitals. Reaching the stated advancements
within a hospital premises however, is infeasible at present in
a technological perspective [4].
Existing Challenges: The medical apparatus market is a
well-established, and exorbitant trade that involve plethora
of manufacturers with a highly scaled design range. Thus,
following challenges should be answered to pragmatically
launching this concept.
•Minimizing latency of the existing network infrastructure
to cater services as RASs.
•Compatibility and inter-operability issues associated
with the employed sensory, haptic, laboratory-measuring,
audio-visual, and robotic appliances.
•Robotic entanglements introduce the ’Robot Ethics’ con-
sensus to medical applications; where ethical legislation’s
should be clearly specified due to human subjects.
•Securing personal health information while preserving
privacy of patients.
•Restriction of the envisioned advancements due to exces-
sive regulations and standardization.
B. Smart Factories
With the COVID-19 outbreak, factories have faced signif-
icant operational challenges, resulting in temporarily shut-
ting down their operation due to strict social distancing
measures. Thus, contact-less and autonomous approaches are
quite essential to industries for seamless operation, and for
minimizing human involvement that offers a solution for
both social distancing and for work force minimization. Such
pervasive approaches are quite adoptable for Industrial IoT
(IIoT) through robotic intervention [5]. This can facilitate
full or partial operations of factories, especially the IoMT
supply chains to facilitate the continuous manufacturing and
dispersion of COVID-19 vaccines and devices. Sensors can
be used to monitor different conditions and the operation in
the Factory environment, report events, and trigger alarms in
a Industry 4.0 smart manufacturing environment. The Digital
Twin (DT) concept can be used to create a digital factory
environment to operate, monitor and optimize the production
chain while continuously collecting and processing data. AR
based collaborative tools can be used for remote maintenance
and operation of industrial equipment. The smart factory
environment can also be extended to have interactive interfaces
that facilitates business management, human resource and
infrastructure management, and security control.
Existing Challenges: Several challenges needs to be ad-
dressed in order to facilitate smart factories to combat the
COVID-19 pandemic.
•Connect all the machinery, sensors, and actuators to be
cyber-physical systems to an IoT network.
•Facilitate an extremely dense network of IoT devices.
•Minimize End-to-End (E2E) delays to about 1ms to sup-
port autonomous control, remote control and actuation.
•Provide extremely reliable wireless communication be-
tween IoT devices (over 99.999%).
•Support ultra low-delay and high-bandwidth access to
powerful cloud computing services analyze Small Data
and Big Data gathered through large number of IoT
sensors and devices.
•Ensure the security of sensitive data and business secrets
while maintaining the privacy of online connected work-
force.
3
C. Consumer Trading
Digitization of economic trading is identified as a pivotal
aspect for overcoming the restrictions imposed by COVID-
19 on consumer trading. For instance, grocery based e-
commerce has incremented up to 74% during the pandemic,
while platforms such as Amazon have increased their sales
in new product ranges including habitual commodities [6].
To facilitate this trend, existing trading platforms should
realize a transformation towards digital product presentation,
marketing, sales, supply chain management, and delivery. For
instance, virtual shopping malls should provide facilitated with
a simulated experience resembling the actual visualization,
tangibility, and mechanics of the product through VR/AR with
haptic feedback. This requires data collected through numer-
ous sensors to harness powerful edge computing capabili-
ties to provide a realistic shopping experience. Furthermore,
intelligent payment solutions, smart vending machines, and
smart shelves can also be implemented as possible solutions.
Unmanned Aerial Vehicles (UAVs) are quite popular with
emerging delivery systems as they are contact-less and reduces
the human workforce requirement. Similarly, other consumer
trading sectors such as fashion industry, architecture, real
estate, and restaurants can also benefit from using numerous
IoT devices and MEC to provide digitized consumer trading
experiences during the pandemic.
Existing Challenges:
Realizing a fully functional IoT based digitized consumer
trading platform requires addressing a number of challenges,
as listed below.
•Efficiently communicate multiple data streams obtained
through numerous IoT sensors in real-time.
•Store and process multiple data streams in real-time
to provide a comprehensive and realistic VR/AR based
shopping experience.
•Process large amounts of Small Data as well as Big Data
to obtain meaningful insights for user profiling, marketing
and providing a customized user experience.
•Offload computationally intensive tasks of resource con-
strained user devices.
•Ensure the security and privacy of consumer data and
information.
D. Contact Tracing
Apart from vaccination, the most effective ways of pre-
venting the spread of the SARS-CoV-2 virus and a rise
in disease are non-pharmaceutical methods such as social
distancing and isolation. The virus has an average incubation
period from infection to symptoms of approximately 5.2 days
and it is estimated that approximately 50% of infections are
asymptomatic [7], [8]. To contain this infection, it is vital to be
able to effectively and rapidly identify all social interactions
during the infectious period and enable proper contact tracing.
This is especially relevant to healthcare workers, who are
hypothesised to be at an increased risk of severe disease due to
higher viral load exposure and their potentially close contact
with infected individuals. Gadgets (i.e. IoT wearable, mobile
devices and dedicated IoT devices) based Contact tracing
enables authorities to identify individuals that have a high risk
of spreading the virus and to subsequently encourage them
to self isolate to limit the spread of the virus. In addition,
UAVs can be employed to conduct city-wide surveillance for
detecting any lockdown violations.
Existing Challenges:
The efficient deployment of IoT based contact tracing
requires to address few challenges as listed below.
•People want to keep their location and movement infor-
mation private without over seen by authorities
•The use cloud based system to store collected IoT loca-
tion data could raise privacy concerns
•Storing row contact tracing information and processing
of information at the cloud will increase the probability
of cyber attacks and operational cost.
E. Working from Home and Online Education
COVID-19 has imposed restrictions across over 185 coun-
tries. Businesses are asking employees to work from home,
while the schools and university are resorted to online learning.
The IoT plays a crucial role in connecting businesses and
employees and respectively, e-learning and students in recent
years. Most of the remote work and education requires internet
connected devices such as laptops, mobile phones, webcams
and microphones. According to UNESCO, over 1.57 billion
learners which represent about 89.4% of total registered learn-
ers are affected [9]. The pandemic opened a new opportunity
for institutions - most of the universes and institutions over the
world replaced the classroom teaching with e-learning [10].
All these sectors are seeking new methods and technologies
to seamlessly continue their usual operations as normal, while
working remotely and practicing e-learning. All these sudden
changes are over-stressing the network since the functionally
of these sectors are now depend mainly on both wired and
wireless communication networks to remain in operation [11].
Existing Challenges: Working from home and e-learning
are bringing new challenges that impact their feasibility and
quality including:
•A great increase in the use of online learning platforms,
social media platforms and XR applications (virtual labo-
ratories), causing a surge in traffic demand on the existing
networks.
•Due to the stay-at-home policy, there is also a sudden
shift in the traffic demand pattern from city centers to
the residential areas.
•Working from home policy has increased the risk expo-
sure of insider threats, since the individuals are a potential
weak factor in preserving security and privacy.
•Employees could boost incident rates through eager-
ness to prove their effectiveness working from home
by bypassing policy or operating under less restricted
parameters.
•New intelligent service to identify the speaker, live tran-
scription, better virtual whiteboards, background noise
and visual distractions need to be addressed for work and
education from home. New algorithms to improve these
capabilities will be important.
4
III. MEC BAS ED SE RVICES FOR IOT/IOMT
MEC can realize new services which can be integrated
with IoT/IoMT applications to mitigate their challenges and
improve the performance, specially in the pervasive context.
This section explains some of such important services related
to IoT/IoMT applications.
The requirements for each application discussed in this
section are specified in TABLE I.
A. Efficient Augmented Reality
AR is one of the prominent wearable based service used
in many IoT applications. For AR based processing, MEC
infrastructure can be leveraged to perform required heavy-duty
tasks that cannot be launched at typical IoT devices. A MEH
can be utilized to perform the AR function. It is possible to
implement main functions of computing, caching, and visual-
ization of the AR process as Mobile Edge Applications (ME
Apps) within a MEH [12]. Such ME Apps can be deployed
as light-weight virtualized entities. AR content within the
MEH process configured for AR cache can be classified as
object and data caching. The computing platform ME App
should conduct computing and graphical processing. Further,
visualization function is composed of feature classification,
video analysis, and tracking sub functions to formalize the AR
perception. Mobile Edge Platform (MEP) of this AR based
MEH should be pre-configured to orchestrate these entities
within the MEH.
B. Low Latency Communication
Cloud computing is capable of providing efficient data
processing capabilities by offloading computationally inten-
sive tasks from devices to powerful cloud servers. However,
cloud services can result in long communication delays and
high packet errors as cloud servers are not located in close
proximity to IoT devices. MEC brings cloud functionalities
such as providing efficient data processing capabilities by
offloading computationally intensive tasks from devices to
the edge of the network to execute computationally intensive
tasks of IoT devices. This is done through harnessing the
computational capabilities in multiple edge nodes located
within close proximity. For instance, MEC can cooperatively
use the computational resources of network nodes such as,
nearby access points, aggregation points of the core network,
micro data centers, cloudlets, and gateways. Furthermore, by
harnessing the high capacity mmWave access and a dense
deployment of MEHs, MEC is also able to connect IoT
devices with low latency access to facilitate delay sensitive
applications and services. Such services will be exposed to
the features of efficient and secure network level protocols
in addition to the enhancements attributed with mmWave
access channels. This will significantly shorten the delays
and jitters, minimize packet errors, speed up application re-
sponse, improve user experience and reduce potential network
congestion [13]. In addition, Terahertz (THz) communication
links can utilize the mostly unused 0.3 - 10 THz range to
strengthen the existing back-haul networks [14]; which can
reduce the latency drastically among the MEC servers, and
eventually for offloading and migration processes. Further,
the virtualization infrastructure in the MEC platform enables
dynamic automated responses that alleviate any processing
or computing delays that might encounter apart from the
network-level latency. Such softwarized strategies incorpo-
rating the optimization processes would further converge to
diminutive delays from the edge infrastructure and networking
[15], [16] point of views; that benefit the services immensely to
deploy their autonomous constructs. Applications of robotics,
and UAVs could benefit from this feature of 5G enabled MEC
for the use cases specified in Section II.
C. Micro Data Center and Wireless Big Data Analysis
MEC can offer new ways of deploying data centres to col-
lect IoT data. It can provide greater flexibility by embedding
MEC resources in giant data centres or clustering it with
multiple small and medium scale servers, which leads towards
the micro data centre concept. Using MEC resources, micro
data centres can be easily deployed within the base stations and
other telecom nodes. With the development of 5G and beyond,
MEC devices will be available ubiquitously to collect IoT
data with enhanced optimization of networking capabilities
[17]. Then, multiple MEC servers in closed proximity can
even form a cluster to scale up resources for these micro data
centres in heavy IoT environments. With IoT and IoMT, a wide
range of massive data is generated, collected, and stored in a
wireless mobile network. This data is called as wireless big
data. When more and more wireless IoT devices are connected,
the amount of wireless big data getting increase. It is necessary
to process such huge amount of IoT data efficiently and with a
minimum delay to enable novel 5G services. Since centralized
cloud approaches have the drawbacks of limited backhual,
high latency, lack of localization, and lack of privacy, MEC can
be used to efficiently process such wireless big data. Moreover,
this can be optimized by coupling with micro data centre
deployments.
D. Edge-AI and Cognitive Assistance
Since modern IoT networks are highly dynamic and diverse,
the edge networks should support Self-organizing networks
(SON) and Self-sustaining Networks (SSN) abilities. Further,
to improve the performance of the network and comput-
ing at the edge, various techniques are used to manage,
orchestrate and sustain the network infrastructure ranging
from topology management (edge site orchestration), to data
and service provisioning. Among such techniques, AI and
Machine Learning (ML) are considered as a key solution
for many challenges. The marriage of edge computing and
AI established a new research area known as “edge AI”
or “Edge Intelligence (EI)”, that is providing AI insight to
most of the widespread edge resources. EI is generating lot
of attention from both the industry and academia including
Google, Amazon, Microsoft, Intel, and IBM. Edge AI used for
wide range of IoT applications including live video analytics,
cognitive assistance, healthcare and Industrial Internet; and
they clearly demonstrate the advantages of edge computing
5
in paving the last mile of AI. In the current context, the
effectiveness of the vaccines can be tracked via Edge AI
means with patient records collected at MEC based micro-
data centres. Furthermore, EI as a service (EIaaS) started
to become a new promising paradigm which is encouraging
the development of new EI platforms with powerful edge AI
functionalities. While ML as a service (MLaaS) focuses on
selecting the proper server configuration and ML framework to
train the model in clouds in a cost-efficient manner; the EIaaS
is focusing on how to perform model training and inference
in resource-constrained and privacy-sensitive edge computing
environments [20].
E. Offloading
MEC infrastructure at the edge is accomplishing a year-
old lacking aspect in the IoT market for performing computa-
tionally intensive or high resource consuming (i.e. storage, or
memory) tasks within resource constrained minimalist devices
through offloading. This fact is severe in IoMT apparatus,
as most of such data should be aggregated and stored se-
curely within a trusted platform. Presumably, data extracted
by end devices are conveyed to the MEC edge platform for
either storage or processing purpose that utilizes the energy
consumption of such apparatus optimally. Processing based
offloading approaches are diverse as they can be varied from
simple sensory aggregated processing (healthcare wearable
devices, contact monitoring, or digital PCR) to actuating via
a control matrix (robots, or drones). Thus, the softwarized
architecture for offloading should be dynamic and configurable
in accordance to the processing and data ingress/ egress model
requisites. Therefore, it is evident that MEC or any other
edge computing paradigm is intrinsic for achieving offloading
functions. Each offloading task can be instigated as a User
Equipment (UE) App from the end device perspective, where
a MEH configured for provisioning the specialized offloading
service is launching the corresponding Mobile Edge (ME) App
within its virtual domain [3]. These ME Apps should be task
intensive, and resources should be allocated dynamically to
leverage the process with minimal spending. Since different
MEHs are configured to provision different offloading ser-
vices, the internal structure of the MEH can be designed and
configured to the service specifications. Transferring the con-
tent from the UE to the MEC edge is another challenge; that
should minimize the energy spending in terms of processing
and transmitting, and the bandwidth. Apart from the latency
that reflects the service quality, energy becomes the dominant
factor for offloading decision making, as most IoT devices
are scarce on resources. The amount of content, content wise
encapsulation, security and reliability measures embedded into
the content that are to be offloaded will determine the energy
spending; in fact, specify the energy constrains. Thus, it is
important to model such offloading schemes in accordance
to the well-known models such as Markov or Lyapunov
approaches to alleviate the discrepancies while in transit [21],
[22]. Further, privacy is a major concern with the offloaded
content, specially in the medical or healthcare context. Privacy
preserving techniques such as Blockchain, improved privacy-
awareness through machine learning, anonymity management,
or physical layer approaches are ideal for mitigating such
issues associated with offloading [23], [24].
TABLE I: Requirements on the networking perspective for realization of the proposed use cases [12], [18], [19]
MEC based application or service Requirements
End-
to-End
Latency
Jitter Packet
Loss
Rate
Bandwidth/
Bit rate
Availability Max.
# of
UEs
Security
Level
Augmented Reality
Stereoscopic 4K (3840x2160 pixels) 120 fps real-time
video stream
<1 ms <10µs 10−3>24 Gbps >99.99999% 1 MEDIUM
4K 120 fps real-time video stream with lossless compres-
sion
<50 ms <2 ms 10−3>12 Gbps >99.99999% 10 MEDIUM
Robot Telemetry / Motion control data stream <2 ms <2 ms 10−4>16 Mbps >99.999999% 10 HIGH
Haptic Feedback data stream <2 ms <2 ms 10−4>16 Mbps >99.999999% 1 MEDIUM
Low-Latency Communication
URLLC channel <0.5 ms <100µs 10−4>2 Mbps >99.99% 1 LOW
Offloading
Less critical offloading channel <20 ms <2 ms 10−4>2 Mbps >99.99% 1 MEDIUM
Command and control data stream <2 ms <2 ms 10−4>16 Mbps >99.999999% 1 HIGH
Sensor fusion channel data stream <2 ms <2 ms 10−4>16 Mbps >99.999999% 10 MEDIUM
Caching
High quality audio stream <100 ms <30 ms 10−2>128 Kbps >99.99% 20 MEDIUM
Stereoscopic 4K 60 fps color coded real time video mon-
itoring
<250 ms <30 ms 10−3>2 Gbps >99.99% 20 MEDIUM
Edge AI
Sensor fusion/ telemetry edge-ingress channel <2 ms <2 ms 10−4>16 Mbps >99.999999% 10 MEDIUM
edge-egress decisive URLLC channel <0.5 ms <100µs 10−4>1 Mbps >99.99% 1 HIGH
Micro Data Centres
Data feeding/ extraction channels <1 s <30 ms 10−3>1 Gbps >99.99% 20 MEDIUM
Channel for priority services <250 ms <10 ms 10−4>2 Gbps >99.99999% 5 MEDIUM
Capabilities of Pre-5G Capabilities of General 5G Capabilities of MEC enabled 5G
6
F. Caching, Dynamic content delivery, Video Streaming and
Analysis
The issue of caching in nascent applications of dynamic
nature is a profound predicament solved by the edge comput-
ing paradigms. MEC infrastructure offers a proximate caching
placement to IoT based services that lack resources at the
device end. In addition, caching in Device-to-Device (D2D)
connections in IoT networks can be performed efficiently using
a clustered architecture coordinated via the MEC edge. In
this case, the popularity of the content can be considered
for maximizing the hit probability [25]. Furthermore, caching
policies based on content appearing frequency, user preference,
Q-learning, and cooperative or non-cooperative aspects can be
augmented to optimize the caching of dynamic audio/ visual
content. Also, depending on the IoT application, caches can
be established as MEHs, or caches inside a MEH where
policy management can be conducted by MEP. However,
streaming protocols should attribute dynamic accessibility in
the networking infrastructure within the MEC edge platform
to provide a smooth user experience.
G. MEC based Security and Privacy
Ensuring security and privacy requirements is critical for
pragmatic IoT based services. Though, resource scarcity ex-
hibited in IoT devices are restricting the application of tamper-
proof security measures. Thus, most state-of-the-art secu-
rity directives are envisaging to employ light-weight security
means (elliptic curve, lattice-based, attribute-based, or physical
unclonable function based) for such end devices and their
protocols. Though, such approaches have their limits with the
considered spectrum of possible threats. In addition, cyber-
threat landscape has been expanded to the attack vectors
of injecting or instilling malicious content/ agents to the
preceding IoT networks, where mitigation is improbable in a
stand-alone or isolated context. Thus, MEC edge environment
offers a unique opportunity to perceive the holistic security
awareness in the considered domain. Security as a Service
(SECaaS) is an innovative concept that caters to the MEC
based service model as presented in [26].
IV. REALIZING IOT US E CAS ES W IT H MEC
It is important to study about, how the MEC architecture
is going to accommodate the diversity required by the use
cases specified in Section II. As shown in Fig. 1, despite the
overall architectural components of the MEC infrastructure
being solid, internal construct of the MEHs and their function
should be amended in accordance to the specifications of the
use case. As most novel services are inclusive of interfaces
that enable rapid online access, localization would further
improve the serving and processing times, that can reach the
level of real-time. Though the novel applications are specifying
more functions or processes for pre-processing, tagging, and
security, localization offered by the MEC proximate infrastruc-
ture is capable of achieving real-time responsiveness for the
specified use cases with 5G radio access technologies. This
section explicates how each use case would be realized from
the technical perspective.
A. MEC for Smart Hospitals
As Fig. 1 - Adepicts, deploying a MEC based In-building
Base Station (IBS) within the vicinity of the hospital improves
the odds in launching the envisioned applications in sub-
section II-A [19]. MEHs can be configured to cater diverse
medical applications leveraging their dynamic softwarized
infrastructure. For IoMT-data aggregation scenarios, MEH
storage resources can be utilized for data storing; can be
formed as a cluster of MEHs for scalability extending to
external MEC domains; can be configured with rapid data
retrieval mechanisms leveraging the autonomous operation.
The external connectivity towards the backhaul that interface
the cloud systems is pertinent for such MEHs. This is required
to retrieve intrinsic medical history, or latest medical guide-
lines for treating patients efficiently. MEHs for RASs can be
formulated with AR based processing platform (explicated in
subsection III-A) that embed a telemetry system for robotic
sensor fusion. In addition, feedback (haptic or otherwise), Si-
multaneous Localization and Mapping (SLAM), and trajectory
tracking of robots should be performed under this MEH. A
resembled approach can be employed for controlling medical
robots launched for automating daily routines of medical
personnel with AR.
Though MEC offers a IBS solution to launch envisaged
services within the hospital vicinity, it creates a single point
of failure with an edge infrastructure disruption, unless a
redundancy scenario is implemented leveraging a cloud plat-
form. Since the backhaul network engagement is minimal
compared to cloud scenarios, reliability and availability aspects
are reliant on the local access networks’ capability or the IBS.
Further, privacy of medical data and ethical practices are key
aspect that should be standardized with federated approaches.
B. MEC for Smart Factories
MEC brings cloud computational capabilities, connectivity
and storage to the edge of the network to facilitate flexible and
real-time processing for IoT and IIoT based smart factories,
as illustrated in Fig. 1 - B. MEC facilitates low latency and
location aware IoT applications due to the close proximity
to devices. This helps achieving low end-to-end delay, low
jitter, high reliability and high bandwidth that is required
for IIoT in smart factories. Furthermore, MEC can support
rapid configuration needed for on-demand and customized
manufacturing in IoT based smart factories by facilitating
dynamic changes in the production conditions. Also, MEC
aids to extend the capabilities of IIoT devices by providing
powerful computing, abundant storage and caching. These
capabilities can be harnessed in a smart factory environment
for autonomous and cloud manufacturing processes. MEC
can also provide the high computational power required for
analyzing large amounts of data collected from heterogeneous
machines and IoT sensors. Video footage on the processes
followed in industries can also be sent to the MEC where
learning algorithms can detect anomalies and make appropriate
decisions to implement proper countermeasures. [27].
Industrial factories are vicinity’s that embed massive amount
of IoT based devices, and the scalability is a key factor in
7
Fig. 1: MEC based IoT Use Cases for Forming Contact-less Strategies
terms of the networking perspective. With enormous amount
of interfaces (wireless or otherwise) that had to be maintained
between the devices and the MEC platform, congestion within
the local area networks is quite eminent. Thus, dynamic and
priority based channel allocation schemes are required to
improve the reliability of such networks.
C. MEC for Consumer Trading
As more individuals turn to IoT as a resource, its use in
consumer and commodity trading will significantly increase,
which will positively impact the business’ growth based on
innovation and new uses of technology. Businesses will exploit
the advantages of using IoT to meet client expectations and
improve the company’s goals for the future.
The MEC based trading infrastructure can be categorized
into product presentation and delivery initiatives as illustrated
in Fig. 1 - C. The VR based product presentation initiatives
will be most preferable in the future. The MEHs configured
for VR processing will be linked to the cooperate clouds that
include the VR specification model of the product. Unlike
AR, VR do not require a computing platform to process
the detection and tracking of realistic perception; though,
8
TABLE II: Significance/Importance of MEC services for potential IoT Applications and their implementation challenges
MEC based Applications and Services Implementation challenges
Potential IoT Applications
Augmented Reality
Low Latency
Communication
Micro Data Centres
and Wireless Big Data
Analysis
Edge-AI and
Cognitive Assistance
Offloading
Caching, Dynamic
content delivery, Video
Streaming and
Analysis
MEC based Security
and Privacy
Security & Privacy
Scalability
Edge Resources
Network Resources
Mobility
Compatibility
Smart Hospitals
Robot assisted surgery H H M H H H M M L H H L L
Remote robotic monitoring and patient care M M M H H H M M M H H L M
Patient vital data aggregation and management L M H M M M H H H M M M H
Smart Factories
Production via autonomous robotic operations L H H H M M L L H M H M H
Digital twin H M H H M H M M M H H L M
Smart sensory data aggregation and management L H H M H L M M H M M M H
Efficient logistic and delivery management L M H H M L L L H M M M M
Consumer Trading
Contact-less/remote product presentation strategies H M H L M H L L M H H H H
Autonomous intelligent interactive agents - Chatbots L M L H M M L L H M M M H
Autonomous optimized contact-less delivery lodging L H H H M M M M H M H H H
Contact Tracing
Distributed data management L L H H M L H H H M M H M
Efficient contact tracing and infectious forecasting L M H H M M H H H M M H M
Contact-less quarantine monitoring L H H H H M H H H H H H M
Working From Home
WFH with virtual environment/ laboratory H H H H M H M M H H H M H
Advanced online caching and presenting H M H H M H M M H H H M H
Low Importance Medium Importance High Importance
requires caching and 3D visualization rendering features in
addition to position and orientation tracking. Each ME App is
subscribed by a distinct product range under different vendors.
Launching AI based chat-bots are less resource intensive in
contrast to VR deployments. The ML based decision engine
that perform intent classification and entity recognition can be
placed inside a MEH linked to external libraries. The message
or speech processing components are acting as the response
agents, operating in line with global databases. ME Apps are
configured by each vendor based on their requirements and
themes to interface the consumers. Product delivery services
are acting separately, as they are typically outsourced to
delivery agents by the prosumers or vendors. Non human
engaged delivery can be implemented as a generic logistic
system, where all the operations are autonomous; including
the deliveries managed through UAVs. UAV based delivery
services can deploy and control their drones via the MEC
platform formed as an offloading model as in [28]. It is obvious
that MEC offers an efficient interface for UAV applications in
contrast to cloud-native scenarios due to its proximate server
environment and rapid response capability.
Online trading is not a priority service in its nature, but
the security level of the transactions along with reliability and
availability should be high due to the monetary content that are
conveyed through the channels. With the UAV based delivery
however, rapid delivery of the content through the network
should be prioritized. Unless, the UAV should attribute the
self-navigation capability within itself.
D. MEC for Contact Tracing
MEC can be efficiently utilized for gadget/IoT based
COVID-19 contact tracing, as depicted in Fig. 1- D. To
prevent the privacy issues in contact tracing, MEC can be
used to implement a distributed data management approach.
With the possibility to deploy localized micro data center,
collected IoT data can stored locally. Moreover, wireless big
data processing at the edge and edge AI techniques can be used
to process such data in a localized manner. In such a way, MEC
can prevent the transmission and storage raw user and IoT data
in the cloud and also sharing them with authorities. Only the
processed information (i.e privacy sensitive markers has been
removed) at the edge will be shared with the authorities. This
will also improve the overall scalability of the contact tracing
mechanism. Moreover, MEC and edge AI can be used for
efficient Command and Control (C&C) for UAVs which play
a vital role in monitoring quarantine violations.
The details and status of contact tracing should be accessible
for relevant authorities at all times. Thus, availability of
9
these servers or storage instances within the MEC platform
should be ensured with proper security levels to prevent any
undue service denying threats. Reliable and hierarchical data
access through dynamic rapid querying that fits this model
should be proposed to maintain the data flow with appropriate
privacy and anonymity measures. Further, data access should
be deployed either with federated data engines running on
MEC or replication to the cloud native data clusters to make
the querying efficient globally and locally.
E. MEC for WFH and Online Education
It is paramount to carefully address (1) the security of the
IoT devices and of the networks used, and (2) the security and
privacy aspects of users who access various online collabo-
rative working/learning platforms and apps. Virtual machine
introspection services can run using the MEC infrastructure
and monitor the behavior of WFH and online education based
apps to detect any attack that intends to compromise the
virtualized entities. AI based intrusion detection systems can
harness the MEC capabilities to monitor the network behavior
and data transmitted in the network. MEC based trusted
platform manager techniques can handle secure authentica-
tion and also verify the integrity of software and executable
programs through validation of operational statistics such
as firmware, operating system kernel, etc. MEHs are also
capable of deploying context aware security mechanisms to
detect anomalous behavior of apps [3]. In addition, XR based
working/learning platforms can harness the MEC capabilities
to process high volumes of data in real-time to facilitate
virtual meetings, virtual classrooms, virtual laboratories, etc.
Most WFH connections are established through Virtual Private
Networks (VPNs). MEC networking infrastructure can be
leveraged to launch Virtual Private LAN Service (VPLS) based
off site remote connections that can be more secure and faster
than individual VPN connections. Usage of MEC for WFH
and online educational activities are illustrated in Fig. 1- E.
Usage of VPN or VPLS level access encapsulation can
overburden the network drastically with the increasing demand
for WFH apps. Managing traffic is becoming a pragmatic
dilemma for the MEC access network with such a demand
for online access. With MEC deployments, the links between
MEC platforms are becoming more prone for traffic. Thus,
managing traffic in these links becomes an issue for reliable
delivery. Feeding online meetings or online lectures with HD
level audio and video streams is going to further burden
the network infrastructure. Therefore, methods and techniques
should be developed for managing this accumulated traffic
introduced with novel demand for online delivery. TABLE II
summarizes the significance of each MEC based service for
IoT use cases.
V. FEASIBILITY EVAL UATION
Proposed use cases in this paper have two different de-
ployment options. Its either through cloud computing or edge
computing. In order to determine the feasibility of the pro-
posed use cases to be deployed with MEC, an evaluation was
conducted via simulating the performance to the following
criteria. Each criteria was selected based on the essential
services required by the stated use cases that relies on MEC
for their pragmatic deployment. In fact, this section validates
the pervasiveness of the MEC paradigm for the stated use
cases. The intention of this evaluation was to study the limits
of each use case from the deployment perspective, and to
determine the better performing scenario. The simulations
for the stated criteria targeting the proposed MEC based
edge computing deployments are running in contrast to the
prevailing cloud computing capabilities. We assume that the
same access network model is applied for both MEC and cloud
scenarios, where the dynamics on mobility is similar for both
instances.
•Criterion 1: General AR applications - for less resource
intensive AR uses such as virtual education and trading
initiatives. Assuming 30 ms, 107, 5 Gbps of processing
delay, CPU cycles, and data rate for achieving a single
task.
•Criterion 2: Critical AR applications - formed for RAS
or AR controlled IIoT deployments. The corresponding
assumed parameters are 5 ms, 109, 20 Gbps, and the
controlling station located 50 km from the operating
premises.
•Criterion 3: Audio/ Video (AV) streaming and caching
applications - suited for video caching or eMBB type
services with parameters 20 ms, 5×106, and 2 Gbps.
•Criterion 4: UAV autonomous C&C applications -
assuming the UAV is operating 2.5 km away from the
MEC locality where the parameters are 10 ms, 106, and
0.5 Gbps.
The parameters for above 4 criteria were selected in ac-
cordance to the limits specified under TABLE I. Each criteria
was assumed to be operating following an offloading scenario,
where UE content is offloaded either to the cloud or MEC
platforms for processing, and the corresponding outcome is
delivered to the UE at the end. The simulation environment
was formulated in accordance to [19]; where the parameters
are specified below.
Fig. 2: Impact of Distance to the Core Network for E2E
Latency
□Computing capacity of the MEC - 5×1010 CPU cycles
□Computing capacity of the cloud - 1011 CPU cycles
□Bandwidth of the MEC access connectivity 50 Gbps
□Bandwidth of the backhaul network - 10 Gbps
10
Fig. 3: Impact of Scalability for Operational Delay
Fig. 4: Impact of Scalability for Availability
□Displacement of the eNB from the MEC - 1 km
□Delays in MEC and cloud access are 0.25 ms and 1 ms
□Distance dependent core network delay - 0.05 ms/km
E2E latency impacted by the distance to the cloud infras-
tructure, Operation Delay impacted by the scalability of the
UEs, and Availability of the each criterion on the scalability
aspects were simulated and the results are illustrated in Fig. 2,
Fig. 3, and Fig. 4 respectively. The availability was computed
in accordance to [29], where the respective mean down times
were estimated referring [30], and implying traffic conges-
tion instances to failures rates. It is observable that MEC
deployments have a lesser E2E delay in contrast to cloud
scenarios; while operational delay is accumulating with the
increasing number of UEs. This is mainly due to the proximate
locality of the MEC deployment. The processing of HD video
in criterion 3 for 1080p with 60 fps requires up to 5 Mbps
bit rate, and a significant processing delay is endured even
for cloud-native scenario in contrast to criterion 4, where the
processing is limited to telemetry data. Thus, E2E delay in
criterion 3 is higher than criterion 4. Fig. 4 indicates that
availability in MEC scenario is significantly higher and within
the limits of the 5G network specifications. However, the high
resource requirement of the specified criteria are limiting the
number of UEs that can operate simultaneously. This proves
that MEC is capable of enabling the IoT applications to realize
the identified key COVID-19 related IoT use cases. In addition,
implementation challenges for various technical aspects are
tabulated in TABLE III.
VI. CONCLUSION
This paper presents how MEC can be used to realize key
IoMT based use-cases as a pervasive approach that enhances
the computing infrastructure. e-Health use cases ranging from
smart hospitals, smart factories, consumer trading, contact
tracing, working from home, and online education have been
discussed, highlighting the challenges and requirements of
each use case. Then, MEC based IoT contact-less services
such as AR, low latency communication, micro data center
and wireless big data analysis, edge-AI and cognitive assis-
tance, offloading, caching, dynamic content delivery, video
streaming and analysis, and MEC based Security and Privacy
are presented elaborating how MEC can realize the identified
IoT use-cases. Simulation results clearly indicate that MEC
based solutions are more suitable for realizing the stated use
cases compared to cloud computing in terms of end-to-end
latency, scalability, and availability. Insights drawn from this
article are expected to encourage both telecommunication and
internet service providers to launch MEC in a global context
to extend the computing infrastructure to the mobile edge, and
overcome the limitations of current technologies that lacks the
pervasiveness.
REFERENCES
[1] K. K. Karmakar, V. Varadharajan, U. Tupakula, S. Nepal, and C. Thapa,
“Towards a security enhanced virtualised network infrastructure for
internet of medical things (iomt),” in 2020 6th IEEE Conference on
Network Softwarization (NetSoft). IEEE, 2020, pp. 257–261.
[2] P. Ranaweera, A. D. Jurcut, and M. Liyanage, “Survey on multi-access
edge computing security and privacy,” IEEE Communications Surveys
& Tutorials, vol. 23, no. 2, pp. 1078–1124, 2021.
[3] ——, “Realizing Multi-Access Edge Computing Feasibility: Se-
curity Perspective,” in 2019 IEEE Conference on Standards for
Communications and Networking (CSCN). IEEE, 2019, pp. 1–7.
[4] G.-Z. Yang, B. J. Nelson, R. R. Murphy, H. Choset, H. Christensen,
S. H. Collins, P. Dario, K. Goldberg, K. Ikuta, N. Jacobstein et al.,
“Combating covid-19—the role of robotics in managing public health
and infectious diseases,” 2020.
[5] Y. Siriwardhana, C. De Alwis, G. Gür, M. Ylianttila, and M. Liyanage,
“The fight against the covid-19 pandemic with 5g technologies,” IEEE
Engineering Management Review, vol. 48, no. 3, pp. 72–84, 2020.
[6] A. Bhatti, H. Akram, H. M. Basit, A. U. Khan, S. M. Raza, and M. B.
Naqvi, “E-commerce trends during covid-19 pandemic,” International
Journal of Future Generation Communication and Networking, vol. 13,
no. 2, pp. 1449–1452, 2020.
[7] Q. Ye, B. Wang, J. Mao, J. Fu, S. Shang, Q. Shu, and T. Zhang,
“Epidemiological analysis of covid-19 and practical experience from
china,” Journal of medical virology, vol. 92, no. 7, pp. 755–769, 2020.
[8] M. Yanes-Lane, N. Winters, F. Fregonese, M. Bastos, S. Perlman-Arrow,
J. R. Campbell, and D. Menzies, “Proportion of asymptomatic infection
among covid-19 positive persons and their transmission potential: A
systematic review and meta-analysis,” PloS one, vol. 15, no. 11, p.
e0241536, 2020.
[9] K. Markelova, “Mapping the world: Education: An unprecedented
crisis,” The UNESCO Courier, vol. 2020, no. 3, pp. 54–57, 2020.
[10] M. S. Elsayed, N. A. Le-Khac, and A. D. Jurcut, “Dealing With COVID-
19 Network Traffic Spikes,” IEEE Security & Privacy, vol. 19, no. 1,
2020.
[11] A. I. Abubakar, K. G. Omeke, M. Öztürk, S. Hussain, and M. A. Imran,
“The role of artificial intelligence driven 5G networks in COVID-19
outbreak: opportunities, challenges, and future outlook,” Frontiers in
Communications and Networks, vol. 1, 2020.
[12] J. Ren, Y. He, G. Huang, G. Yu, Y. Cai, and Z. Zhang, “An Edge-
computing based Architecture for Mobile Augmented reality,” IEEE
Network, vol. 33, no. 4, pp. 162–169, 2019.
11
TABLE III: Implementation Challenges of MEC based IoT Applications
Technical
Aspect
Issue/ Challenge/
Threat
Description Possible Solutions
Security Social engineering
threats on IoT users
COVID-19 pandemic has forced the users with lesser aware-
ness/literacy on online practices to engage in remote work
via IoT devices. This fact has been leveraged by cyber-
criminals to launch phishing type attacks to attain confidential
information.
Awareness is the key for avoidance. The institutions
should educate consumers regarding possible threats,
while preventive mechanisms such as spam filters, anti-
phishing add-ons, or firewalls should be pre-installed.
Malware intrusions Circumstances with the pandemic are used to tempt online
users into download and execute spyware, adware, or ran-
somware type agents for unauthorized access. These malware
can compromise the MEC edge components.
Advance and adaptive intrusion detection and prevention
systems are required while enhancing the security at
Internet access points within organizations.
Untrusted service
outsourcing
WFH initiatives have forced the employees to use third-
party web-based streaming, collaborative, educational, and
file sharing tools exclusive to the company network domains.
Their unverified security and limited functionalities are com-
promising the trade secrets.
Strengthen the remote access procedures of the institutions
and employ secure tunneling approaches for cooperate
communications and collaborations.
Denial of Service
(DoS) and Jamming
type threats
The excessive use of online services are inviting adversaries
to launch DoS attacks at channels connecting the IoT device
layer and the edge. These will impede the services while
jamming on UAV type devices will change its trajectory.
Detecting the illegitimate traffic via a trusted platform
manager would allow mitigation at routing level. Further,
anti-jamming techniques to secure UAVs.
Privacy Contact tracing The unavoidable tracing of the contacts of the infected is
exposing the private information including their whereabouts.
This is a complete violation of EU GDPR legislation’s,
and the process is neglecting the personal dissent in some
countries.
Information of the contacts can be restricted to localized
domain while AI, and blockchain like technology can be
integrated to secure the private data of individuals from
the authorities, where only a social security number of the
contact will be sent to them.
News, printed, digi-
tal, and social media
Governments used fear cultivated from media as a mean of
restricting the movements of citizens in the early days of the
pandemic. Such practices have violated the privacy rights of
exposed personnel and institutions.
Concrete legislation’s should be formed by the authorities,
specific to pandemic situations for media usage.
Health data
extracted from
IoMT wearables
The vitals of the patients are extremely private information
that are extracted from wearables by medical services. With
MEC, IoMT services are provisioned with a localized pro-
cessing infrastructure, though privacy violations are immi-
nent.
Blockchain can be employed to contrive an anonymous
data network where only a representative value is con-
veyed to the IoMT servers, not the actual value.
Edge and
Network
Resources
with
Scalability
Wireless access
channels
Increasing IoT and IoMT devices are exceeding the existing
spectrum allocations (or bandwidth).
Utilizing more frequency bands as in mmWave technology
and advance spectrum re-using strategies.
Traffic bottlenecks Rapid accumulation of all sorts of online traffic is creating
bottlenecks in certain access points of the network. Such
points are affecting the entire network performance towards
the edge.
Delegating traffic to various mediums optimally, so that
the network performance is not reliant on a single medium
such as RF.
MEC resource de-
pletion
Increasing number of UEs will eventually deplete the re-
sources of the edge infrastructure. This will terminate in-
proximity un-prioritized IoT services and reject the new
requests.
Establishing a migrating strategy that utilizes the closest
resourceful MEC.
Mobility Switching MEC
edges
Highly mobile device will keep sifting from one MEC BS to
another rapidly.
An efficient and fast migration strategy is required.
Coverage of IoMT
wearables
Wearables embedded to the human body do not possess a
direct transmission capability to a network hub, and relies
on short range technologies such as Bluetooth. With a highly
mobile device, these technologies become obsolete for com-
munication.
Developing a re-transmission protocol for mobility de-
tected situations.
Energy Depletion It is obvious that mobility is degrading the energy of IoT
devices, specially for location based services.
Attaching energy harvesting systems to the IoT devices.
Compatibility IoMT hardware
integration
compatibility
Due to heterogeneity exhibited with various IoT manufactur-
ers, integrating IoT devices to the medical premises systems
become quite questionable.
Policies, and protocols should be established for integra-
tion for IoMT devices while universal connection options
should exist on each IoT device.
Integration issues
with the edge
At software level, with different IoT based operating systems
are attempting to connect to the edge, there will be obvious
compatibility issues.
Integration policies and protocols should be standardized
by the authorities such as ETSI.
12
[13] M. Merluzzi, P. Di Lorenzo, S. Barbarossa, and V. Frascolla, “Dynamic
computation offloading in multi-access edge computing via ultra-reliable
and low-latency communications,” IEEE Transactions on Signal and
Information Processing over Networks, vol. 6, pp. 342–356, 2020.
[14] L. A. Grieco, G. Boggia, G. Piro, Y. Jararweh, and C. Campolo, “Ad-
hoc, mobile, and wireless networks.”
[15] A. M. Somarin, Y. Alaei, M. R. Tahernezhad, A. Mohajer, and M. Barari,
“An efficient routing protocol for discovering the optimum path in
mobile ad hoc networks,” Indian Journal of Science and Technology,
vol. 8, no. S8, pp. 450–455, 2015.
[16] A. Mohajer, M. Bavaghar, and H. Farrokhi, “Reliability and mobil-
ity load balancing in next generation self-organized networks: Using
stochastic learning automata,” Wireless Personal Communications, vol.
114, no. 3, pp. 2389–2415, 2020.
[17] A. Mohajer, M. Barari, and H. Zarrabi, “Big data-based self optimization
networking in multi carrier mobile networks,” Bulletin de la Société
Royale des Sciences de Liège, vol. 85, pp. 392–408, 2016.
[18] 3GPP, “Study on Communication Services for Critical Medical
Applications,” 3GPP Technical Report 22.826, 2019, last accessed 13
July 2020. [Online]. Available: https://portal.3gpp.org/desktopmodules/
Specifications/SpecificationDetails.aspx?specificationId=3546
[19] P. S. Ranaweera, M. Liyanage, and A. D. Jurcut, “Novel MEC based
Approaches for Smart Hospitals to Combat COVID-19 Pandemic,” IEEE
Consumer Electronics Magazine, 2020.
[20] Z. Zhou, X. Chen, E. Li, L. Zeng, K. Luo, and J. Zhang, “Edge
Intelligence: Paving the Last Mile of Artificial Intelligence With Edge
Computing,” Proceedings of the IEEE, vol. 107, no. 8, pp. 1738–1762,
2019.
[21] G. Yang, L. Hou, X. He, D. He, S. Chan, and M. Guizani, “Offloading
time optimization via markov decision process in mobile-edge comput-
ing,” IEEE Internet of Things Journal, vol. 8, no. 4, pp. 2483–2493,
2020.
[22] Y. Hu, T. Cui, X. Huang, and Q. Chen, “Task offloading based
on lyapunov optimization for mec-assisted platooning,” in 2019 11th
International Conference on Wireless Communications and Signal
Processing (WCSP). IEEE, 2019, pp. 1–5.
[23] D. C. Nguyen, P. N. Pathirana, M. Ding, and A. Seneviratne, “Privacy-
preserved task offloading in mobile blockchain with deep reinforcement
learning,” IEEE Transactions on Network and Service Management,
vol. 17, no. 4, pp. 2536–2549, 2020.
[24] X. He, R. Jin, and H. Dai, “Physical-layer assisted privacy-preserving
offloading in mobile-edge computing,” in ICC 2019-2019 IEEE
International Conference on Communications (ICC). IEEE, 2019, pp.
1–6.
[25] S. Wang, X. Zhang, Y. Zhang, L. Wang, J. Yang, and W. Wang, “A
Survey on Mobile Edge Networks: Convergence of Computing, Caching
and Communications,” IEEE Access, vol. 5, pp. 6757–6779, 2017.
[26] P. Ranaweera, V. N. Imrith, M. Liyanag, and A. D. Jurcut, “Security as
a service platform leveraging multi-access edge computing infrastruc-
ture provisions,” in ICC 2020-2020 IEEE International Conference on
Communications (ICC). IEEE, 2020, pp. 1–6.
[27] B. Chen, J. Wan, A. Celesti, D. Li, H. Abbas, and Q. Zhang, “Edge com-
puting in iot-based manufacturing,” IEEE Communications Magazine,
vol. 56, no. 9, pp. 103–109, 2018.
[28] P. A. Apostolopoulos, G. Fragkos, E. E. Tsiropoulou, and S. Papavas-
siliou, “Data offloading in uav-assisted multi-access edge computing
systems under resource uncertainty,” IEEE Transactions on Mobile
Computing, 2021.
[29] L. Jereb, “Network reliability: models, measures and analysis,” in
Proceedings of the 6th IFIP workshop on performance modelling and
evaluation of ATM networks, Tutorial papers, Ilkley, UK, 1998, p. T02.
[30] Ponemon. (2021, January) Data center downtime at
the core and the edge: A survey of frequency,
duration and attitudes. [Online]. Available: https:
//www.vertiv.com/490a6d/globalassets/documents/reports/ponemon/
vertiv-ponemon-datacenterdowntimesurveyreport_321796_0.pdf
Pasika Ranaweera is currently pursuing his PhD studies in School of
Computer Science, University College Dublin, Ireland. His research directives
extend to the areas of light-weight security protocols, 5G and MEC integration
technologies, Privacy preservation techniques, MEC security, and IoT security.
URL: https://csintranet.ucd.ie/phd-student/pasika-sashmal-ranaweera
Chamitha de Alwis is currently working as a Senior Lecturer in the De-
partment of Electrical and Electronic Engineering, University of Sri Jayewar-
denepura, Sri Lanka. His research interests are 5G, 6G, IoT, Blockchain and
Network Security. URL: http://eng.sjp.ac.lk/eeeng/staff/dr-chamitha-de-alwis
Anca D. Jurcut is an Assistant Professor at School of Computer Science,
University College Dublin, Ireland. Her research interests focuses on network
and data security, security for internet of things (IoT), security protocols,
formal verification techniques and applications of blockchain technologies in
cybersecurity. URL: https://people.ucd.ie/anca.jurcut
Madhusanka Liyanage (S07, M16, SM20) is working as Assistant Professor/
Ad Astra Fellow at School of Computer Science, University College Dublin,
Ireland. He is also an adjunct professor at the University of Oulu, Finland.
His research interests are SDN, IoT, Block Chain, mobile and virtual network
security. URL: http://madhusanka.com
