VISVESVARAYA TECHNOLOGICAL UNIVERSITY

BELAGAVI

A
SEMINAR REPORT
On
“Multi-Access Edge Computing”

Submitted in partial fulfillment of the

SEMINAR
In
INFORMATION SCIENCE AND ENGINEERING
VII SEMESTER WEB TECHNOLOGY LABORATORY WITH MINI PROJECT
(15CSS86)

By
Jones G (1HK16IS116)
Under the guidance of

Prof. Devi Sivasankari

Department of Information Science and Engineering
HKBK College of Engineering, Bengaluru.
2019-2020

HKBK COLLEGE OF ENGINEERING

( Accredited by NAAC )
22/1, Nagawara, Bengaluru – 560045.
E-mail: [email protected], URL: www.hkbk.edu.in
BENGALURU-560045.

DEPARTMENT OF INFORMATION SCIENCE AND ENGINEERING

VISVESVARAYA TECHNOLOGICAL UNIVERSITY

SEMINAR REPORT

ON

“Multi-Access Edge Computing”

Submitted in partial fulfillment for

SEMINAR (15CSS86)

VIII SEMESTER

2019-2020
SUBMITTED BY:

Jones G (1HK15IS116)
HKBK COLLEGE OF ENGINEERING
22/1, Nagawara, Bengaluru – 560 045.
( Accredited by NAAC )

DEPARTMENT OF INFORMATION SCIENCE AND

ENGINEERING

CERTIFICATE

Certified that the Seminar work entitled “Multi-Access Edge Computing” is a bonafide
work carried out by Jones G (1HK16IS116), in partial fulfillment of SEMINAR (15CSS86),

VIII Semester, Information Science and Engineering of the Vishveshvaraya Technological

University, Belgaum during the year 2018-2019. It is certified that all corrections/suggestions
indicated for Internal Assessment have been incorporated in the Report deposited in the departmental
library. The seminar report has been approved as it satisfies the academic requirements in respect of
15CSS86 - Evaluation of Seminar Work Prescribed by the said Degree .

---------------------------- ----------------------------- -------------------------

Signature of Guide Signature of Co-ordinator Signature Of the HoD.
(Prof. Devi Sivasankari ) ( Prof. Asghar Pasha ) ( Dr A Syed Mustafa)
ACKNOWLEDGEMENT

I would like to place my regards and acknowledgement to all who helped in making this
seminar possible. I thank all those whose guidance served as a beacon of light and crowned
our efforts with success.

First of all I would take this opportunity to express our heartfelt gratitude to the management
committee - Chairman Mr. C. M. Ibrahim, Director Mr. C.M. Faiz Mohammed and the
Principal Dr. M S Bhagyashekar for all the infrastructures provided to complete the
SEMINAR in time.

I deeply indebted to Dr. Syed Mustafa, HOD, Information Science and Engineering for the
ineffable encouragement he provided for the successful completion of the project.

A special and an earnest word of thanks to the guide Prof. Devi Sivasankari for their
constant assistance, support, patience, endurance and constructive suggestions for the
betterment of the Seminar.

I’m extremely thankful to the teaching and non-teaching staff of the Department of
Information Science and Engineering for their valuable guidance and cooperation throughout
our dissertation.

I thank my parents for their support and guidance provided to us to finish the seminar well
ahead of time. I thank my friends who lent their support in every way possible to make sure
the mini project has been completed. Last, but not least I would like to thank God for giving
us this opportunity to do everything in the appropriate time to finish this seminar.

Name: Jones G .
USN: 1HK15IS116

iv
ABSTRACT

Multi-access edge computing (MEC), formerly mobile edge computing, is an ETSI-defined

network architecture concept that enables cloud computing capabilities and an IT service
environment at the edge of the cellular network and, more in general at the edge of any network.
The basic idea behind MEC is that by running applications and performing related processing
tasks closer to the cellular customer, network congestion is reduced and applications perform
better. MEC technology is designed to be implemented at the cellular base stations or other
edge nodes, and enables flexible and rapid deployment of new applications and services for
customers. The emergence of Multi-Access Edge Computing (MEC) technology aims at
extending cloud computing capabilities to the edge of the wireless access networks. MEC
provides real-time, high-bandwidth, low-latency access to radio network resources, allowing
operators to open their networks to a new ecosystem and value chain. Moreover, it will provide
a new insight to the design of future 5th Generation (5G) wireless systems. This technology
describes five key technologies, including Network Function Virtualization (NFV), Software
Defined Networking (SDN), Network Slicing, Information Centric Networking (ICN) and
Internet of Things (IoT), that intensify the widespread of MEC and its adoption.

v
Table of Contents
1. INTRODUCTION

1.1 DEFINITION........................................................................................................... 07

1.2 PRINCIPLE............................................................................................................. 07

1.3 BRIEF.......................................................................................................................07

2. NETWORK FUNCTION VIRTUALIZATION

2.1. NETWORK FUNCTION VIRTUALIZATION................................................. 09

3. SOFTWARE DEFINED NETWORKING

3.1. SOFTWARE DEFINED NETWORKING.........................................................10

4. INFORMATION CENTRIC NETWORKING

4.1 INFORMATION CENTRIC NETWORKING................................................... 11

5. NETWORK SLICING

5.1 NETWORK SLICING....................................................................................... 13

6. INTERNET OF THINGS

6.1 INTERNET OF THINGS.................................................................................. 14

7. CHALLENGES AND FUTURE WORK

7.2 NFV......................................................................................................................15

7.3 SDN...................................................................................................................... 16

7.4 ICN.......................................................................................................................17

7.5 NS......................................................................................................................... 18

7.6 IoT........................................................................................................................ 19

7.7 INTEGRATION PATH..................................................................................... 20

8. SNAPSHOTS................................................................................................................22

CONCLUSION.......................................................................................................... 25
REFERENCES....................................................................................................................26

vi
Introduction Multi-Access Edge Computing

Chapter 1
INTRODUCTION

1.1 Definition

Multi-access edge computing (MEC), formerly mobile edge computing, is an ETSI-

defined network architecture concept that enables cloud computing capabilities and an IT
service environment at the edge of the cellular network and, more in general at the edge of any
network.

1.2 Principle

The basic idea behind MEC is that by running applications and performing related
processing tasks closer to the cellular customer, network congestion is reduced and applications
perform better.

1.3 Brief

MEC technology is designed to be implemented at the cellular base stations or other

edge nodes, and enables flexible and rapid deployment of new applications and services for
customers. The emergence of Multi-Access Edge Computing (MEC) technology aims at
extending cloud computing capabilities to the edge of the wireless access networks. MEC
provides real-time, high-bandwidth, low-latency access to radio network resources, allowing
operators to open their networks to a new ecosystem and value chain. Moreover, it will provide
a new insight to the design of future 5th Generation (5G) wireless systems. This technology
describes five key technologies, including Network Function Virtualization (NFV), Software
Defined Networking (SDN), Network Slicing, Information Centric Networking (ICN) and
Internet of Things (IoT), that intensify the widespread of MEC and its adoption.
Firstly, since MEC servers are located close to the source of information, they have
direct access to the devices and local resources. This proximity is very useful to capture key
information for analytics and in Machine-to-Machine (M2M) communication scenarios.
Secondly, as the MEC servers are running close to end devices, they reduce the latency by
reacting faster to improve user experience and to minimize congestion in other parts of the
network. Thereby the user will experience ultra low latency and high bandwidth. Thirdly, when
MEC servers are part of the wireless access network, a local service can leverage low-level
signaling information to determine the location of each connected device. This property will
Dept. Of ISE, HKBKCE 7 2019-2020
Introduction Multi-Access Edge Computing

encourage use cases which require location based services. Finally, MEC servers can exploit
real-time network data such as radio conditions and network statistics to offer context related
services to the particular applications. This can also differentiate the mobile broadband
experience and be monetized. Thus, MEC plays a key role to achieve the vision of the fifth
generation (5G) wireless networks which are expected to reach 1 ms latency and high
bandwidth along with the quantified users’ quality of experience (QoE). In addition to MEC,
few other edge computing paradigms are also emerging such as Mobile Cloud Computing
(MCC), Fog computing, and Cloudlets. However, MEC is considered as better choice for 5G
mobile networks than other owing to its compatibility with cellular networks and heavy backup
by mobile standardization organizations such as ESTI. In spite of its huge potential, the
realization of MEC should be upheld by a multitude of underlying technologies. In this article,
we examine five of those key enabling technologies, including Network Function Virtualization
(NFV), Software Defined Networking (SDN), Information Centric Networking (ICN), Network
Slicing, and Internet of Things (IoT), and illustrate how to utilize them to accelerate the
adaption and development of MEC systems. Besides identifying open challenges, we also
pinpoint tangible integration paths and future directions for MEC.

1.4 Techniques

 Network Function Virtualization (NFV)

 Software Defined Networking (SDN)
 Information Centric Networking (ICN)
 Network Slicing
 Internet Of Things (IoT)

Dept. Of ISE, HKBKCE 8 2019-2020

Network Function Virtualization Multi-Access Edge Computing

Chapter 2
NETWORK FUNCTION VIRTUALIZATION
2.1 Network Function Virtualization
NFV proposes to utilize virtualization technologies to decouple physical network
equipment from the functions that run on them. Via this, different Virtual Network Functions
(VNFs) can be implemented in software running on one or more industry standard physical
servers. The VNFs can be relocated and instantiated at different physical network locations
without necessarily requiring the purchase and installation of new hardware.

Moreover, NFV is regarded as one of the key enablers for deployment of MEC in 5G
networks. Both NFV and MEC technologies can be used together in 5G mobile networks to
elevate computing capacity to meet the increased networking demands. Both MEC and NFV
share similar characteristics. For instance, MEC architecture is also based on a virtualized
platform which is similar to NFV, as depicted in Figure 1. Both technologies feature stackable
components and each has a virtualization layer. Accordingly to ESTI, it is beneficial to reuse
the infrastructure and infrastructure management of NFV to the largest extent possible, by
hosting both VNFs and MEC applications on the same platform to enhance the computing
experience. In addition, MEC can use the NFVI (NFV Infrastructure) as the virtualization
platform to run mobile edge applications alongside other VNFs. Therefore, MEC applications
also appear as VNFs in the NFV environment and parts of mobile edge orchestration can be
delegated to the NFVO (NFV Orchestration).

On on hand, the main benefit of using MEC in NFV is to achieve low latency. MEC
offers the possibility to host virtualized network functions closer to the user devices. Moreover,
Service Function Chaining (SFC) is a part of NFV systems which use to connect to VNFs to
follow an order and the data traffic to be flowed through the chain for carrying out the whole
service. The use of MEC enhances the performance of SFC because deploying the functions
closer to the end users decreases the latency and eliminates long-haul transmission of data
traffic for carrying out the whole service.

On the other hand, the use of NFV in MEC will increase the scalability of MEC
applications. NFV provides the high scalability by scaling in and out the networks resources
depending on need and application usage. Thus, combine use of these two technologies will
deliver a dynamic, quick and scalable computing platform for 5G ecosystem.
Dept. Of ISE, HKBKCE 9 2019-2020
Software Defined Networking Multi-Access Edge Computing

Chapter 3
SOFTWARE DEFINED NETWORKING

3.1 Software Defined Networking

SDN is an emerging network concept that proposes to decouple the control plane
functions from data plane of a switch. Moreover, it eliminates the use of vendor specific back-
box hardware and promotes the use commodity switches in data plane.

Transferring network control functionalities to centralized entities demands SDN

controllers to be located closer to data plane to reduce latency for critical applications. In this
regard, MEC can be ideal solutions to satisfy the latency requirement. MEC complements the
SDN advancement of the transformation of the mobile network into a softwarized networks,
ensuring highly efficient network operation and service delivery. Thus, the popularity of SDN
in various domains including 5G and IoT will further fuel the adaption of MEC concept as well.

SDN has capabilities of orchestrating the network, its services and devices by hiding the
complexities of the heterogeneous mobile environment from end users. Thus, SDN has a
significant potential for mitigating the barriers and restrictions that multi-tier MEC
infrastructure will encounter. Illustrates a usage possibility of SDN for MEC ecosystem. For
instance, The SDN control mechanism can lower the complexity of MEC architecture by
offering a novel approach to the networking and utilizing the available resources in a more
efficient manner.

SDN can dynamically route the traffic between tier-MEC servers and cloud servers to
provide the highest quality of service to end users. Moreover, SDN paradigm concentrates the
network intelligence at the central software-based controller. This will relieve the relatively
simpler MEC devices from executing the complex networking activities such as flow
management, service discovery and orchestration.

Dept. Of ISE, HKBKCE 10 2019-2020

Information Centric Networking Multi-Access Edge Computing

Chapter 4
INFORMATION CENTRIC NETWORKING

4.1 Information Centric Networking

To support the ever increasing bandwidth demand and low latency for Internet
applications such as 4K/HD Videos, 3D games, and AR (Augmented Reality)/ VR (Virtual
Reality), several networking technologies are developed over the past decades. Most of these
technologies focused on utilizing caching, replication and content distribution in optimum ways.
Similar to the MEC, ICN is another network concepts which can satisfy this demand. In
particular, ICN is an Internet architecture that puts information at the center where it needs to be
and replaces the client-server model by proposing a new publish-subscribe model.

Several benefits can be achieved by exploiting the synergy between MEC and ICN. The
use of ICN can solve some MEC issues related to the content delivery and application level
reconfiguration. ICN can offer high speed content delivery between the MEC and central cloud
systems. Application level reconfiguration is challenging in MEC systems, since a session re-
initialization is required whenever a session is being served by a non-optimal service instance.
This process always increases the session migration delay and significantly affects the low
latency applications. ICN can reduce the application level reconfiguration delay by minimizing
the network configuration delay for MEC applications. Due to the servicecentric networking
characteristics in ICN technology, it allows fast resolution of named service instances.

The coexistence of ICN and MEC can also improve the performance of the edge storage
and caching function at the edge networks. This is enabled by two features of ICN naming
location independent data replication and opportunistic caching at strategic points in the
network. These features benefit both realtime and non-realtime 5G applications where a set of
users share the same content.

ICN will significantly improve the efficiency of session mobility in MEC networks with
the optimal operational cost and bandwidth utilization for signaling traffic. In contrast to the
anchor-based mobility approach used in current MEC networks, ICN handles session mobility
by using application bound identifier and location split principles which have significantly
reduced control and user plane overheads.

Dept. Of ISE, HKBKCE 11 2019-2020

Information Centric Networking Multi-Access Edge Computing

The real-time context aware applications could be accomplished with the correct
coordination between MEC platforms. ICN provides considerable opportunities for context-
aware data distribution in the networks by allowing content distribution over unreliable radio
links and transparent mobility between heterogeneous network. Due the latency support of
MEC, ICN-MEC integration is important to provide services for high mobility 5G applications
such as tactile Internet and autonomous vehicles.

Dept. Of ISE, HKBKCE 12 2019-2020

Network Slicing Multi-Access Edge Computing

Chapter 5
NETWORK SLICING

5.1 Network Slicing

Network Slicing has also emerged as a key concept for providing an agile and dynamic
networking platform on demand. It allows multiple virtual networks to be created on top of a
common shared physical infrastructure.

Upcoming 5G architecture will utilize both MEC and net-work slicing [10] along with
other technologies. MEC and network slicing can be used together to provide in different
application domains in 5G. One of such application domains is Massive IoT [11]. In order to
support massive IoT systems, the network should be able to satisfy the requirements such as
massive cost reduction, network scalability and edge analytics.

Network slicing with the MEC-based analytics and security assets can be used to deliver
these requirements. Another use case is critical communications for delay critical applications
such Tactile Internet, autonomous driving and industrial Internet. The key requirements to
enable the critical communications are reduced latency and traffic prioritization. While MEC
can be used to reduce the latency, network slicing can support traffic prioritization. As shown in,
network slicing can help to divide the MEC resources to different slices based on tenants’
demands.

Moreover, network slicing can enable dynamic and in short life cycles for network
services. This feature will enable new value creation opportunities where resource sharing
among virtual MNOs (Mobile Network Operators), services and applications in time share
manner. MEC provides edge analytics and faster security assets for better network slicing
which will lead to massive cost reduction and increase of network scalability.

MEC has been identified as one of the key attributes to realize the aforementioned
network slicing extensions in 3GPP (3rd Generation Partnership Project) toward full multi-
tenancy. Thus, the synergy between MEC and network slicing will play a critical role in the
deployment of 5G applications.

Dept. Of ISE, HKBKCE 13 2019-2020

Internet Of Things Multi-Access Edge Computing

Chapter 6
INTERNET OF THINGS

6.1 Internet of Things

IoT is an ecosystem of connected physical objects that are accessible through the
Internet. Currently, IoT supports a myriads of application areas. Since IoT devices are low
powered with low memory footprints and less processing power, centralized cloud computing is
used for storing and processing in many IoT environments.

Although cloud computing enables the outsourcing of storage and processing

functionalities, a conventional cloud faces several challenges such as the single point of failure,
lack of location awareness, reachability, and high wide area network (WAN) latency. In this
content, MEC has risen up in order to fill the gap between the centralized cloud and IoT devices
by providing many mutual advantages. To illustrate the connection, we explain the role of MEC
in different IoT domain.

Firstly, MEC allows IoT to filter large volumes of data generated by massive IoT
applications at the edge of the network, saving the time and cost of transmitting data to the data
centers. Moreover, if the IoT devices do not need global level services, with the cloud
computing capability, MEC servers can process the packets and provide the required services
without sending to the core network. This will reduce the traffic volume in the core network and
reduce the latency.

Secondly, MEC facilitates rapid decision making based on the locally processed data by
reducing the End-to-End (E2E) delay. This is very important in the critical IoT applications
(e.g., remote surgeries, smart grid, autonomous vehicles, and video conferencing), which have
very high demands of reliability, availability, and low latency.

Thirdly, MEC will improve the scalability as the number of IoT connections increase
and reduce the drain on the battery due to the less transmission time between the device and
application server.

Dept. Of ISE, HKBKCE 14 2019-2020

Challenges and Future Work Multi-Access Edge Computing

Chapter 7
CHALLENGES, FUTURE DIRECTIONS AND
INTEGRATION PATH

7.1 Challenges, Future Directions and Integration Path

Up to this point, we illustrated the mutual benefits of using each driving technology in
MEC. We will summarize how can we utilize them to improve various features and create the
demands (e.g., IoT deployment) for MEC.

In this chapter we discuss the obstacles and challenges related to the integration of
different technologies. Based on each category, we share our insights on future directions and
integration plan.

7.2 NFV
Main research challenges and obstacles for NFV-MEC integration are the absent of
standards, immaturity of technologies, deployment complexity and new security risks. These
obstacles and challenges to be address to achieve the full integration benefits of MEC and NFV
in 5G networks.

Both NFV and MEC are recent technology are evolving through the phases of
implementation and requires standardization emanating from collaboration of industry and
researchers over an agreed platform. Specially, the interfaces and architectural components
which are required to MEC NFV integration should be defined at global level. Otherwise, wide
spread adaptation will hinder due to the compatibility issues.

Most NFV and MEC projects will face a steep learning curve in getting their
infrastructure to work as expected because of their heavy dependency on non-standardized
implementations. Due to the immaturity of both technologies, the updates are released
frequently. Thus, keeping up in an operational deployment model is hard to achieve.

Dept. Of ISE, HKBKCE 15 2019-2020

Challenges and Future Work Multi-Access Edge Computing

Minimizing the latencies through optimal utilization of resources can be achieved with
the efficient deployment of MEC services. However, it is difficult to optimize the MEC services
when they are depending on complex system components such as NFV. Specially, de-facto
NFV standard implementations such as OpenStack, is also notoriously difficult to learn, deploy,
and use.

NFV-MEC integration creates several new security challenges as well. On one hand,
MEC introduces software components such as MEPM, VNFM to NFV deployments. These
components are not a part of traditional NFV model and they cause to create a longer chain of
trust. On the other hand, the NFV features such as resource pooling can lead to the sharing of
risk between multiple unrelated MEC domain. For instance, an attack on a certain VNF might
affect other VNFs running on the same VM (Virtual Machine) or physical server.

7.3 SDN
There are a few research challenges such as new security risks and modification of APIs
(Application Programmable Interfaces) which need to be addressed to achieve the full
integration benefits of MEC and SDN technologies.

The use of SDN introduces new security threats on MEC systems. SDN has many
security threats including SDN protocol weaknesses, information disclosure through
interception, flow poisoning, side channel attacks and DoS (Denial of Service) on SDN
controller. Since MEC extends cloud computing capabilities to the edge of mobile networks, the
level of protection that can be offered to the edge hosts is low compared to what is obtainable in
traditional large data centers. Thus, the integration of MEC with SDN will further reduce the
protection of SDN systems and the impact of SDN based attacks will results the service
degradation on MEC systems.

In addition, MEC has its own security issues. For instance, proposed architectural
modifications in MEC create a number of security vulnerabilities such as malicious mode
problems, privacy leakages, and VM manipulation. The impact of these MEC threats on the
open network based SDN is more devastating. In contrast to traditional black-box type network
devices, SDN uses software programmable common standard backhaul devices. It will not only
ease the work of network administrators but also allow malicious attackers to deploy attacks.

Dept. Of ISE, HKBKCE 16 2019-2020

Challenges and Future Work Multi-Access Edge Computing

MEC-SDN interworking will also introduce several connectivity challenges. Similar to

SDN southbound, northbound, east/west interfaces, MEC also have three interfaces, namely
1) Northbound connections, which are the connections between MEC servers and a
Cloud service (public or private),
2) Southbound connections, which are the connections between the MEC servers and
the Edge devices and
3) East/West connections, which are the connections between MEC themselves, so that
MEC servers can communicate without requiring, cloud connectivity.
Since they serve similar purpose in high level, it is necessary to merge similar interface
to reduce the signaling overheard. Furthermore, use of many inferences reduces the security of
the network.

It is also necessary to define APls so that applications and services can program network
functions and SDN network directly bypassing control and management to optimize the
performance, For instance, such APIs are need to support ultralow latency applications.
Otherwise, information exchange between MEC and SDN systems will introduce additional
delays in network operations.

7.4 ICN
As per today, MEC and ICN are complementary concepts that are mostly deployed
independently. Good coordination is needed to obtain the best outcomes of their synergies.
Proper APIs have to be defined in order to communicate between the systems. Although there is
on-going research to define interfaces for MEC-NFV and MEC-SDN integrations, the interface
for MEC-ICN communication is yet to be defined. Moreover, ICN-MEC integration requires a
cross-domain industry collaborations and standard development organizations.

Specially, it is important to develop system control orchestrator and/or coordination

architecture to enable the cooperation of two systems. Moreover, such architecture should
focused on automatic and autonomic system control rather than the traditional
provisioning/configuration or distributed control of networking systems.

The real advantages of MEC can be achieved by obtaining context information such as users’
location, other users in vicinity, condition and resources in the environment. Although ICN can
provide different levels of context information (application, network and device level), their
Dept. Of ISE, HKBKCE 17 2019-2020
Challenges and Future Work Multi-Access Edge Computing

simultaneous retrievals are still challenging. Most of the current ICN research is focused on
providing the basic functionality, rather than on utilizing the available context information to
improve network parameters such as Quality of Service (QoS) and Quality of Experience (QoE).
In ICN-MEC collaborations, it is required to examine typical scenarios encompassing different
5G applications (AR/VR, autonomous driving, Tactile Internet) with varying context.

Moreover, MEC-ICN integrated systems face severe challenges due to authorization and
access control issues in ICN systems. User/level authorization is a significant challenge in ICN
due to the the lack of user-to-server authentication. ICN systems can not use traditional access
control schemes based on Access Control List (ACL). The in-network caching function of ICN
enable the possibility to deliver Named Data Objects (NDOs) on demand basis. In this
environment, ICN entities have to maintain an identical control policy over NDOs for each
consumer to support ACL based access control. However, such mechanism is challenging due
to privacy issues and computational overhead.

7.5 NS
There are several challenges that has to be address to achieve the true cooperate benefits
of network slicing and MEC technologies.
The inter-system vertical coordination between the two technologies should be clearly
structured and modeled for efficient information sharing. This vertical coordination can be
chieved via two ways. First method is to define APIs between management systems of slicing
and MEC to share the available resources. Second method is to use physical resource
coordination aimed to efficient resource handling through policy and analytics. However,
standardization these interfaces have to be performed to synchronize the various research and
development activities around the globe.

If the MEC servers can offer service composition with finegrained network functions, it
will enhance the scalability tosupport different vendors. Each coarse grained function at the
MEC server can be further divided into many sub-functions. Nevertheless, the granularity of
those networking functions should be defined carefully in such a way to comply with the
available standardized interfaces. Otherwise, same as in the above case the interfaces need to be
recognized by the respective standardization entities.

Dept. Of ISE, HKBKCE 18 2019-2020

Challenges and Future Work Multi-Access Edge Computing

When multiple RATs are accommodating the 5G paradigms, there should be some ways
to multiplex them on specialized or dedicated hardware. Although network slicing may lead to
virtualize RAN instances, it is mandate to ensure radio resource isolation and manage efficiency.
In order to assist RAN virtualization for slicing, Software Defined ) RAN controllers can be run
at the MEC servers.

Even though the high-level description of a concrete slice in terms of infrastructure and
network functions is defined, the physical realization of E2E slice orchestration is still to be
established. As the intermediary edge computing platforms, the MEC servers play a vital role to
support E2E slice orchestration and management by correlating cloud and radio resources.

7.6 Iot
Although IoT is becoming a mature technology, the constant addition of new IoT
devices to the consumer market and their versatile connectivities create myriads of security and
scalability issues. When MEC is coincided with IoT, these challenges will arise in a broader
manner. Security, privacy, and trust management are three important research areas of IoT
which have many synergies. Edge computing users are increasingly vulnerable to security
threats as more IoT devices and applications use the edge to transmit information. The
heightened exposure of user’s data in MEC and IoT may create many possible ways which the
sensitive data can be breached. With the growing intelligence of the smart devices, there can be
the possibility of one IoT device betraying the implicit trust of another IoT device. Typically,
the IoT devices are programmed to automatically trust another connected device and to share
data without a validation process. If all the devices natively trust each other and share data, it is
hard to identify whether there is a misbehaving device. This may create a big issue specially
with the absence of a perimeter around the network edge prohibits firewalls to block out MEC
security threats. Moreover, the frequent back and forth transfer of data from a device to the
network edge increases the opportunity for breaching data and hacking devices. In MEC
systems, it is challenging to identify, authenticate, and authorize devices and the data they
generate from the edge to cloud and back while maintaining a latency of some milliseconds.

Scalability and mobility are two other important features that MEC provides for IoT
applications. In many applications, IoT devices will require scalability of services by applying
load balancing mechanisms which can be managed by a cloud orchestrator in MEC nodes. As
edge devices uses different access technologies including 3G, 4G, 5G, Wi-Fi and Wi- Max the
Dept. Of ISE, HKBKCE 19 2019-2020
Challenges and Future Work Multi-Access Edge Computing

aspect of heterogeneity and scalability should be catered in smooth functioning of MEC

operations. Similarly, it is challenging to achieve non-negligible impact on caching and
computation offloading decisions with the user mobility which will cause frequent handovers
among edge servers.

7.7 Integration Path

We explain here the integration paths and pinpoint tangible steps to realize them.
1) Control Level Orchestration:
To achieve added benefits in MEC-enabled networks, different technologies must work
together. Given that 5G is going to integrate those five driving technologies with MEC, as
depicted, it is expected to enable them simultaneously in 5G. Meanwhile, such integration will
face challenges at control level. For instance, SDN, NFV, Network Slicing and MEC are
independent technologies. Each technology utilizes its own orchestrator and management
entities such as SDN controller, NFV orchestrator, network slicing manager and mobile edge
platform manager. In this respect, a synergy between these control entities is needed to jointly
optimize the network resources and also to create efficient Service Function Chains (SFCs) for
each user application.

2) Synchronization of Standardization Process:

In order to achieve orchestration in MEC systems, different technological components
need to inter-communicate with each other. This demands that communication interfaces have
to be defined at architecture level. However, the standardization of different technologies are
coordinated by different organizations, e.g., MEC and NFV by ESTI, SDN by ONF, ICN by
IETF, IoT by IEEE,and Open Internet Consortium (OIC). Therefore, the cooperative efforts of
different standardization bodies are needed to make synergies between the interfaces defined for
different technologies. As a good example, ETSI has already started defining the interfaces for
NFV and MEC integration. However, these efforts need to expand for other
technology domains as well.

3) Hardware Limitations and Dependencies:

The integration of driving technologies demands changes not only at the control plane
but also at the data plane and hw/sw components. For instance, SDN-enabled switches and
devices are needed at the infrastructure layer to implement SDN. Similarly, ICN enabled
switches are needed to enable ICN functionality. Installation of such multi-technology hardware
Dept. Of ISE, HKBKCE 20 2019-2020
Challenges and Future Work Multi-Access Edge Computing

is challenging. First, standardization of different technology integration should be finalized so

that vendors can start building such multi-technology hardware equipments. Second, extensive
hardware resources are needed to implement multitechnology concepts. Therefore, these
hardware limitations and dependencies has to be solve to obtain full benefits of technology
integration.

4) Security and Privacy:

To deploy MEC at large scale, security and privacy must be enforced by all five
technologies across multiple layers in MEC. In particular, since IoT is closely coupled with
public and enterprise infrastructure, this cyber-physical setting renders ample use cases and also
generates security challenges for MEC to integrate NFV, SDN, ICN and network slicing. The
principle of security and privacy by design should hence be mandated in the integration process.

5) AI as a key integration enabler:

Recently artificial intelligence (AI) and machine learning (ML) have been resorted to
create smarter and autonomous wireless systems. In the 5G context, AI can directly benefit the
driving technologies such as SDN and NFV to be integrated into MEC. For instance, AI-based
edge orchestrators can be used for better system and host level management functions for
various NFV based use cases. AI and MEC together (i.e., edge automation) will combat
towards low latency for real-time services, better orchestration, enhanced security, and backhaul
cost savings.

Dept. Of ISE, HKBKCE 21 2019-2020

Snapshots Multi-Access Edge Computing

Chapter 8
SNAPSHOTS

Fig. 1: MEC in NFV Architecture

Fig. 2: The use of SDN for MEC

Dept. Of ISE, HKBKCE 22 2019-2020

Snapshots Multi-Access Edge Computing

Fig. 3: Use of Network Slicing and MEC in different 5G applications

Fig. 4: Integration of Driving Technologies in MEC Systems

Dept. Of ISE, HKBKCE 23 2019-2020

Snapshots Multi-Access Edge Computing

Dept. Of ISE, HKBKCE 24 2019-2020

Conclusion Multi-Access Edge Computing

CONCLUSION

This helps in analyzing five integration technological directions that can accelerate the
utilization and deployment of MEC in 5G mobile networks, including NFV, SDN, ICN,
Network Slicing and IoT. We highlight the benefits of using each technology in MEC systems
as well as identify the remaining challenges and integration path to realize the full integration
benefits. With careful integration, the suite of these solutions will form the future of MEC
deployments.

Dept. Of ISE, HKBKCE 25 2019-2020

Conclusion Multi-Access Edge Computing

REFERNCES

[1] Y.-C. Hu, M. Patel, D. Sabella, N. Sprecher, and V. Internet Architecture. ACM, 2015, pp. 8–13.
Young, “Mobile Edge Computing : A Key Technology
Towards 5G,” ETSI White Paper, vol. 11, no. 11, pp. [12] P. Porambage, J. Okwuibe, M. Liyanage, M.
1–16, 2015. Ylianttila, and T. Taleb, “Survey on multi-access edge
computing for internet of things realization,” IEEE
[2] R. Mijumbi, J. Serrat, J.-L. Gorricho, N. Bouten, F. Communications Surveys and Tutorials, 2018.
De Turck, and R. Boutaba, “Network Function
Virtualization: State-of-the-art and Research [13] F. Bonomi, R. Milito, P. Natarajan, and J. Zhu,
Challenges,” Communications Surveys & Tutorials, “Fog Computing: A Platform for Internet of Things
vol. 18, no. 1, pp. 236–262, 2016. and Analytics,” in Big data and internet of things: A
roadmap for smart environments. Springer, 2014, pp.
[3] L. Gupta, R. Jain, and H. A. Chan, “Mobile Edge 169–186.
Computing–An Important Ingredient of 5G
Networks,” IEEE Software Defined Networks, [14] D. Grewe, M. Wagner, M. Arumaithurai, I. Psaras,
Newsletter, 2016. and D. Kutscher, “Information-Centric Mobile Edge
Computing for Connected Vehicle Environments:
[4] “Mobile Edge Computing (MEC); Deployment of Challenges and Research Directions,” in Proceedings
Mobile Edge Computing in an NFV environment,” of the Workshop on Mobile Edge Communications.
ETSI Industry Specification Group (ISG) White Paper, ACM, 2017, pp. 7–12.
February 2018.
[15] L. Iliadis, I. Maglogiannis, and V. Plagianakos,
[5] D. Kreutz, F. M. Ramos, P. E. Verissimo, C. E. Artificial Intelligence Applications and Innovations.
Rothenberg, S. Azodolmolky, and S. Uhlig, Springer, 2018, vol. 520.
“Software-Defined Networking: A Comprehensive
Survey,” Proceedings of the IEEE, vol. 103, no. 1, pp.
14–76, 2015.

[6] B. Blanco, J. O. Fajardo, I. Giannoulakis, E.

Kafetzakis, S. Peng, J. P´erez-Romero, I. Trajkovska,
P. S. Khodashenas, L. Goratti, M. Paolino et al.,
“Technology Pillars in the Architecture of Future 5G
Mobile Networks: NFV, MEC and SDN,” Computer
Standards & Interfaces, vol. 54, pp. 216–228, 2017.

[7] A. V. Vasilakos, Z. Li, G. Simon, and W. You,

“Information Centric Network: Research Challenges
and Opportunities,” Journal of Network and Computer
Applications, vol. 52, pp. 1–10, 2015.

[8] R. Ravindran, A. Chakraborti, S. O. Amin, A.

Azgin, and G. Wang, “Realizing ICN in 3GPP’s 5G
NextGen Core Architecture,” arXiv preprint
arXiv:1711.02232, 2017.

[9] K. Samdanis, X. Costa-Perez, and V.

Sciancalepore, “From Network Sharing to Multi-
tenancy: The 5G Network Slice Broker,”
Communications Magazine, vol. 54, no. 7, pp. 32–39,
2016.

[10] N. Alliance, “Description of Network Slicing

Concept,” NGMN 5G P, vol. 1, 2016.

[11] N. Nikaein, E. Schiller, R. Favraud, K. Katsalis,

D. Stavropoulos, I. Alyafawi, Z. Zhao, T. Braun, and
T. Korakis, “Network Store: Exploring Slicing in
Future 5G Networks,” in Proceedings of the 10th
International Workshop on Mobility in the Evolving
Dept. Of ISE, HKBKCE 26 2019-2020