Content uploaded by Madhusanka Liyanage
Author content
All content in this area was uploaded by Madhusanka Liyanage on Apr 21, 2021
Content may be subject to copyright.
Identifying Factors Enabling the Enhancement of
Service Migration of Multi-Access Edge Computing
Pasika Ranaweera∗, Anca Delia Jurcut†, Madhusanka Liyanage‡
∗† ‡School of Computer Science, University College Dublin, Ireland
‡Centre for Wireless Communications, University of Oulu, Finland
Email: ∗pasika.ranaweera@ucdconnect.ie,†anca.jurcut@ucd.ie ,‡madhusanka@ucd.ie, ‡madhusanka.liyanage@oulu.fi
Abstract—Edge computing is a novel concept proposed to
overcome the limitations of the prevailing cloud-based telecom-
munication networks. Various concepts have emerged with edge
computing that requires proper investigation prior to deployment.
Migration of services within the edge computing nodes/ base
stations is an imminent aspect of the envisaged paradigm that
has created a lot of attention. The selection of the optimum edge
node to migrate the service is such an issue that restricts the
advancement of edge paradigms. The sole focus of this research is
to identify and validate the factors enabling the optimal migration
decision considering the Multi-Access Edge Computing paradigm.
I. INTRODUCTION
Multi-Access Edge Computing (MEC) is one of the leading
edge computing paradigms introduced by ETSI for overcoming
the limitations of cloud computing based network infrastruc-
tures; by placing the storage and processing infrastructure in
proximity to the User Equipment (UE) [1]. Edge computing,
evolving beyond cloud computing in terms of mobile network
integration and heterogeneity support for Internet of Things
(IoT) devices, enables the deployment of emerging 5G ap-
plications. Such stipulated use cases of 5G include: massive
Machine Type Communication (mMTC), Autonomous Vehicle
(AV) Driving, Unmanned Aerial Vehicles (UAV), and Ultra
Reliable Low-Latency Communication (URLLC) [2]. But due
to the architectural changes in the mobile network based stor-
age and processing infrastructure, unprecedented challenges of
service migration, mobile offloading, technological integration,
communication, mobility management, security, and privacy
are emerged.
Service migration is the process of transferring executable
content incepting in an application or a service, among cloud
and edge computing environments. These migrations are occur-
ring either from cloud-to-edge or edge-to-another-edge envi-
ronment. Typically, a particular Mobile Edge Service (MES) is
not available at all the edge nodes due to the envisaged hetero-
geneity and scalability of IoT resembled services. Therefore,
in a situation where a UE/ IoT device/ AV/ UAV is crossing
over to a different coverage area from the serving edge node
(referred as a handover), the serving MES should be migrated
to the roaming edge node for mobility management and service
continuity as depicted in Fig. 1. Thus, migration is opening
novel avenues for researchers to investigate various aspects of
managing migration channel capacity, edge resource utilization
during migration, and security of the migration channel from
the edge computing perspective; minimizing latency, migration
down-time, and selecting the optimum edge node with MEC
capabilities to migrate the service are aspects important for
MEC Edge B
MESAV
MEC Edge A
UAV/ Drone
AV
MESUAV
Where to migrate?
Traversed UE
Location
Ttravel
Tmigration
Tconfig
Ttravel Tmigration Tconfig
<+
MEC-1
MEC-2
MEC-3
MEC-4
Storage
Processing
Migration Channel
d
Fig. 1: Issues of MEC based service migration
maintaining the Quality of Service (QoS) and Quality of
Experience (QoE) standards from the MES perspective.
Among the stated research directives of the service mi-
gration phenomenon, a strategy to select the optimum MEC
enabled edge node to migrate (i.e. when there are multiple
edge nodes in the proximity of the UE that possess different
capabilities) is an aspect that lacks in the current research.
Eventhough the closest edge node might grant the highest radio
access capacity, its current resource utilization, status of the
migration channel, and the availability of the considered MES
are factoring in for the selection process. Thus, it is imperative
to identify the factors that aids us to select the optimum edge
node. In this research, our prime focus is to investigate such
factors/ parameters that governs the service migration process
in addition to guiding the optimum migration selection.
II. MEC-BAS ED ENABLING FAC TORS TO ENHANCE
SERVICE MIGRATION
The factors or parameters enabling the governance of
service migration phenomenon can be specified from com-
munication and operational capability of the edge node, and
backhaul capacity of the migration channel.
1) Communication and Operational Capability of the
MEC Edge Node
Communication Capability : Signal-to-Interference-Noise-
Ratio (SINR) measurement of a radio channel gives a good
perception on the Received Signal Strength (RSS), that corre-
sponds to the communication capacity of the migrating MEC
node. The aerial distances derived from geo-locations (assum-
ing obstacle-less line-of-sight radio links) can be employed to
determine the SINR, where data rates in bps can be computed.
A permissible communication range (R- minimum data rate,
SINR level [3], or number of edge nodes) should be established
to conduct the selection process.
Operational Capability : the operational capability of the
considered MEC edge node can be determined by the consum-
ing computing (C) and storage (S) capacities at the considered
instance. The factors of latency, jitter, and priority level of
the MES are contributing to the Cand Sparameters. QoS
Class Identifier (QCI) standards and specifications are aiding to
model these two factors in line with the emerging applications
and services [4].
2) Backhaul Capacity of the Migrating Channel
The edge-to-edge Bandwidth (BW) of the backhaul link
is critical for modelling the service migration process. The
existing backhaul links are typically employed for signaling
purposes with limited capacity. Though, novel services tend to
utilize these links for improving QoE aspects with embedded
intelligence. As migrations are less-occurring events, there
is no guarantee that available backhaul capacity would be
sufficient for migration initiation. For the MEC node selection
model, migration time (computed from available link capacity
and size of the migrating content) for a specific MES can be
considered as a viable input parameter.
III.VALIDATIN G TH E PROP OS ED ENABLING FAC TORS
Two scenarios can be considered to validate the proposed
factors. In the legacy scenario MEC edge node is selected
based only on the communication capability (i.e. data rate)
as presented in [3]. The proposed scenario considers com-
munication, operational, and backhaul capabilities for the
selection process. A simulation was carried out considering
25 edge nodes located at Dublin city, Ireland (i.e. longitude
53.3243282◦
∼53.3654212◦, and latitude −6.2956804◦
∼
−6.2071241◦), with their actual geo-locations and BWs ex-
tracted from [5]. We assumed that all the nodes were MEC
enabled. Further, MESs were incepted based on the QCI levels
1,2,3,4,5,6,7,8,9,70,80,84 to randomize the Cand Sconsump-
tion. The 3GPP propagation model in [6] was followed to
compute the data rates. In addition, parameters in TABLE
I were considered to perform the simulation, where Rwas
considered as 10.
The simulation was carried out for continuous 500 trials
where the UE location was randomized withing the grid to
determine the success of launching the migrated services at
the roamed MEC edge node based on its resource availability.
It is observed from Fig. 2, the success rates of both legacy and
proposed scenarios are 33.4% and 98.8% respectively. Thus,
the proposed enabling factors are suitable for selecting the
optimum MEC edge node for migration.
TABLE I: General Simulation Parameters
Parameter Values
Number of eNBs 25
Average eNB BW 15 MHz
Maximum transmission power of a eNB 46 dBm
Average resultant antenna gain at eNBs 5 dBm
Average resultant noise power -92 dBm
Number of MESs Normalized µ/σ100/ 90
Maximum eNB computing capacity 100 GHz [7]
Maximum eNB storage capacity 100 TB [7]
Maximum computing capacity of a MEC App 60 MHz
Maximum storage capacity of a MEC App 100 GB
Fig. 2: Simulation results on the success of launching the
migrated services.
IV. CONCLUSION
This study was conducted to identify the enabling param-
eters that govern the service migration process of MEC. The
presented validation proves the effectiveness of the identified
parameters defined in terms of communication, operational
capability of edge nodes, and backhaul capacity of the mi-
grating channels. These factors will provide a baseline for
formulating solutions to current issues of service migration:
security, latency, mobility and handover management.
ACKNOWLEDGMENT
This work is supported by Academy of Finland in 6Genesis
(grant no. 318927) projects.
REFERENCES
[1] P. Ranaweera, A. D. Jurcut, and M. Liyanage, “Survey on multi-access
edge computing security and privacy,” IEEE Communications Surveys
& Tutorials, 2021.
[2] C. de Alwis, A. Kalla, Q. V. Pham, P. Kumar, K. Dev, W. J. Hwang,
and M. Liyanage, “Survey on 6G Frontiers: Trends, Applications, Re-
quirements, Technologies and Future Research,” IEEE Open Journal of
the Communications Society, pp. 1–1, 2021.
[3] J. Plachy, Z. Becvar, and E. C. Strinati, “Dynamic resource allocation
exploiting mobility prediction in mobile edge computing,” in 2016 IEEE
27th Annual International Symposium on Personal, Indoor, and Mobile
Radio Communications (PIMRC). IEEE, 2016, pp. 1–6.
[4] M. Mamman, Z. M. Hanapi, A. Abdullah, and A. Muhammed, “Quality
of service class identifier (qci) radio resource allocation algorithm for lte
downlink,” PloS one, vol. 14, no. 1, p. e0210310, 2019.
[5] Eir. Eir (ireland) cell tower map. [Accessed November 30, 2020].
[Online]. Available: https://www.cellmapper.net/
[6] E. Access, “Further advancements for e-utra physical layer aspects,”
3GPP Technical Specification TR, vol. 36, p. V2, 2010.
[7] 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.
