Page 69 - ITU Journal Future and evolving technologies Volume 2 (2021), Issue 6 – Wireless communication systems in beyond 5G era

ITU Journal on Future and Evolving Technologies, Volume 2 (2021), Issue 6

MULTI‑TIER MULTI‑TENANT NETWORK SLICING: A MULTI‑DOMAIN GAMES APPROACH

1
1
Sunday Oladayo Oladejo , Stephen Obono Ekwe , Lateef Adesola Akinyemi 1,2
1 Department of Electrical Engineering, University of Cape Town, Cape Town, South Africa, Department of Electronic and
2
Computer Engineering, Lagos State University, Epe, Lagos State, Nigeria
NOTE: Corresponding author: Sunday Oladayo Oladejo, oldsun002@myuct.ac.za

Abstract – The 5G slice networks will play a critical role in meeting the stringent quality‑of‑service requirements of differ‑
ent use cases, reducing the Capital Expenditure (CapEX) and Operational Expenditure (OpEX) of mobile network operators.
Owing to the lexibility and ability of 5G slice networks to meet the needs of different verticals, it attracts new network players
and entities to the mobile network ecosystem, and therefore it creates new business models and structures. Motivated by this
development, this paper addresses the dynamic resource allocation in a multi‑slice multi‑tier multi‑domain network with dif‑
ferent network players. The dynamic resource allocation problem is formulated as a maximum utility optimisation problem
from a multiplayer multi‑domain perspective. Furthermore, a 3‑level hierarchical business model comprising Infrastructure
Providers (InPs), Mobile Virtual Network Operators (MVNOs), Service Providers (SPs), and slice users are investigated. We
propose two schemes: a multi‑tier multi‑domain slice user matching game scheme and a distributed backtracking multiplayer
multi‑domain game scheme in solving the transformed maximum utility optimisation problem. We compare the multi‑tier
multi‑tenant multi‑domain game scheme with a Genetic Algorithm (GA) Intelligent Latency‑Aware Resource (GI‑LARE) al‑
location scheme, and a static slicing resource allocation scheme via Monte Carlo simulation. Our indings reveal that the
proposed scheme signi icantly outperforms these other schemes.
Keywords – 5G, multi‑domain, multi‑tenancy, MVNO, network slicing, resource allocation, SP

1. INTRODUCTION The multi‑domain concept for NS refers to the pooling
of network resources from multiple InPs which are then
The Network Slicing (NS) paradigm entails the abstrac‑ virtualised and sliced to meet the needs of slice users
tion of the physical mobile network and its hard re‑ and diverse verticals [7]. Verticals are business and pub‑
sources into logical autonomous networks via network lic entities that run their service applications on mobile
softwarisation technology [1]. To this end, it has been networks. Instances of verticals are defence agencies,
widely acknowledged that NS will greatly improve the transport systems, health institutions, power and energy
resource utilisation of the mobile networks and signi i‑ utilities, disaster and emergency agencies, manufactur‑
cantly reduces their Capital Expenditure (CapEx) and Op‑ ing, and agricultural chains. The importance of multi‑
erational Expenditure (OpEx). Consequently, efforts have domain NS is emphasised in resource aggregation, cov‑
been geared towards achieving these goals through di‑ erage area extension, reliability of services, and the op‑
verse resource management frameworks, owing to com‑ portunity to further make the business model Incentive
plex business models of NS. However, with the lexible Compatible (IC) for network players [8, 9]. On the other
and agile characterisation of NS architecture, its busi‑ hand, in the single domain model, the NS resources are
ness model departs radically from the traditional mod‑ primarily from the virtualised resources of only one InP.
els [2, 3, 4]. Hence, new mobile entities such as Infras‑ A vast corpus of literature on NS has focused on the sin‑
tructure Providers (InPs), Mobile Virtual Network Oper‑ gle domain model despite its non‑ lexibility and limited
ators (MVNOs), Over‑the‑top Providers (OTTs), Content resource constraints which make it unsuitable for the fu‑
Providers (CPs), Service Providers (SPs), Backhaul Oper‑ ture mobile networks [8, 10]. In Fig. 1(a)‑(c), we illustrate
ators (BOs), and Vehicular‑Content Providers (VCPs) have the traditional, single, and multi‑domains multi‑tenant
become critical industry players. NS paradigms. This work focuses on the M‑TTSD net‑
To exploit the capabilities of 5G and beyond mobile net‑ work. In a multi‑domain system, several InPs provide in‑
works, the self‑interests of the diverse network players frastructure and network resources services, and to this
would have to be iciently managed. To this end, the end, the infrastructure and network resources owned by
challenge of managing network resources to achieve op‑ an InP represent a network domain. The multi‑tenancy
timal network performance and economic advantages to paradigm entails several independent MVNOs (i.e., other‑
the network players is non‑trivial. This paper examines wise known as tenants) share the same physical infras‑
the resource allocation bottleneck in a Multi‑Tier multi‑ tructure and network resources owned by an InP. The
Tenant multi‑Slice multi‑Domain (M‑TTSD) 5G and be‑ term “tenants” may be extended to SP depending on the
yond network [5, 6]. Speci ically, we consider an M‑TTSD
5G network model with a business model consisting of SP,
MVNO, InP, and slice users.

© International Telecommunication Union, 2021 57