Resource Allocation for Network Slicing in 5G Telecommunication Networks: A Survey of Principles and Models

With the rapid and sustained growth of network demands, 5G telecommunication networks are expected to provide flexible, scalable, and resilient communication and network services, not only for traditional network operators, but also for vertical industries, OTT, and third parties to satisfy their different requirements. Network slicing is a promising technology to establish customized end-to-end logic networks comprising dedicated and shared resources. By leveraging SDN and NFV, network slices associated with resources can be tailored to satisfy diverse QoS and SLA. Resource allocation of network slicing plays a pivotal role in load balancing, resource utilization, and networking performance. In this article, we focus on the principles and models of resource allocation algorithms in 5G network slicing. We first introduce the basic ideas of the SDN and NFV with their roles in network slicing. The MO architecture of network slicing is also studied, which provides a fundamental framework of resource allocation algorithms. Then, resource types with corresponding isolation levels in RAN slicing and CN slicing are analyzed, respectively. Furthermore, we categorize the mathematical models of resource allocation algorithms based on their objectives and elaborate them with typical examples. Finally, open research issues are identified with potential solutions.

Introduction

The emerging fifth generation telecommunication networks (5G) are conceived to offer a large amount of end-to-end network services for diverse requirements. These requirements are created by not only traditional mobile communication applications, but also vertical market segments, such as automatic driving, unmanned aerial vehicles, telemedicine, massive Internet of Things (mIoT), and so on [1]. To serve different application scenarios, verticals may require distinct network services by means of differentiated quality of service (QoS), service level agreements (SLAs), and key performance indicators (KPIs). In this case, the “one-size-fits-all” and “one-network-fits-all” types of fourth generation (4G) telecommunication networks (i.e., all network devices follow the same pipeline with few considerations on service customization) is no longer suitable, which are evolved into customized network paradigms by leveraging network slicing.