Network Slicing for 5G and Beyond Systems

Network Slicing (NS) is a holy grail nowadays for future of networking, and is poised to transform the Internet and the services eco-system. Other closely associated technologies on which NS is built on are NFV, SDN, and OpenFlow. Since 5G would enable a range of new applications and services by delivering higher bandwidth, lower latency, and significant scalability, one size fits all approach of the past would not work. This has made network slicing in 5G an essential requirement of the overall 5G architecture. The concept of network slicing enables logically partitioning the network into slices based on different considerations of traffic types, application types, customers, industry verticals, etc, such that NS provides isolation between slices for the reasons of maintaining QoS/SLA, security and potential interference. Simply stated, the operators are pursuing the strategy of “divide and conquer.” A very important concept in NS is that of Orchestrator(s) that allows slices to be configured in the access network (RAN and fixed access), the core network, and very high capacity backbone transport which almost always is fibre-based. Although there are several standards organizations and industry fora driving network slicing standardization, inter-operability and implementation details, 3GPP is at the core of driving the entire effort. This WWRF Outlook endeavours to capture the broad landscape of network slicing from basic concept to potential business and economic models. And, in-between, it describes how slices are configured in the RAN, the core network and the transport network. Since new advancements and new generations of wireless tend to overlook the requirements of the developing world (with close to over 3.5B people) who have yet to experience the value of the Internet, this Outlook dedicates a significant amount of space to the role of network slicing to bridge the digital divide as we usher a new era of 5G and B5G technologies for social impact. As network management and operations is a significant aspect of any network, this topic is covered as well in this paper followed by a discussion on implementation and testing of network slicing in future networks. On the business models aspect, network slicing is likely to unleash a plethora of new opportunities for the operators as they would be able to price the slices differently based on the QoS agreements, services offered, applications supported and who the customer is. Finally, since network slices would be tariffed differently, much different from the present day data-type agnostic best effort Internet, there potentially might be issues of network neutrality. Furthermore, since these slices would run across national boundaries, subject to different regulations and policies, it would seem that some level of harmonization and international cooperation would be required for network slicing to realize its full potential. Therefore, it would be necessary to be able to explain NS to all stakeholders, national SDOs, and government policy makers and regulators. While network slicing is predicted to lead to the promised land, but this road is not without significant pot holes along the way. Some significant challenges that must be addressed to realize the vision of E2E slicing concern re-designing RAN and CN for slicing, spectrum slicing, E2E service assurance to meet QoS/SLA, inter-operability across multiple operator network domains, maintaining service concurrency enabled by network slicing, scheduling issues when a user device wishes to access multiple slices simultaneously, addressing security issues, including distributed denial of service (DDoS) attacks, and slice management to ensure they are adhering to the agreed or configured KPIs for individual slices.