MEC-Assisted End-to-End 5G-Slicing for IoT

Abstract

The Internet of Things (IoT) has emerged as a key horizontal technology enabler within the 5G ecosystem, characterized by supporting the heterogeneity of end-devices, radio-access technologies, and services. Network slicing techniques allow to configure dedicated networks with assured resources to specific users or applications over a common physical infrastructure. This paper describes a novel end- to-end network slicing framework for IoT services in 5G systems, based on a Multi-Access Edge Computing (MEC) and central cloud architecture that permits the flexible and dynamic placement of micro-services encapsulated as Virtualized Network Functions (VNFs) along the end-to-end path. The critical analysis of the validation results highlights the performance and flexibility gains of our proposal. It allows deploying end-to-end slices over a number of participating domains in a very short time, it is able to isolate slices with guaranteed Quality of Service (QoS) parameters, and shows great scalability in terms of number of connected end-devices and application services.