Reliable Software-Defined RAN Network Slicing for Mission-Critical 5G Communication Networks

Emerging Industrial Internet of Things (IoT) applications impose challenging and diverse requirements on underlying communication infrastructures. Traditionally, each vertical industry utilizes dedicated communications networks. In contrast, the 5th Generation of Mobile Communication Networks (5G) aims to fulfill Quality of Service (QoS) requirements for various vertical industries. This is achieved by deploying several, virtually dedicated networks, known as slices, on top of a unified, physical communication infrastructure. While various solutions have been presented for Core Network Slicing, the here presented approach aims to provide an end-to-end solution by allocating Radio Access Network (RAN) resources, realizing slicing for mission-critical applications via a novel scheduler design. A proof of concept is provided by way of detailed empirical evaluations, based on IoT scenarios from the energy sector (i.e., Smart Grids). The proposed RAN Network Slicing solution is shown to reliably sustain service guarantees of critical applications, while coexisting with non-critical services. Application-dependent, dynamic, inter-slice resource sharing enables an efficient use of available RAN spectrum. Finally, we demonstrate dynamic adaptation of slices to channel quality, ensuring reliable operation of Industrial IoT.