Latency-Based 5G RAN Slicing Descriptor to Support Deterministic Industry 4.0 Applications

Abstract

5G networks can support the development of the Industry 4.0. To this aim, 5G must be able to guarantee the deterministic latency requirements that characterize many industrial applications. This objective can be achieved using network slicing, a novel 5G paradigm that exploits the softwarization of networks to create different logical instances of the network over a common network infrastructure. Each instance is configured to support specific applications. Slicing can be applied at the Core Network or at the Radio Access Network (RAN). This study focuses on RAN slicing since the RAN typical accounts for a large part of the end-to-end delay. RAN slicing splits (and configures) resources at the RAN level between the slices in order to adequately serve nodes with a particular profile. This includes identifying the necessary radio resources per slice. To date, most proposals define slices in terms of the number of required radio resources. While this descriptor can account for bandwidth or rate requirements, it does not adequately reflect the latency requirements characteristic of many Industry 4.0 applications. This paper proposes a novel latency-based RAN slice descriptor and demonstrates that the new descriptor improves the capacity of RAN slicing to meet the latency requirements of Industry 4.0 applications with deterministic periodic traffic.