5G RAN Slicing for Deterministic Traffic

A key aspect of 5G is the facilitation of network convergence. It enables the support of a broad range of use cases including real-time applications over the same shared network infrastructure. One major challenge in addressing the various latency requirements is the efficient sharing of resources for heterogeneous services under the presence of different scheduling timescales. State-of-the-art solutions propose preemptive approaches to satisfy instantaneous time-critical demands, which however can deteriorate the communication for preempted users as well as the overall rate efficiency in the network. We propose a slicing method tailored for deterministic traffic, as commonly found in industrial applications. By exploiting deterministic properties of resource demands, we determine an optimal share of reserved and preempted resources on a per-slot basis to efficiently satisfy all time-critical demands. Additionally, we introduce a preemption restricted slice to protect users with higher reliability requirements. Under this proposed model, we derive an online scheduling algorithm which maintains a better fairness and higher throughput compared to state-of-the-art solutions under high demand of deterministic traffic.