SYNC-ADAPTIVE RESOURCE SCHEDULING IN PRIVATE 5G FOR TIME-CRITICAL APPLICATIONS

Abstract

Private 5G networks enable Industry 4.0 through ultra-reliable low-latency communication (URLLC) for time-critical applications such as closed-loop control of PLCs, robots, and AGVs, requiring sub-microsecond end-to-end synchronization. Conventional static schedulers assume fixed timing accuracy, leading to over-provisioning or failures under variable GNSS/PTP conditions. This paper introduces a sync-adaptive resource scheduling algorithm that dynamically tunes time-slot granularity and allocation based on real-time Pulse-Per-Second (PPS) error measurements.The core model formulates joint optimization as a mixed-integer linear program: minimize latency subject to PPS error constraints, QoS requirements, resource limits, and multi-tenant slicing. Tighter sync (e.g., <100ns error) unlocks aggressive mini-slotted patterns with priority queuing; coarser sync (>500ns) activates conservative redundancy and wider slots. Solved via priority-embedded DDPG heuristics in O-RAN RIC, it supports Near-RT adaptation.Evaluations in ns-3 simulated private 5G environments show 25-40% latency reduction versus PF/RR baselines, with 99.999% reliability under interference and mobility. Robustness extends to TSN integration, enhancing deterministic performance for OT/IT convergence without excessive resources. This scheduler advances ORAN xApps for adaptive URLLC in dynamic industrial settings.