Digital Twin-assisted belief-state reinforcement learning for latency-robust ISAC in 6G networks

Collection of recent arXiv papers (April 2026) on 5G/6G systems and edge-cloud AI for networking. Key contributions include: a Digital Twin-assisted belief-state RL controller for latency-robust ISAC that compensates for delayed telemetry; NeuralEmu, an ML-driven high-fidelity 5G scheduler emulator; clustering methods for traffic-matrix prediction; twin-in-the-loop planning for resource allocation in federated split learning; StreamGuard, a DPI-based RAN architecture for QoE-aware subflow prioritization with packet marking and subflow shaping; SWE-Bench 5G, a benchmark that evaluates AI coding