Technical Report: A Hierarchical Dynamically Weighting Deep Reinforcement Learning Method for Multi-UAV Multi-Task Coordination

This collection of networking and ML papers highlights multiple advances and safety/availability risks for time-sensitive and safety-critical networks. TSNBench demonstrates that current LLMs perform well on multiple-choice TSN questions but fail catastrophically on worst-case delay (WCD) computations (large MAPE), meaning LLM-assisted configuration or automation could produce unsafe TSN schedules and violate real-time guarantees. PolicyCache-SDN offers a mitigation pattern (auditable per-path policy envelopes and local online adaptation) that reduces distribution-shift fragility. LUDB++ and E