Safeguard URLLC Services in Industry 4.0 Through Reinforcement Learning Scheduling

The potential of Industry 4.0 will be fully enabled through 5G network development. The 5G network functions need to deal with all diversity of applications and their stringent requirements to guarantee the proper function of industrial use cases. When dealing with 5G for industrial scenarios, it is crucial to design resource management functions to fulfill the network slice requirements and protect critical slices such
$as$
URLLC when required. In this work, we propose a radio resource scheduler for the Industry 4.0 scenario considering three services available: eMBB, URLLC, and mMTC. In this context, we design a reinforcement learning agent
$as$
an inter-slice scheduler and safe-guard the URLLC slice since it is usually associated with critical applications. The obtained results show that the proposed method obtained better performance in protecting the URLLC slice and also in minimizing the total number of violations to all the slice's requirements.