Resource Allocation in IRSA-Assisted NOMA for Massive URLLC Using Lightweight Q-Learning | Publicación