Hierarchical RL-Based Slotted ALOHA Protocol with Transmit Power Control

This paper proposes a distributed transmission slot selection and transmit power control algorithm based on hierarchical Q-learning to reduce the probability of packet collision and maximize spatial reuse in aerial broadcasting networks based on slotted ALOHA protocol where unmanned aerial vehicles (UAVs) send their location information to avoid collision among UAVs. In the proposed algorithm, each UAV operates as an independent agent, the inner loop learns the optimal transmission slot and the outer loop learns the optimal transmit power under a given slot selection, enabling autonomous parameter selection according to the environment. The simulation results show that the proposed algorithm significantly reduces the packet collision rate compared to existing methods and improves both slot utilization and communication stability.