Reinforcement Learning–Based Hybrid Optical–RF Framework for Ultra Reliable Low Latency Communication (URLLC) Networks

This article proposes a deep reinforcement learning-based hybrid optical–RF communication framework for Ultra-Reliable Low-Latency Communication (URLLC) networks under dynamic wireless and atmospheric channel conditions. The proposed system integrates OptiSystem-based optical channel modeling, MATLAB R2019b Simulink RF subsystem implementation, and a Deep Q-Network (DQN)-based adaptive transmission controller for intelligent hybrid link selection. The adaptive transmission problem is formulated as a Markov Decision Process (MDP), enabling dynamic selection among RF-only, FSO-only, and cooperative hybrid transmission modes according to real-time RF/FSO channel quality, queue status, and URLLC latency requirements. Performance evaluation is conducted using an integrated OptiSystem–MATLAB co-simulation framework with 100 Monte Carlo simulation runs under varying atmospheric attenuation and RF fading conditions. In addition, a laboratory-scale experimental hybrid optical–RF testbed operating at 1550 nm and 28 GHz is implemented to validate the practical feasibility and real-time operational performance of the proposed framework under realistic operating conditions. Simulation results demonstrate that the proposed DRL-based framework achieves an average latency reduction of approximately 28–35% compared to conventional RF-only transmission while improving throughput performance under favorable optical channel conditions. The adaptive hybrid framework maintains reliability approaching URLLC-grade communication performance across varying channel environments through intelligent adaptive transmission control. Experimental validation demonstrates close agreement with simulation predictions, with average deviation remaining below 6% across evaluated latency, reliability, and throughput metrics. The obtained results demonstrate that the proposed reinforcement learning-driven hybrid optical–RF communication framework can approach stringent URLLC latency and reliability targets under varying wireless and atmospheric channel conditions, supporting its application in mission-critical 5G and emerging 6G communication systems.