Hybrid optical amplifiers (HOAs) are essential for improving signal strength and extending transmission reach in high-capacity optical communication systems. However, conventional amplifier gain control methods, such as static configurations and heuristic adjustments, often fail to adapt efficiently to dynamic traffic conditions and nonlinear impairments in emerging 6G-ready elastic optical networks (EONs). This study proposes an AI-driven optimization framework for hybrid optical amplifiers aimed at improving transmission performance and network reliability. The proposed approach integrates supervised learning for Quality of Transmission (QoT) prediction with reinforcement learning-based gain control to dynamically adjust amplifier parameters. An OptiSystem-based simulation model was developed to represent a realistic optical transmission system consisting of Raman amplification and Erbium-Doped Fiber Amplifiers (EDFAs). System performance was evaluated using key parameters including Optical Signal-to-Noise Ratio (OSNR) and Bit Error Rate (BER) under different optimization strategies. Simulation results demonstrate that the proposed AI-driven method significantly outperforms conventional techniques. The average OSNR improved from 17.8dB in static gain configuration to 22.4dB, while BER decreased from 3.2 × 10⁻³ to 4.8 × 10⁻⁴. Compared with heuristic optimization, the AI-based approach achieved more consistent signal quality through adaptive gain adjustments based on real-time network conditions. These findings indicate that intelligent amplifier optimization can significantly enhance signal integrity and transmission efficiency in elastic optical networks, providing a scalable solution for autonomous optical network management in future 6G communication infrastructures.
Publication Date: 2026-06-15