Enhancing Robustness in Hybrid FSO/RF Systems: A Feedback-Free Approach Leveraging Deep Learning for Real-Time Power Allocation

The sixth-generation (6G) mobile network and the Industrial Internet of Things (IIoT) demand high-speed, low-latency communication in diverse environments. Hybrid free-space optical (FSO)/radio frequency (RF) systems offer a promising solution by combining the strengths of both technologies. However, maintaining robustness under dynamic atmospheric conditions remains a challenge. To tackle this issue, we propose a feedback-free hybrid FSO/RF system that analyzes atmospheric conditions from camera-captured images and utilizes a deep learning network for real-time power allocation. By dynamically optimizing link utilization, it mitigates obstacles and weather fluctuations while eliminating feedback overhead, enhancing efficiency. Simulations show that our feedback-free system maintains superior bit error rate (BER) and outage performance while offering greater stability and adaptability to dynamic environments. Compared to state-of-the-art methods, it achieves more reliable performance under varying conditions. By leveraging camera-based perception instead of feedback links, it optimally adjusts FSO/RF utilization, providing a practical, scalable framework for next-generation IoT applications and ensuring reliable communication under varying conditions.