Air Strike Task-Oriented Intelligent Communication System Based on Diffuse Reflection Visible Light

As an important technology for 6G, Visible Light Communication (VLC) is highly suitable for internal cabin communication amid the current trend of cable-free and modular communication. This paper proposes a VLC system implementing "OOK + Manchester coding"based on deep learning. The system combines two neural networks to extract signal features and adopts a classification method for signal demodulation, adapting to the characteristics of high occlusion and non-line-of-sight (NLOS) paths in the missile interior. The "train-first-then-use"characteristic of artificial intelligence aligns well with the task features of air-to-ground missiles - long mounted flight time and short autonomous flight time. It can utilize the pre-launch time to complete data training, thereby supporting in-missile signal transmission after launch. Practical tests have shown that a communication rate of 1 Mbps can be achieved in a 1.2-meter-long