Deep Learning based RS-OFDM scheme considering mobility environment for Optical Camera Communication system.

Rolling-Shutter Orthogonal Frequency-Division Multiplexing was presented in IEEE 802. 15.7a Task Group, which proposed new standard for Optical Camera Communication (OCC). OFDM is a digital communication modulation method used in high rate wireless communication to reduce the channel distortion occurred by multi-path effect. Based on this advantage, OFDM has been extensively employed in optical wireless communication for many applications such as the IoT, V2V, V2X, and Smart Factory. However, the movement effect is a significant impact for OWC systems since it results in an optical channel variance in processing time, which decreases performance of OCC system. In the movement environment, to detect the starting f:ame, conventional shows disadvantages, then we have to find new algorithm to increase OCC performance. In this study, the deep learning algorithm was employed instead of Van De Beek algorithm, which reveals disadvantages compared this approach. In a 2 m/s velocity environment, our technique enables the OCC system to communicate over long range (up to 18 m) with low error.