딥러닝기술을 활용한 무선광통신시스템 신호생성에 관한 연구

Abstract

Many state-of-the-art technologies, such as Cloud Computing, AI(Artificial Intelligent), IoT(Internet of Things), and VR(Virtual Reality), have been developed rapidly with communication systems. Out of various communication systems, OWC(optical wireless communication) having advantages in viewpoints of high-frequency band, security of information, and free allocation frequency, which makes at a promising a next-generation communication. However, the performance of OWC suffered from deteriorating channel effects. Thus, appropriate signal formats for OWC channels are needed. In this thesis, I propose a DL-based signal format generation/recommendation system for OWC links. The system incorporates a physical optical channel model and deep learning neural network for generating signal sets for binary OWC system. It is simulated by a computer with assumed physical conditions for OWC. In this study, I specifically focus on generating signal sets for a parallel LED array to decide binary symbols (HIGH and LOW) for binary communication by classifying the recovered image pattern at the charge-coupled device(CCD).