An Introduction on Machine Learning for Visible Light Communications

Visible light communication (VLC) is expected to act as an alternative candidate in next-generation wireless optical communication. To alleviate the pressure on the conventional radio frequency spectrum and realize the high-speed wireless communication, the challenge is that the complex VLC channel characteristics and the nonlinear response from inperfect optoelectronic devices, which will bring large attentuation and nonlinearity penalty. In addition to the limited performance of the conventional digital signal processing algorithms, machine learning techbology are commomly used in optical communication due to its strong ability in mapping the nonlinear relationship between input and output. In this report, we will summarize the latest progress on the application of machine learning algorithms and structures such as K-means, DBSCAN, FLANN, and DNN in VLC system. Our goal is to assist the readers in refining the motivation, structure , performance and cost of some typical machine learning techniques for future VLC system to tap into hitherto unexplored applications and services.


Professor Nan Chi is with School of Information Science and Engineering, Fudan University, China. She received the BS degree and PhD degree in electrical engineering from Beijing University of Posts and Telecommunications, China. She is the author or co-author of more than 300 papers and has been cited more than 6500 times. She has been awarded as The National Science Fund for Distinguished Young Scholars, the New Century Excellent Talents Awards from the Education Ministry of China, Shanghai Shu Guang scholarship. Her current research interests include advanced modulation format, optical packet/label switching, optical fiber communication and visible light communication. She is a fellow of the OSA.