Fiber-wireless integrated systems with ultra-high capacity low-latency and high-reliability

Wireless communication;Fiber-wireless convergence;Digital signal processing

Lu, Feng ; Electrical and Computer Engineering Chang, Gee-Kung Weitnauer, Mary Ann Ma, Xiaoli Barry, John R. Ramachandran, Umakishore

University:Georgia Institute of Technology

Department:Electrical and Computer Engineering

关键词: Wireless communication;    Fiber-wireless convergence;    Digital signal processing;   

Others  :  https://smartech.gatech.edu/bitstream/1853/61124/1/LU-DISSERTATION-2018.pdf

美国|英语

来源: SMARTech Repository

【 摘 要 】

An enhanced fiber-wireless integrated system is designed using advanced system design and digital signal processing (DSP) techniques. The capacity and reliability are improved with reduced latency in the converged network. For the optical section, the peak-to-average power ratio is reduced using advanced DSP and the network architecture is optimized with novel light sources in analog links. In addition, the adaptive parameter tunning is designed for digital link enhancement. All these schemes provide enhanced capacity and reliability with higher efficiency. For millimeter-wave fiber-wireless convergence, advanced waveforms including orthogonal chirp division multiplexing, generalized frequency division multiplexing and power-division non-orthogonal multiple access are proposed and implemented, to support applications emphasizing capacity, reliability or device density. With verifications using a millimeter-wave fiber-wireless converged setup, these waveforms demonstrate better performance in various channel conditions. Finally, for the all-spectrum fiber-wireless system, the converged fiber, millimeter-wave, free-space optics and visible light communication system is designed aiming to enhance the network capacity, density and reliability by utilizing extra spectral resources. The novel system design and DSP techniques presented in the dissertation provide a reference for future wireless system standardization and network implementation. Further waveform design for all-spectrum system and machine learning based optimization are discussed for future research.