Thesis Examination Committee

Prof Mordecai Jay GOLIN, CSE/HKUST (Chairperson)
Prof Amine BERMAK, ECE/HKUST (Thesis Supervisor)
Prof Shahriar MIRABBASI, Department of Electrical and Computer Engineering, University of British Columbia (External Examiner)
Prof Hoi Sing KWOK, ECE/HKUST
Prof Levent YOBAS, ECE/HKUST
Prof Brahim BENSAOU, CSE/HKUST

Abstract

Recently, visible light communication (VLC) has gained much attention not only for its utilization in indoor applications, but also it is paving the way towards outdoor applications related to intelligent transportation systems (ITS). Much work has been done to increase the data rate, however very few works found in literature deal with increasing the robustness and communication distance of a VLC link. In this thesis, we propose to address these challenges in four phases.  

In the first phase, we propose a polarization based transceiver methodology to mitigate channel interference thereby increasing the robustness of the optical link. Experimental results demonstrate a 32.6% more robust VLC link compared to the conventional transceiver under severe optical interference.

In the next phase, an adaptive threshold under-sampled technique has been proposed that enables the receiver to sample at slower shutter speeds thereby increasing the communication distance of the link. We extend this work to propose a hybrid phase-frequency modulation to establish a record 160 m long communication link, making it the longest communication distance achieved till date using visible light.

In the third phase, we propose channel modeling of a camera communication system using perspective projection. The work is extended to estimate the channel length of a MIMO camera communication system, which is a critical parameter specifically for distance-critical applications.  

In the last phase, we use the channel model to predict the behavior of the system in a real-world scenario specifically in ITS applications. Vehicle vibration analysis was analyzed to quantify the extent of vibrations in a real-world environment. Based upon the simulations, we propose a methodology to detect and track a moving transmitter, while maintaining a robust communication link between the transmitter and the receiver. The proposed methodology caters for the worst case vehicle vibrations, validated through experimental results.