Interference Suppression in Massive MIMO VLC Systems
The focus of this dissertation is on the development and evaluation of methods and principles to mitigate interference in multiuser visible light communication (VLC) systems using several transmitters. All components of such a massive multiple-input multiple-output (MIMO) system are considered and transformed into a communication system model, while also paying particular attention to the hardware requirements of different modulation schemes. By analyzing all steps in the communication process, the inter-channel interference between users is identified as the most critical aspect. Several methods of suppressing this kind of interference, i.e. to split the MIMO channel into parallel single channels, are discussed, and a novel active LCD-based interference suppression principle at the receiver side is introduced as main aspect of this work. This technique enables a dynamic adaption of the physical channel: compared to solely software-based or static approaches, the LCD interference suppression filter achieves adaptive channel separation without altering the characteristics of the transmitter lights. This is especially advantageous in dual-use scenarios with illumination requirements. Additionally, external interferers, like natural light or transmitter light sources of neighboring cells in a multicell setting, can also be suppressed without requiring any control over them. Each user's LCD filter is placed in front of the corresponding photodetector and configured in such a way that only light from desired transmitters can reach the detector by setting only the appropriate pixels to transparent, while light from unwanted transmitters remains blocked. The effectiveness of this method is tested and benchmarked against zero-forcing (ZF) precoding in different scenarios and applications by numerical simulations and also verified experimentally in a large MIMO VLC testbed created specifically for this purpose.