Research

CROWD Networks: Coexisting Radio and Optical Wireless Deployments

In this project, we analyze CROWD networks in the context of 5G. The majority of wireless capacity gains have come from cell densification. In CROWD networks, we envision a continuation of this trend with the deployment of ultra dense optical wireless small cells that coexist with conventional RF small cells. This work focuses on deployment strategies - including provisioning and implementation - as well as handover and allocation techniques for dynamic multi-cell/multi-user indoor environments. This work stemmed from concepts developed in my dissertation work with the LESA ERC and is the primary focus of our team's awarded NSF grant on Coexistence of Directional Communications within 5G: The Case for Visible Light Enhanced Small-Cells.

SDVLC: Software Defined Visible Light Communication

Software Defined Radio (SDR) has proven to be an effective and practical tool in the area of RF communication systems, essentially allowing flexible and rapid exploration of signal processing techniques for RF. The software defined concept can also be adapted to other physical communication media; we consider an optical channel. We demonstrate a Software Defined Visible Light Communication (SDVLC) solution implementing an illumination quality optical front end to adapt an SDR platform to the VLC channel. A “software defined” system implements the entire communications stack in a general purpose processor rather than specialized hardware. This enables rapid implementation of new testing scenarios in order to facilitate research in VLC. For example, a key research topic for VLC through LED lighting is modulation design – creating a light waveform that most effectively carries data under the constraints of the hardware and the environment. We utilize the open-source GNURadio SDR platform to enable implementation of new modulations designed specifically for VLC. This platform also facilitates development of system hardware. It uses a common physical interface that lets us easily swap in and test different prototype modules and components. This helps us isolate and understand the key performance bottlenecks of the hardware in order to direct our efforts to those areas. Some videos of the work can be seen here: SDVLC Setup, SDVLC Signal Blockage, SDVLC HetNet, SDVLC MIMO1, SDVLC MIMO2

CandLES: Communication and Lighting Environment Simulator

As research in the area of visible light communication has grown, we have found a need for common research platforms. The CandLES software toolkit is an open source software package for modeling and simulating indoor environments. The software is built in Matlab and consists of a code base and graphical interface for users or developers to generate environments and evaluate the illumination and communication performance. We have released a preliminary version of the software on github (here) and plan to continue adding features in the near future. If you are interested in collaborating and/or contributing to this project, please let me know!

Interference in IM/DD Optical Wireless Networks

When considering dense environments with many simultaneous wireless connections, interference analysis is a primary concern. Given the different constraints and operating points in VLC systems, interference models do not fit the conventional RF models. Specifically, we have shown that the presence of a dominant interfering VLC cell can cause the interference to follow the distribution of the interfering transmitted signal. This is contrary to the common assumption in RF systems where interference is modeled as Gaussian due to the combination of many interfering signals and central limit theorem. In addition, the average optical power constraint of VLC and IR systems does not directly indicate the signal variance as is the case with electrical power constraints in RF systems. We have shown that the relationship is modulation specific; but we have analyzed bounding conditions for interference variance based on the various constraints, along with the related bounds on signal to interference plus noise ratio.