ESR 5

Ömer Dalgic

Project: Resource-constrained VLC

Host: Scuola Universitaria Professionale della Svizzera Italiana SUPSI (Switzerland)

PhD: TU Delft (Netherlands)

Supervisor: Dr. D. Puccinelli

I was born in Batman/Turkey in 1993. A little later after my birth, my family moved to Istanbul which is the most populated city in Turkey. My childhood had passed within a very crowded group of friends. In Turkey, playing football in the street was the most popular game during my childhood and I remembered that we were playing until late in the evenings. Apart from Math and Physics, I began to be interested in literature and politics from the beginning of my high school term. I could state that two of my favorite books are 'The Trial (Der Prozess)' from F. Kafka and 'Catcher in the Rye' from J.D. Salinger. In Turkish Literature, Nazım Hikmet and Yusuf Atılgan are my favorite literary personalities. On the one hand, I decided to move into engineering fields for my education because I preferred interest with mathematical modelling. My Bachelor and Master education is more related to Telecommunication Engineering. In the last year of my master’s education, I feel more fascinated with wireless data and energy transfer concepts. Thanks to ENLIGHT'EM project, I could improve myself in this field while having chance to study in Embedded Systems with plentiful collaboration opportunities from both academy and industry. For my daily life, I like to take long walks, have a chat with my relatives, and interest in literature.

My research will focus on communication between light emitting diodes (LEDs) as transmitters and smartphone cameras as receivers. My focus on LED-to-smartphone camera communication is motivated by the pervasiveness of smartphones in today’s world. We plan to develop an adaptive link layer that enables multiple types of LEDs to communicate with smartphone cameras. I will investigate the properties of visible light communication (VLC) channel in the context of LED-to-smartphone camera communication. Moreover, I plan to design energy-aware communication strategies to facilitate LED-to-smartphone camera communication in LED-to-camera applications, namely IoT applications that leverage VLC-based augmented reality (VbAR).

VbAR applications are subset of internet of things (IoT) applications for LED-to-smartphone camera communication, and it provides getting information from everyday objects thanks to LED on them. This can be achieved by letting light sources on everyday objects blink at high speed to encode data, and by having smartphone cameras decode the received signal to extract data from the objects. For instance, a smartphone could display a Wi-Fi Router’s SSID and password with the help of a nearby LED. However, this approach results in binary transmission and is limited by the relatively low frame rate of today’s smartphone cameras. It is fairly typical for today’s phone cameras can achieve 240 fps. Moreover, the rolling shutter image acquisition method, which may be employed with most CMOS camera sensors, can provide a higher data rate compared to the camera frame rate by reading data row-by-row instead of the entire image at once. This method may allow transmission data rates that are significantly higher than the camera frame rate but it is applicable for relatively short distances comparing to binary transmission in VbAR applications.

I aim to build a robust mathematical model and to implement a system prototype to demonstrate the coexistence of VbAR and rolling shutter on Android phone applications. My research begins with mathematical modelling to describe the effect of various parameters (such as LED size, camera fps, etc.) on the balance between reliability and throughput and on whether rolling shutter is preferable to binary transmission. My model will allow users to enter the parameters and investigate output variables such as the range, the achievable data rate, and the BER for each transmission type. Finally, we will design a system prototype using commercial off-the-shelf hardware and custom software for an experimental evaluation.

A key contribution of this line work is to provide flexibility for VbAR applications. In our work, we aim to investigate two research problems in this space:

1) We plan to propose a theoretical model to investigate the tradeoffs between leveraging the rolling shutter effect and settling for binary transmission. There are many types of LEDs and cameras, and they have different requirements for both rolling shutter and binary transmission. The model parameters will include LED size, smartphone-to-LED distance, communication channel properties, and camera frame rate. The model will also naturally account for smartphone mobility. Because user could get closer to LED or move away from LED to get information, this situation automatically brings the management of mobility for VbAR. The goal of the model is to determine whether an AR application should leverage rolling shutter as opposed to binary transmission depending on the values of the parameters. While we intuitively expect that the rolling shutter effect will be more beneficial for shorter distances, at the time of writing there exist no quantitative guidelines for AR application designers to apply.

2) In addition to our analytical modeling work, we also plan to design and implement a proof-of-concept system. We will develop a design that provides dynamic transitions between binary transmission and rolling shutter based on a subset of the parameters investigated by the model (only those that can be measured in real time).