Veljko Pejovic - Research
General Interests
Wireless networks, mobile sensing systems, resource-efficient systems, machine learning, technologies for development, technology and society.
Research Statement
Communication technologies need to be adaptable to the context in which they operate. The context is described in part by physical peculiarities of the surroundings, by which wireless transmission is directly influenced. The context also includes social environment, as this environment both affects and is affected by communication technologies. As a networking researcher, I keep these dual components in mind, so that communication technologies I develop can both efficiently utilize physical resources and result in positive societal change.
In my PhD research I was focused on bridging the digital divide and providing Internet connectivity to rural developing regions. I visited rural area networks and investigated the use of existing technologies and issues that hamper Internet proliferation. On-site experience helped me discover causal relationships between the technology and the society and set guidelines for my efforts in designing networking solutions specifically tailored for rural area context. Consequently, I concentrated on key problems in rural wireless networking: I developed protocols that provide resource-efficient long distance wireless connectivity through software-defined radios over white spaces; furthermore, I designed solutions that harness the existing locality of online interaction to facilitate local file sharing and enable free voice communication within a village, crucial for closely knit rural communities in developing regions.
My research on wireless networks for rural areas taught me how important it is to have more than a single perspective when solving a complex real-world problem. Thus, I decided to extend my interests beyond just computer science. During my academic career I took courses and established collaborations with researchers from diverse departments: sociology, film and media studies, psychology, and electrical engineering.
For my postdoctoral research I am very pleased to be working on a large interdisciplinary project with a goal that I believe in. UBhave project investigates the power and challenges of using mobile phones and social networking for Digital Behaviour Change Interventions (DBCIs). In my research I harness smartphones to extract information about human actions and their environment. Then, using machine learning models, I devise relationships between different aspects of context and human behaviour. Such models are later used, for example, to improve user-device interaction or to deliver DBCIs at the right moment. My work consists of both theoretical investigation of machine learning models that can be employed for successful behaviour modelling, as well as of designing and building real-world mobile systems. My duties also include experiment design, mobile application deployment and data gathering and analysis.