Monday, November 10, 2014
12:00 PM
Computer Studies Building, Room 209
Shahriar Nirjon
HP Labs
Staying in Touch: Seamless Connectivity and End-device Localization in Mobile Environments
Being able to “staying in touch” in a mobile environment has two fundamental challenges. First, the device needs to stay connected to the Internet so that it can communicate with other devices, and second, it needs to know where it is so that it can access various location-based services. In this talk, I will describe my approach of addressing these two challenges of mobile computing: seamless mobile network connectivity and a GPS based indoor localization service.

First, I will introduce MultiNets, which is a modified Android operating system – having the capability of dynamically switching wireless network interfaces (3G and WiFi) on a mobile device. Network interface switching happens based on some predefined policies – such as saving energy, increasing throughput, and offloading data traffic. Based on our empirical observation on TCP sessions on a mobile device, we devise an algorithm to perform network interface switching without interrupting on-going TCPs. Evaluation of MultiNets in a real-world scenario demonstrates that it could save 33% more energy, and achieve near-optimal WiFi offloading and throughput, when compared to the state-of-the-art Android OS.

I will then describe my latest work on a GPS based indoor localization service. This work challenges the common belief that the GPS receivers do not work indoors. We propose the first system, named COIN-GPS, which is capable of obtaining a location fix inside single-storied buildings by directly using GPS signals and without requiring any extra infrastructural setup. This solution relies of hardware-software symbiosis where a high-gain directional antenna and signal processing algorithms are used to estimate the GPS coordinates with a median error of less than 10 m.

Bio: Dr. Shahriar Nirjon is a Research Scientist in the Networking and Mobility Lab at the Hewlett-Packard Labs in Palo Alto, CA. Shahriar is interested in building experimental systems that involve embedded sensors and mobile devices, connectivity, and mobile data analytics. Research challenges that he deals with include practical issues in physical world sensing, user-contexts and mobility, real-time issues, and resource constraints of the embedded platform.

Shahriar received his Ph.D. in Computer Science from the University of Virginia in 2014 under the supervision of Prof. John A. Stankovic. Shahriar has won a number of awards, including two Best Paper Awards – at the Mobile Systems, Applications and Services (MOBISYS 2014), and the Real-Time and Embedded Technology and Applications Symposium (RTAS 2012). Shahriar has worked as a Research Intern at Microsoft Research, Redmond, WA (Summer 2013) and at Deutsche Telekom Lab, Los Altos, CA (Summer 2010).