Personal computing is currently undergoing a major revolution: the growth of mobile computing. The number of mobile devices is on a rapidly increasing trajectory, including millions of Bluetooth and WiFi enabled devices such as cameras, PDA, cell-phones, and music players. Although each device has a specific purpose, they are all used to create, store, and share vasts amounts of personal data. Along with this growth comes a data management nightmare: systems frustrate users with the lack of availability of data, complications in transferring data from one device to another, and limited battery lifetime.
This CAREER research is designing and building mobile systems that address power management, variable connectivity, and usability concerns. A variety of techniques are employed, from unique combinations of hardware and software, to innovative techniques in measurement in modeling. The research includes building real systems, in order to demonstrate the benefits of these approaches to end-users.
The first issue to be addressed is power management in always-on devices. Through a hierarchy of connected low-power embedded subsystems, the mobile devices efficiently enter low-power modes to provide data transfer opportunities. Second, because many devices frequently disconnect from their peers, it is necessary to use third party devices to transfer data updates. Using techniques from Distributed File Systems, as well as Disruption Tolerant Networking, the research is building systems that take advantage of mobility to transfer data between normally disconnected devices. Third, efficiently transferring data using mobile devices requires efficient operating support for storing data.
This CAREER grant funds these research activities, as well as ongoing collaborations with four other colleges in the Western Massachusetts area. Through interactions with two highly-ranked all-women's colleges, the PI is encouraging more female students to pursue graduate school in Computer Science.
