Ph.D Defense by David Lillethun
Title: ssIoTa: a System Software Framework for the Internet of ThingsDavid LillethunPh.D. CandidateSchool of Computer ScienceGeorgia Institute of TechnologyDate: Tuesday, March 24, 2015Time: 12:00 PM (Noon) - 2:00 PMLocation: KACB 3402Committee:Dr. Kishore Ramachandran (Advisor, School of Computer Science, Georgia Institute of Technology)Dr. Karsten Schwan (School of Computer Science, Georgia Institute of Technology)Dr. Mustaque Ahamad (School of Computer Science, Georgia Institute of Technology)Dr. Santosh Pande (School of Computer Science, Georgia Institute of Technology)Dr. Flavio Bonomi (IoXWorks, Inc.)Abstract:The Internet of Things (IoT) is an emerging technology fueled by the proliferation of sensing devices. IoT is characterized by massive scale, wide-area distribution, machine-to-machine (M2M) interactions, heterogeneous devices, and hierarchical structure. While much of the work in IoT thus far has focused on connectivity and addressing, sensor device constraints, interoperability of disparate devices and protocols, and data access and querying/filtering on simple values, the application space becomes truly interesting when intelligent analysis algorithms may be applied to live streaming sensor data. Developers have thus far produced live streaming analysis applications as turnkey solutions from the ground up, due to the lack of systems support. We present a framework for designing system support for such live streaming analysis in the Internet of Things, and our system implementation of that framework, ssIoTa.When the Internet of Things is fully realized, it will consist of billions of sensors producing a massive amount of data, and widely distributed across the globe. Old models of bringing all the data to a data center for later analysis are no longer suitable. Bringing computation to the edge of the network has been proposed to reduce latency and the amount of data that must be transmitted across the network core. To support this model, we have designed ssIoTa to work across multiple sites, allowing it to support computation near the edge while still enabling analysis across sensors in different locations. Specifically, we present a qualitative analysis of distributing a geospatially indexed sensor registry by adapting known techniques, including distributed hash tables (DHTs). We also present simulation results for DistAl, our method and algorithm for distributing analysis computation among geographically distributed computational resources.
- Workflow Status: Published
- Created By: Tatianna Richardson
- Created: 03/12/2015
- Modified By: Fletcher Moore
- Modified: 10/07/2016