I worked on algorithms related to large-scale wireless sensor networks. Researchers are now envisioning sensor networks scaling countries, oceans, the planet, and even further. To be able to study and analyze networks of that scale, efficient algorithms have to be developed in order to process that level of complexity.

By using graph coloring algorithms on a Random Geometric Graph, I was able to develop an algorithm which can efficiently segment a graph into independent, dominating sets which can be used as “backbones” in a network. The idea is that, at any given time, only one backbone would need to be active which would allow for significant power savings without a significant decrease in network coverage.

A more complete description of the algorithms involved including an interactive simulation is available here. A write-up describing the paper and my solution can be found here: Wireless Sensor Networks (796)