Date of Award

8-2013

Document Type

Thesis

Degree Name

Master of Science (MS)

Legacy Department

Computer Science

Advisor

Hallstrom, Jason O

Committee Member

Malloy , Brian A

Committee Member

Sorber , Jacob M

Abstract

In the recent past, there has been a phenomenal increase in monitoring the physical world using wireless sensor networks comprising tiny computing devices with integrated sensors. These devices are deployed in large numbers, often across large spaces and within hostile environments.
Wireless sensor networks are used to gather meaningful data and to enable important applications. They must satisfy some basic requirements. Specifially, they must be maintenance free, inexpensive, reliable, and scalable. The devices used in these networks are deployed in the hundreds to thousands and rely on battery power. It is expensive to change these batteries in such large networks, both in terms of battery cost and personnel time. The cost of an individual device plays an important role in the cost of a sensor network. Further, if these networks are deployed in safety critical contexts, we cannot risk having incorrect data or missing important data. Finally, sensor networks must be able to accommodate new devices and gracefully handle device failures.
This thesis describes a hardware/software solution for wireless sensor networks which is maintenance free, inexpensive, reliable, and scalable. In this thesis, I present a wireless sensing device which harvests solar energy and stores it in a Li-Ion battery. The device works on solar energy during the daytime and relies on battery power during the night. This addresses the problem of maintaining remote devices. The design is also focused on reducing component costs. The cost is low enough to discard the individual devices without significant concern. Finally, using these devices, I present a network protocol and reference implementation which makes data reception reliable, while supporting network scalability.

Share

COinS