Date of Award

12-2009

Document Type

Thesis

Degree Name

Master of Science (MS)

Legacy Department

Computer Science

Advisor

Hallstrom, Jason O

Committee Member

McGregor , John D

Committee Member

Malloy , Brian A

Abstract

To meet the increasing demand for monitoring of the physical world, there has been an increase in the development of wireless sensor network applications. The TinyOS platform has emerged as a de facto standard for developing these applications. The platform offers a number of advantages, with its support for concurrency, power-efficient operation, and resource-constrained hardware chief among them. However, the benefits come at a price. Even without the TinyOS platform, the inherent parallel and distributed nature of these applications makes it difficult for developers to reason about program behavior. Further, the TinyOS programming model adopts
asynchronous, split-phase execution semantics. Developers must explicitly manage program control state across event-handlers, components, and devices. This makes the design, debugging, and comprehension of these programs even more difficult.
In this work, we describe an animation framework for TinyOS programs, designed to enhance
the comprehension of their runtime behavior. The framework enables application developers
to specify, in the form of an XML configuration file, the runtime elements to be captured within a
given system and the manner in which those elements should be displayed. The resulting visualization presents an animated play-back sequence of the events that occurred during execution. The framework also provides a visual representation that connects causally-related events in a distributed network. We describe the design and implementation of the animation framework and present an analysis of the runtime overhead it introduces.

Share

COinS