Nathan Cooprider's Research
Nathan Cooprider's Research
Graduate Research
Developing software for embedded systems is an exercise in resource
management and tradeoffs. Elements such as stack overflow, ram usage, code
size, and duty cycle all become critical. However, minimizing one resource
often results in sacrificing another. The exact balance desired for a
system is often application specific. I research tools for adjusting the
resource use of embedded systems and tools for evaluating tradeoffs.
My graduate advisor is Dr. John
Regehr. He heads the Software Tools for Sensor
Networks at the University of Utah.
Projects
-
Safe-TinyOS - It is an experimental version of TinyOS 2
that uses the Deputy compiler to enforce type and memory safety at run
time
- CIL Tools for TinyOS - a collection of binaries,
scripts, and instructions designed to make getting started with our tools
easier
- CComp - an extension to cXprop
which takes information learned by the analysis and uses it to compress
global variables into a minimum number of bits to result in RAM savings
- cXprop - a static analysis tool for C programs. It
was originally started as a simple extension to CIL, but it grew to be
substantially more
- ASSHes -
Application and System Specific Heuristics
I interact with the Flux Research
Group a little. They run Emulab,
among other things. Two parts I find particularly interesting are their mobile
robotic WSN and their wireless network.
I also have a list of my publications,
presentations, and awards.
Undergraduate Research
A picture may be worth a thousand words, but sometimes if it tries to say
that much then much of the information gets lost. This can be seen with
graphs and data plots. Many people have difficulty understanding 3D
graphs, and often people struggle with 2D graphs. However, most data comes
in dimensions higher than three. The research I performed looks at how to
effectively display data of more than three dimensions, or data in
hyperspace.
The
Hyperspace Research Group
at BYU is the oldest research group at BYU. This group studies meaningful
representations of multivariate data. Common (and only partial) solutions
to this problem include parallel axes, icons, star plots and 4D cartesian
rendering. This group is headed by
Dr. Robert Burton.
My particular work with Dr. Burton and his Hyperspace Research group
involved working with Dr. Eser Kandogan's multivariate display technique
called Star Coordinates. My work expanded his model from two to three
dimensions and thus made the system more intuitive for the user. I
presented my proposal for this project in the Spring Research Conference
at BYU. Here are the results of our
research.
Useful Internet Research Tools
Conferences, Journals and Workshops
Fellowships