Amy Bell is investigating applying wavelets to signal compression for applications such as multimedia Internet.
Sometimes it is imperative that a communication signal reaches its destination --regardless of interference or other unknown channel effects. In these situations, standard modulation schemes have their limitations.
A novel technique under investigation by Amy Bell, a specialist in digital signal processing, offers a unique advantage by sending the signal at multiple rates simultaneously through the channel. "The point is that if you do not know what part of the channel is open, you can send multiple copies. Whichever part of the channel is open, one of your copies gets through. You don't need to know what your channel looks like."
The technique, called fractal modulation, relies on wavelets, which are a mathematical tool that provides an alternative to classical Fourier signal analysis. "Wavelets are better able to represent time-varying signal behavior," she explained. "Unlike the Fourier transform's decomposition of a signal into its frequency components, the wavelet transform provides a view of the signal from the time and scale perspectives."
Bell is also working on extending wavelets into multiwavelets and investigating their application for efficient compression. "One of the important applications for this work is the Internet. We're all interested in sending video, but it takes a huge amount of time to send-and even then the video quality is questionable. We are looking at new methods for compressing data so that the reconstruction is timely and of sufficient quality."
Bell first worked in wavelets research while earning her Ph.D. at the University of Michigan. Her doctoral work investigated applying wavelets to a classic unsolved problem in signal reconstruction - the phase retrieval problem.
"We came up with some solutions," she said. "Like many engineering solutions, they are based on assumptions, so for certain classes of signals, our approach offers unique advantages."
Bell enjoys her theoretical research, but is also interested in applied, industrial projects. One project, on which she is collaborating with other faculty in the department, concerns improved location accuracy in cellular networks. "This is directly related to the FCC requirement that cellular service providers offer mobile E-911 service," she said.
A balance between applied industrial projects and more basic research is important to Bell. "It helps you keep things in perspective. With the theoretical projects, you realize that advances can take years - even decades. With industrial projects, you realize the urgent and ongoing needs faced by today's engineers."
Before returning to graduate school for her doctorate, Bell worked for five years as an engineer. Her first position, with what was then Westinghouse Corp., was in manufacturing engineering with F-16 fighter planes. She then moved into simulation modeling for facility design. She also spent several years with Systems Modeling Corporation designing and developing discrete event simulation models for various manufacturing environments.
"So many aspects of industry are fundamental and independent of the individual problems," she commented. "I find that my industrial experience provides great examples in class and is invaluable when I mentor students who will soon enter industrial positions."
Bell is particularly interested in continually improving her teaching. During her graduate schooling she earned two awards for teaching excellence, and she meets regularly with a faculty study group connected with Virginia Tech's Center for Excellence in Undergraduate Teaching.
"Some things that I've tried in the classroom work better than others," she said. Recently she has had success with extra credit quizzes during the first 15 minutes of lecture to get students working in small groups on a warm-up activity for that lecture's subject material.
Bell is also concerned with the relatively low numbers of women in electrical and computer engineering. "Sometimes the academic environment discourages women students." Since coming to Tech in 1997, she has organized get-togethers for women students to meet one another, develop a support network, and share information regarding classes, scholarships, and employment opportunities.
"We need to provide an equitable environment for women
in engineering because it is simply absurd to recruit from only
half the potential pool."