exit bar




Special Report
New Faculty Members

 April 1999




When a Best Effort
Isn't Good Enough


For Luiz DaSilva, a best effort is not good enough. "Traditionally, computer networks have worked in a best-effort fashion. They provide you with the best available connection at the time, but offer no guarantee on the quality," he explained. "A prime example of this is the Internet."

A specialist in computer telecommunications networks, DaSilva is working to improve integrated networks. "We are accustomed to good quality from our telephone services, now we'd like to integrate data, voice, multimedia, etc., in the same network with some quality guarantees," he said.

Some organizations have successfully deployed guaranteed-quality networks, such as Virginia Tech's net.work.virginia, which was established to bring broadband service to educational organizations across the state. The high-capacity network uses Asynchronous Transfer Mode (ATM) to deliver simultaneous transmission of fully interactive voice, data, and video services.

"Integrating these services into a single network brings many advantages such as a single interface for the user, economies of scale in infrastructure costs, and maintenance," he explained. "The challenge is to provide differentiated levels of service that guarantee the appropriate quality to each user and application while making efficient use of the resources available. In other words, how do you deal with times of congestion in the network - guaranteeing that the inevitable degradation in quality of service happens in a graceful manner?"

In addition to quality of service, DaSilva is exploring issues involving bringing very-high-speed connections to end users. "In order for users to be able to take full advantage of broadband services, it's crucial that we address the last-mile problem," he said. "Several alternatives, using wired and wireless media, are being investigated at present, such as cable modems and asymmetrical digital subscriber lines (ADSL). One of the major initiatives in this area, led by Virginia Tech's Center for Wireless Telecommunications (CWT), is the development of high-speed wireless Internet access using the LMDS spectrum."

This past fall, DaSilva joined the Virginia Tech faculty as an assistant professor in Northern Virginia and a member of the college's new Alexandria Research Institute (ARI). The institute focuses on the development and deployment of information technology.

"This is an ideal location for networks education and research," he said. "The Washington, D.C. area is one of the world's major political and communication centers. The proximity to information technology industry leaders and major government research means that we are able to interact more closely with sponsors and enhance our collaborative efforts," he said.

Drawing from his experience working at IBM Brazil, DaSilva is very comfortable interacting with technology firms. Before going back to school for his Ph.D., DaSilva spent six years with IBM Brazil, most recently as a product specialist/brand manager. His responsibilities included introducing the IBM AS/400 line of computers into the Brazilian market, as well as technical support and marketing planning.

He was also responsible for training IBM Brazil employees in the AS/400 line. "That was part of what motivated me to want to teach. It gave me exposure to teaching to an audience of professionals, much like those at the Northern Virginia Center," he said.

"My experience at IBM gave me a good idea of how the industry works," he said. "I understand corporate structure, the mechanics of development, and how products get to market."

"However, I wanted to do more research and teaching." So he returned to the University of Kansas to earn a Ph.D., where he worked at the Information and Telecommunication Technology Center on performance measurement and analysis for a wide-area ATM network testbed. During that time, he also participated in industry-sponsored research on pricing for commercial networks with multiple service classes.

"It is important for engineering professors to understand both their technical field and the industries involved. The time between pure research and product development in this field has become so condensed that researchers, developers, teachers, and students all need to speak the same language."


The Bradley Department
of Electrical and Computer Engineering
Virginia Tech

Last Updated, July 10, 1999
Questions or comments about the content: eqb@rightwordonline.com
Technical questions or comments: webmaster@birch.ee.vt.edu