Researchers use smart radios to advance cognitive network capability
Cognitive radios present an exciting new frontier for the world of wireless telecommunications. Virginia Tech’s Center for Wireless Telecommunications has received a three-year National Science Foundation (NSF) grant to extend its work in the field of cognitive radio to advance cognitive network capability.
Cognitive radios are “intelligent” cell phones or police radios that determine the best way to operate in any given situation. Instead of following a set of predefined protocols, as regular radios do, cognitive radios configure to their environment and their user’s needs.
“The new cognitive radios are similar to living creatures in that they are aware of their surroundings and understand their own and their user’s capabilities and the governing social constraints,” said Charles W. Bostian, Alumni Distinguished Professor of Electrical and Computer Engineering and director of the Center for Wireless Telecommunications.
A radio’s actions arise from a rational process that predicts probable consequences and remembers all of its failures and successes. “The radios are treated like animals that learn to evolve over time with their changing environment,” said electrical engineering Ph.D. candidate Tom Rondeau. "Basically, the cognitive engine is a brain that reads the radio’s 'meters' and turns the radio’s 'knobs' in order to get the desired outcome."
The grant will fund research that will, for the first time, allow the radios to share a distributed knowledge base to use for individual and collective reasoning and learning. "Every cognitive radio has a knowledge base it has learned over time. Now, our research makes it possible for all cognitive radios to be connected in a network where that information can be shared," said Bostian. "You will now have a whole network of radios working and sharing information in what we characterize as a cognitive network.
"We approached this network like you would approach a community of animals. First, we teach the radios to seek others of their own kind and recognize their environment, thus creating a network. Then the Center for Wireless Telecommunications team teaches the different cognitive radios to work together."
The Center for Wireless Telecommunications team will be conducting the first large scale tests of cognition in a well-defined wireless network environment. The experiments will provide quantitative answers to such very practical questions as, “Can cognitive techniques allow ‘Wi-Fi like’ services in locally unoccupied TV channels?” Meaning ‘Wi-Fi like’ services could eventually be available in once previously unusable TV spectrums.
The research will have many important outcomes, said Bostian. First, a cognitive engine will be fully implemented in any wireless network, making it a cognitive one. Cognitive engines will be implemented into the widely available GNU radio, which is a low-cost software-defined radio developed specifically for experimentation in projects like this one. Finally, there will be a practical assessment of the performance advantages of cognitive wireless networks.
The grant was awarded through the NSF’s NetS Programmable Wireless Information Networks Program. Bostian is principal investigator. Co- principal investigators are Associate Professor Michael Hsiao and Assistant Professor Allen MacKenzie in the Bradley Department of Department of Electrical and Computer Engineering, and Sheryl Ball associate professor of economics in the College of Science. Additional sponsorship is provided by Anritsu Company, which will provide much needed testing and measurement equipment. The Virginia Tech researchers will use the equipment to prototype key cognitive radio functions and measure the performance of the resulting designs.