BIN LI

Researchers have developed a cognitive radio that uses algorithms modeling human learning to automatically optimize communications across wireless channels in unanticipated situations. The radio formalism is expected to be of particular use in emergency response situations and in military applications. Commercial applications may also benefit from the dynamic utilization of the spectrum.

The team, led by Charles Bostian, has implemented the formalism in a working broadband wireless cognitive radio testbed. The system uses smart transmitters and smart receivers embedded with a distributed cognitive engine based on a multi-tiered architecture. A broadband channel sounder senses and a Wireless Channel Genetic Algorithm (WCGA) models wireless channels at the waveform or symbol level. A Wireless System Genetic Algorithm (WSGA) performs on-the-fly evolution of the radio’s operational parameters, while a Cognitive System Monitor (CSM) handles the cognitive functions, short- and long-term memory and control.

The team is testing and improving the system and plans to extend the techniques to build a network testbed and apply the genetic algorithm approach to the MAC/Data Link Layer.