A New Approach to Learning System in Cognitive Radio Based on Elman Networks

Satellite transponders play a very important role in establishing a logical connection between the transmitting and receiving earth stations. In recent years, the Cognitive Radio (CR), a special class of Software Defined Radio (SDR), based transponders have attracted special attention due to their unusual working behavior. The modern transponder designs are getting smarter and smarter as they are not heavily relying on the specifications of the hardware components anymore. It adapts itself to the exposed environment by learning about its characteristics such as the electromagnetic spectrum being utilized. This paper is introducing a new approach of learning systems based on Elman Networks, which were especially proposed for the cognitive systems but unfortunately were not explored for realizing the CR transponders.