In an earlier article title “cognitive radio: The idea of radio enviroment map (REM)” I provided an insight into the REM’s concept. It’s based on this discusssion that I briefly present a description of an architecture of radio enviroment map (REM)-enabled cognitive engines.
The system architecture of the REM-enabled CR model is illustrated below;
The system architecture of the REM-enabled CR model is illustrated below;
In this model, the REM consists of multidomain information. During operation, the cognitive radio (CR) observes its operational environment via sensor(s), and synthesizes necessary situation awareness (SA) of the current radio scenario by leveraging the sensing results and REM information.The CR reasoner then determines an appropriate utility function based on the policy and the goals, by considering the specific application and radio scenario. The utility function maps the current state of the CR, usually represented by an array of chosen metrics, to a value for indicating how close the state is toward the desired (or optimal) CR state. The most pertinent performance metric(s) should be taken into account and incorporated into a utility function to meet the CR’s goal fo this application and radio scenario. By leveraging experience and knowledge, the cognitive engine (CE) can choose the most efficient reasoning and learning method and make improved crosslayer adaptations (if necessary) subject to the constraints of regulation, policy, and radio equipment capability.
