It is now widely recognized that wireless communications systems don't exploit the whole available frequency band. The idea has naturally emerged to develop tools to better use the spectrum. Cognitive Radio (CR) is the concept that meets this challenge. The CR is a form of wireless communication in which a transmitter/receiver can detect intelligently communication channels that are in use and those that are not, and can move to unused channels. This optimizes the use of available radio frequency spectrum while minimizing interference with other users. CRs must have the ability to learn and adapt their wireless transmission according to the ambient radio environment. The application of Artificial Intelligence (AI) approaches in the CR is very promising because they are essential for the implementation of CR networks architecture. They must be able to coexist to make CR systems practical, which may cause interference to other users. To solve the problem of congestion, CR networks use Dynamic Spectrum Access (DSA). In order to deal with this problem, the idea of cooperation between users to detect and share spectrum without causing interferences is introduced. The authors found a large number of suggested works relating to spectrum access, those using Auctions, a large number of approaches use the Game theory, but those using Markov chains are fewer. However, some research has been done in this area using Multi Agent Systems (MAS).
