Raphael M. Guedes,  José Ferreira de Rezende. 

Cognitive radios have emerged as a promising technology in spectrumsharing among secondary users (SUs). Due to difficulties faced by this solution,many proposals consider the existence of a channel, called control channel.Through this channel SUs can exchange coordination messages. However,due to a number of factors, it is unlikely that a single control channel is simultaneouslyavailable to all SUs. One solution is to use a multi-channel approach,where nodes can hop from channel to channel in order to find other users toexchange information. Thus, there must be a channel sequence, which most ofusers can communicate. This paper discusses the use of Reinforcement Learningas a possible tool for discovering those channel sequences.

http://www.lbd.dcc.ufmg.br/colecoes/sbrc/2011/006.pdf

Processando consulta.