Learning RAT Selection Game in 5G Heterogeneous Networks

In this letter, we consider the problem of RAT selection in a heterogeneous cellular network where each user chooses among multiple access technologies. The competition of the users is modeled as an incomplete information game where players are not aware other players' actions. A RAT selection strategy is proposed and is shown that it converges to the Nash equilibrium of the game. Moreover, we consider a more practical case in which the users have no information about other users. For this case, we propose a Q-learning-based algorithm, which converges to the Nash equilibrium of the first problem.