On the implementation of channel selection for LTE in unlicensed bands using Q-learning and game theory algorithms

The use of Long Term Evolution (LTE) in the unlicensed 5 GHz band, referred to as LTE-U or Licensed Assisted Access (LAA), is a promising enhancement to increase the capacity of LTE networks and meet the requirements of future systems. This paper analyses the use of fully distributed channel selection mechanisms for facilitating the coexistence among different LTE-U and/or Wi-Fi systems operating in the same band. Specifically, the paper focuses on a Q-learning and a Game Theory based approach. The implementation considerations of both approaches are discussed in relation to current 3GPP specifications and a comparison in terms of performance is presented to analyze the convergence time, the signaling requirements and the impact of errors in the throughput estimation.Peer Reviewe