Q-learning for opportunistic spectrum access
Omar I. Alsaleh, Bechir Hamdaoui, Alan Fern · 2010
The expected shortage in spectrum supply is well understood to be primarily due to the inefficient, static nature of current spectrum allocation policies. In order to address this problem, FCC promotes the so-called Opportunistic Spectrum Access (OSA). In short, the idea behind OSA is to allow unlicensed users to use unused licensed spectra so long as they do not cause interference to licensed users. In this paper, we propose Q-OSA, a learning scheme that enables effective OSA, thus improving spectrum efficiency. Q-OSA does not require prior knowledge of the environment's dynamics, yet can still achieve high performance by learning from interaction with the environment.