Q-learning for dynamic channel assignment in cognitive wireless local area network with fibre-connected distributed antennas

Cognitive wireless local area network with fibre-connected distributed antennas (CWLAN-FDA) is a promising and efficient architecture that combines radio over fiber, cognitive radio and distributed antenna technologies to provide high speed/high capacity wireless access at a reasonable cost. In this paper, a Q-learning approach is applied to implement dynamic channel assignment (DCA) in CWLAN-FDA. The cognitive access points (CAPs) select and assign the best channels among the industrial, scientific, and medical (ISM) band for data packet transmission, given that the objective is to minimize external interference and acquire better network-wide performance. The Q-learning method avoids solving complex optimization problem while being able to explore the states of a CWLAN-FDA system during normal operations. Simulation results reveal that the proposed strategy is effective in reducing outage probability and improving network throughput.