Green Wi-Fi Implementation and Management in Dense Autonomous Environments for Smart Cities

Abstract

Advanced informatics technologies facilitate the construction of green smart cities, especially the Wi-Fi implementation and management, for rapidly increasing personal Wi-Fi devices in autonomous environments residing in nonoverlapped channels often result in low energy efficiency and severe cochannel interference. In this paper, a green Wi-Fi management framework is constructed in order to reduce the overall energy consumption through turning off a portion of access points (APs) and aggregating their users to the other active APs. A Tabu-search-assisted active AP selection algorithm is proposed to minimize the power consumption with a seamless wireless converge. For the active APs, based on our defined metric airtime cost that is integrated by the in-range interference and the hidden terminal interference, a reinforcement-learning-aided AP self-management algorithm is proposed to dynamically adjust APs’ channels in the partially overlapped channel space. Extensive simulations and field experiments demonstrate that the power consumption can be reduced by about 65%, and the airtime cost of APs can be reduced by 50% compared with the typical least congestion channel search algorithm.