Energy-aware WiFi network selection via forecasting energy consumption

Covering a wide area by a large number of WiFi networks is anticipated to become very popular with Internet-of-things (IoT) and initiatives such as smart cities. Such network configuration is normally realized through deploying a large number of access points (APs) with overlapped coverage. However, the imbalanced traffic load distribution among different APs affects the energy consumption of a WiFi device if it is associated to a loaded AP. This research work aims at predicting the communication-related energy that shall be consumed by a WiFi device if it transferred some amount of data through a certain selected AP. In this paper, a forecast of the energy consumption is proposed to be obtained using an algorithm that is supported by a mathematical model. Consequently, the proposed algorithm can automatically select the best WiFi network (best AP) that the WiFi device can connect to in order to minimize energy consumption. The proposed algorithm is experimentally validated in a realistic lab setting. The observed performance indicates that the algorithm can provide an accurate forecast to the energy that shall be consumed by a WiFi transceiver in sending some amount of data via a specific AP.