Kalman Prediction Based Proportional Fair Resource Allocation for a Solar Powered Base Station

Abstract

This paper considers optimising transmissions of base station (BS) equipped with a solar panel. The BS is assumed to serve a distributed wireless network of nodes or gateways on orthogonal channels. Considering the daily periodicity as well as short term variations of solar energy, and the differences in channel gain from the BS to the nodes, a scheduling method for efficient delivery of maintenance messages from the BS to the wireless nodes that it serves to (such as wireless sensor network gateways) is investigated. The proposed solution is an online resource allocation algorithm, called PTF-On, that can predict the base station's energy arrival profile, and then, act upon this profile to determine the best power and time allocation that will maximize the throughput in a proportionally fair way.