Scheduling wireless power transmission over a fading channel

In this paper, we consider the problem of causally scheduling wireless power transmission in a simple wireless power transmission system which includes a wireless power transmitter (WPT) and a wireless power receiver (WPR). The channel used for wireless power transfer from WPT to WPR is modeled as a finite-state block channel. It is assumed that both WPT and WPR have a knowledge of current channel state information (CSI) and statistical information about future CSI, which is used to make an optimal decision to maximize energy efficiency. Naturally, the scheduling problem converts into a Markov decision process. Under the constraints of transmission completion time, we use stochastic dynamic programming to derive algorithms finding decision thresholds. With these thresholds, we obtain the optimal scheduling policies. Numerical results manifest that the proposed policies can result in significant energy savings, especially for short task.