在這篇論文裡，我們考慮一個利用多天線波束成形的通訊系統，下行鏈路進行訊息及無線能量傳輸。然而，波束成形的系統通常需要由接收機回傳接收機和發送機之間的通道狀態資訊(CSI)。波束成形效能會因發射端所知的通道狀態資訊不完美而下降。雖然愈精確的通道狀態資訊可以提供越好的波束成形效益，但在接收端能量與傳輸時間有限情況下，通道估計或回傳通道狀態資訊所需的能量和時間不可能無限加大以達到精確的通道狀態資訊。吾人考慮一個點對點的MISO通訊系統，由下行時間接收端會以功率分流器的方式進行訊息及能量傳輸，並於上行時間回傳通道狀態資訊。 論文第一部分探討的情況為接收端直接將當下獵取到的能量全部用於回傳通道狀態資訊，以最大化下行鏈路資訊吞吐量為目標，透過簡化最佳化問題來找 出功率分流器參數以及上下行時間分配的次佳解。第二部分則是將接收端增加一個能量儲存裝置，由接收端視通道狀態以及能量儲存裝置的狀態來決定功率分流器參數以及上下行時間分配，並利用馬可夫決策過程(Markove decsision process, MDP)根據對過往的觀察與現今的狀態找出最佳的傳輸策略。 ;In this paper, we consider a communication system using multi-antenna beamforming, downlink information and wireless energy transmission. However, the beamforming system typically requires the receiver to return the channel state information (CSI) between the receiver and the transmitter. Beamforming performance is degraded by imperfect channel state information known to the transmitter. Although the more accurate channel state information can provide better beamforming benefits, but the energy and time required for CSI feedback can’t be infinitely increased to achieve perfect since the receiver′s energy and feedback time of the CSI are limited. We consider a point-to-point MISO communication system. At the downlink time, the receiver will be the way the power splitter for simultaneous information and energy transfer, and feedback channel state information at the uplink time. The first part of the paper explores the situation that the receiver can directly use the energy obtained in the current time to feedback the channel state information. To maximize the downlink information throughput, the suboptimal solution of the power splitter parameter and downlink time allocation is found by simplifying the optimization problem. The second part is to add an energy storage device to the receiver. The power splitter parameters and the downlink time allocation are determined by the receiver depending on the channel state and the state of the energy storage device. Then, we use Markov decision process (MDP) according to the past observation and the current state to find the best transmission policy.
