| Summary: | This paper studies a new design of a wireless-powered relay network to improve its throughput performance by opportunistic energy harvesting (EH) and information relaying (IR) at the relay node. To enable the relay to implement opportunistic EH and IR, a new protocol, namely, the adaptive harvest-store-forward (AHSF) protocol, is proposed. In the AHSF protocol, the relay is allowed to adaptively switch between an EH mode and a power-splitting (PS)-based IR mode at the beginning of each time frame; moreover, the relay can adaptively determine the amount of energy for IR based on the wireless channel conditions over one finite time horizon, which consists of multiple consecutive time frames. Then, under the proposed protocol, a throughput maximization problem is formulated to optimize the relay operation mode (i.e., EH mode or IR mode) and resource allocation (i.e., time slot, PS ratio, and power allocation) in the IR mode in a finite time horizon. The formulated problem is coupled over time and, thus, is intractable. To address the intractability, dynamic programming is employed to transform the problem into a series of Bellman equations. Based on the Bellman equations, the optimal decision criteria for the relay operation mode is derived, and a time-coupled optimization problem for the resource allocation in the IR mode is addressed by decoupling into two subproblems for two adjacent time frames. By solving the subproblems via convex programming, an optimization policy with causal channel-state information (CSI), which can be implemented online in real time, and an optimization policy with full CSI, which is used to reveal the upper bound of the achieved throughput, are achieved. The simulation results verify that the proposed design can yield significant throughput gains over the existing wireless-powered relaying schemes.
|
|---|
