Transmission Policy of Two-Way Relay Networks With Multiple Stochastic Energy Harvesting Nodes
Transmission Policy of Two-Way Relay Networks With Multiple Stochastic Energy Harvesting Nodes
复制标题
DOI：
10.1109/access.2019.2922222
复制
发表时间：
2019
期刊：
影响因子：
3.9
通讯作者：
Wei Li;Meng-Lin Ku;Yan Chen;Yichen Wang;Zhonghua Liang
中科院分区：
文献类型：
--
作者：
Wei Li;Meng-Lin Ku;Yan Chen;Yichen Wang;Zhonghua Liang
利用真实数据驱动的随机能量采集(EH)模型，研究了能量采集(EH)双向中继网络的在线最优传输策略和中断性能。在该网络中，两个太阳能源节点都配备了有限大小的电池，并通过一个中继节点相互通信。研究了继电节点的两种供电方式：固定供电(FP)和太阳能供电作为EH节点的有限大小电池供电。通过调整所有EH节点的发射功率以适应因果太阳能EH条件、衰落信道状态及其电池的储能，优化了TWR网络的长期中断概率。因此，该框架被描述为一个马尔可夫决策过程(MDP)，奖励函数被定义为网络条件中断概率的补集，并从理论上推导出其精确的闭合形式。在MDP模型的基础上，计算了网络的最优传输策略和期望中断概率。在FP中继模式下，揭示了最优策略的二维开关阈值结构，并显示了两个EH源节点功率动作的二维零作用、非保守和功率收敛特性。该特殊属性指示两个EH源节点之间的传输交互。进一步发现，网络的中断性能极限由网络电池的不足概率决定，它反映了两个源节点的EH能力对网络性能的影响。此外，在EH中继模式下，还推导了更为复杂的最优策略的三维开关门限结构和网络的中断性能极限，分别反映了EH中继对网络功率分配和中断性能的影响。最后，通过计算机仿真验证了理论分析的正确性。
The online optimal transmission policy and outage performance are investigated for energy harvesting (EH) two-way relay (TWR) networks by exploiting a real-data-driven stochastic EH model. In this network, two solar-powered source nodes are both equipped with finite-sized batteries, and communicate with each other via a relay node. Two energy supply modes are studied for the relay node: fixed power (FP) or solar-powered as an EH node with a finite-sized battery. The long-term outage probability of the TWR network is optimized by adapting the transmission power of all EH nodes to the causal solar EH condition, the fading channel states and the energy storage of their batteries. Thus, the framework is formulated as a Markov decision process (MDP), and the reward function is defined as the complement of the network’s conditional outage probability, whose exact closed-form is derived theoretically. Based on the MDP model, the optimal transmission policy and expected outage probability of the network are computed. In the FP relay mode, a two-dimensional (2D) on–off threshold structure of the optimal policy is revealed, and it shows the 2-D zero-action, non-conservative and power-converged properties of the power actions of the two EH source nodes. This special property indicates the transmission interaction between the two EH source nodes. Further, the outage performance limit of the network is found out to be determined by the deficiency probability of network batteries, and it reflects the impact of the EH capabilities of the two source nodes on the network performance. Moreover, in the EH relay mode, a more complicated three-dimensional on–off threshold structure of the optimal policy and the outage performance limit of the network are also derived, reflecting the impact of the EH relay on the network’s power allocation and outage performance, respectively. Finally, computer simulations are demonstrated to verify the theoretical analysis.