基于射频充电的无线传感网数能协作策略研究

Recently, power beacons (PBs), which beamform energy to sensor nodes, become a key solution to solve energy shortage problem in Wireless Sensor Networks (WSNs). However, past works on PB assisted WSNs usually transmit data in one-hop network such that increases energy consumption of data transmission. Further, they ignore time-varying channel gains that impact on energy harvesting and link capacity. In addition, there are lack of works on multiple PBs scenarios. To this end, this project proposes the problem of jointly optimizing the charging of nodes, transmission power of PBs, as well as data routing and link scheduling of sensors in time-varying environments. It aims to solve the following research issues. First, it studies the coupling relationship between charging nodes, charging power, data routing and link scheduling. Second, it considers timely response of a PB in time-varying environments. Third, it explores charging node selection scheme under multiple PBs scenario. To address the aforementioned problems, this project will employ Mixed Integer Linear Programming (MILP) and Markov Decision Process (MDP). It will design heuristic algorithms to solve problem instances involving large-scale WSNs, and also design features to be incorporated into linear approximation function that is used to handle large state-action space. Apart from that, the project will consider a distributed Q-learning or MDP approach for scenarios with multiple PBs. This project is expected to jointly optimizes data transmission and charging strategies in a PB-assisted multi-hop WSNs in order to further increase energy efficiency. As a result, this project will help develop wireless charging solutions that realize the future Internet of Things (IoTs).

射频能量射束因高可控性和较低传输损耗成为远距离充电的重要技术。通过多跳网络优化数据传输是无线传感网降低节点能耗的基本方法。然而，当前综合考虑两者的研究不够深入，对于结合多跳网络的射频充电策略、时变环境下快速响应机制和多射频能源协作充电的研究比较匮乏。因此，针对以上不足，本项目将研究以下问题：充电策略、路由选择和链路调度的耦合关系；时变环境下射频能源的快速响应与决策原理；多射频能源协作下能量分配与数据传输规律。本项目考虑节点能量与数据存储的约束，以节点间数据公平采集为优化目标，通过混合整数线性规划与马尔可夫决策过程建模，控制充电策略与数据传输，设计快速响应时变环境的数能协作网络协议。本项目创新性地综合充电策略、路由选择、时变环境和多射频能源协作等因素，从提高节点采集能量和降低传输能耗两方面，进一步提升网络能量利用效率，预期将完善射频充电无线传感网数能协作的理论基础，扩展其在民生领域的应用。

射频充电因高可控性和较低传输损耗成为远距离充电的主要技术；优化多跳网络数据传输是降低网络节点能耗的基本方法。当前协同优化射频充电策略和数据传输路由的研究较为匮乏；考虑信道和太阳能辐射时变特性的研究不够充分；多射频能源协作下能量分配与数据传输规律的研究有待加强。因此，本项目解决以下科学问题：充电策略、路由选择和链路调度的耦合关系，时变环境下射频能源的快速响应与决策原理，多射频能源协作下能量分配与数据传输规律。本项目首先确立射频充电无线传感网数学模型，考虑节点非线性能量转化过程，以最大最小化采集节点在多个时隙的采集速率，综合优化射频能源功率分配、节点路由选择、链路调度，构建混合整数线性规划模型。..在处理该模型的时变环境方面，本项目设计射频能源充电策略和节点数据采集的完全分布式协议D-MRA，使射频能源和网络节点基于本地信息完成最大最小化节点采集速率，仿真结果表明D-MRA可达最优结果的58.25%，并可在大型网络中运行。此外，本项目还引入认知无线电技术，设计基于博弈论的节点完全分布式功率分配协议，使节点完全基于本地信息完成发送功率设置。..在多射频能源协作方面，本项目针对应用场景，考虑物联网需要确保采集数据的新鲜程度，从而引入信息年龄度量数据新鲜程度。本项目以最小最大化各个采集节点在多个时隙的信息年龄为优化目标，综合优化数据采集、路由选择、链路调度、射频充电等，构建混合整数线性规划模型，并证明节点信息年龄在该网络中的上下界。本项目还设计中心控制协议CLR和分布式DPS。仿真结果表明，两种协议分别为最优结果的1.6倍和1.95倍。为处理复杂数据采集任务，本项目将用户需求构建成不同虚拟网络，以最大化虚拟网络映射数量构建混合整数线性规划模型并设计启发式算法。仿真结果表明，所设计的算法可以达到最优结果的90.31%。

内容获取失败，请点击重试