同频并行无线充电中的群广播调度问题研究

With the increment of wireless charging efficiency, it is possible to wirelessly charge sensor node tens of meters away, which helps to extend the applications ofwireless sensor networks. The project proposes frequency-synchronized parallel wireless charging technique which can significantly accelerate the charging for at some positions of the sensor network. However due to the radio interference effect, nodes at different positions in the network can obtain much different power from the chargers, thus requiring appropriate scheduling. Aiming at this phenomenon, this project studies the implementation of frequency-synchronized parallel wireless charging and the group broadcasting scheduling problems therein. These problems differ from the classic combination optimization problems in three aspects: utility of each scheduled resource has broadcasting-like mapping relationship with the constrained objects; utility of the scheduled resources cannot be calculated independently and does not follow the monotony rule; special optimization object that minizes the number of scheduled activities. Research on these problems will have wide meaning in strategy programming applications with non-monotonous broadcasting mode inside..Innovations of the project are: observing the local accelelated charging efficiency and proposing the idea of frequency-synchronized parallel wireless charging; extracting a series of representative computation problems different from classic combination optimization problems from various typical charging scenario; and studying phase-controllable frequency-synchronized parallel wireless charging which will further improves the charging efficiency and leads to new nonmonotone group broadcasting scheduling problems with adjustable utility of the scheduled resources.

无线充电效率的快速提升，使长距（>10米）无线充电成为可能，由此给传感器网络应用赋予新的活力。本项目提出同频并行无线充电思想，可倍增局部无线充电效能，同时也因无线电干涉效应而导致不同区域充电效率极不均衡，因而需要合理调度。项目将研究同频并行无线充电实现技术及其群广播调度问题。该类问题不同于经典组合优化问题之处在于：调度对象的效用与约束对象间呈多对多映射关系；调度对象的效用不可独立计算且不具备单调性原则；优化目标较特殊——为最小化调度对象的行为数。该类问题的研究，对非单调广播下策略规划类应用，亦有参考意义。.项目创新性体现在：观察到同频率多路无线充电信号的局部倍增效应，并提出同频并行无线充电思想；对不同典型充电场景下同频无线充电站的调度需求，抽象为一系列较具代表性的新型组合调度问题；探索相位可控下的同频并行无线充电技术，在进一步提高充电效能下，研究调度对象效用可变下的非单调群广播调度问题。

无线充电技术，可解决传感器网络长期工作的能耗问题，因而受到广泛关注。近年来无线充电效率的提升，使得十余米开外无线充电成为可能，由此极大扩展了传感器网络的应用。由于无线充电这种全向能量扩散、点域能量接收的充电方式，电磁波大部分能量在空间散失。本项目提出同频并行无线充电思想，可倍增局部无线充电效能，例如在 路信号强度及相位相同下，同频并行无线充电的功率为 倍于非同频并行无线充电的充电功率，从而显著提升无线充电速度。然而，同频并行无线充电也因无线电干涉效应而导致不同区域充电效率极不均衡，因而需要合理调度。项目研究了同频并行无线充电实现技术及其群广播调度问题。该类问题不同于经典组合优化问题之处在于：调度对象的效用与约束对象间呈多对多映射关系；调度对象的效用不可独立计算且不具备单调性原则；优化目标较特殊——为最小化调度对象的行为数。项目最终实现了发射功率达4W的915MHz频段无线充电站，提出并实现了各无线充电站发射同频同相电磁波的控制方法——即各充电站统一接收457MHz高频信号作为输入源，对其进行倍频后，经不同移相控制，然后功放辐射出去。项目给出干涉效应下并行无线充电的充电模型，并进行了实验验证。项目完成了给定充电站部署和被充节点部署下的三种调度算法设计（两种贪心+一种遗传算法）。项目完成了给定被充节点部署下的最佳充电站部署算法（两种贪心算法）。项目将这两个问题转化成子模覆盖问题，并证明了其时间复杂度和性能上界为lnMT+1。本项目所研究的群广播调度问题不仅仅存在于同频并行无线充电技术中，也可运用于策略规划应用领域，因为策略的效用也常常呈现区域上的非单调性和时域上的非线性特征。本项目所研究相应的群广播调度问题，对相关策略的设计亦具有参考意义。