基于信息年轮的物联网信息采集与数据传输机制研究

In recent years, the research on real-time sensing in the Internet of Things based on the Age of Information draws great attention. Because the Age of Information provides new theoretical foundation and technical method for minimizing the gap between the sensing delay and the finite information age. The key problem of realizing this technology is decided by the coupling design and optimization of the information acquisition and data transmission during the sensing process, for effectively minimizing the sensing delay in the real-time environment sensing. To this end, we explore the characteristics of the practical sensing environment, take the age of information as the performance metric of real-time sensing and mainly focus on the coupling design and optimization of the mechanisms of information acquisition and data transmission. In this project, we will do the following works to solve the problem. 1) For the heterogeneous sensing jobs and their random states, we design the distributed dynamic sampling strategy and meanwhile optimize the allocation of the limited bandwidth resources among the various sensing jobs. 2) For the constrained power supply of the sensing node, we design the energy-saving sampling and transmission power control mechanism, and optimize the energy allocation between the sampling and transmission. 3) For the complex transmission topology, we design the feedback sampling and cooperative transmission mechanism to achieve the dynamical matching between the information acquisition and the time-varying transmission environment. Overall, this project is to provide useful and insightful guidelines for the practical application of the real-time sensing in the Internet of Things.

近年来，基于信息年轮的物联网实时感知研究受到广泛关注，因为信息年轮为缩小感知时延与状态信息时效之间的差距提供了新的理论依据和技术方法。该技术实现的一大关键问题在于，如何针对物联网感知的信息采集与数据传输过程进行耦合设计与优化，以有效提升监测设备对于环境中状态变化的感知实时性。针对上述关键问题，本项目从实际物联网感知场景的特征出发，以信息年轮作为感知实时性的评价指标，重点关注物联网实时感知的信息采集与数据传输机制的耦合设计与优化，具体包括：（1）针对感知业务多样性和状态变化随机性，设计分布式动态采样机制，同时优化有限带宽资源在不同业务之间的分配；（2）针对传感节点供能受限，设计节能采样与传输功率控制机制，优化实时可用能量在采样与传输两者之间的分配；（3）针对传输拓扑复杂，设计反馈采样与传输协作机制，实现信息采集过程与时变传输环境的动态适配。研究结果将对物联网实时感知的实际应用提供有益参考。

近年来，基于信息年轮的物联网实时感知研究受到广泛关注，因为信息年轮为缩小感知时延与状态信息时效之间的差距提供了新的理论依据和技术方法。本项目从实际物联网感知场景的特征出发，以信息年轮作为感知实时性的评价指标，围绕多源异构任务感知与推断模型建立、节点自主获能与数据传输联合机制设计以及节点高效供能理论与方法三个方面，主要完成了以下创新性工作。首先充分考虑了实际情况，如感知任务在物理空间中的位置特性、任务的随机性和时效性以及用户处理感知任务的能力异质性等等因素，设计了更为有效和具有普适性的数据感知、传输与推断模型。其次，针对高动态和复杂的物联网感知应用场景下节点无法获得稳定能量供给的情况，基于一维离散空间和二维连续空间分别提出了基于深度强化学习的传感节点移动轨迹自适应优化机制，在能量获取与数据传输之间取得了平衡。最后，针对目前物联网采集与传输系统中传感节点的供能瓶颈，从静态充电、移动充电、全向充电和定向充电等多个维度充分研究了对传感网络进行无线供能的可行性并且针对充电能耗成本进行了优化，为保障物联网传感节点在传统电网不可接入的感知场景下得以获取稳定的电能以进行正常运作提供了一种新思路。研究结果将对物联网实时感知的实际应用提供有益参考。