基于RFID阵列的可标记无源感知机制研究

Nowadays, “batteryless sensing” has become a key supporting technology in intelligent sensing area, as it can extract energy from the environment to support computing, sensing, and communication. The state-of-art device-free sensing schemes, such as “WiFi-Radar sensing”, however, cannot effectively distinguish sensing signal sources from multiple objects, which leads to the issue of the unmarked sensing. In this project, we will study the RFID-based approach for “marked” batteryless sensing. On one hand, the backscatter communication of RFID.system is very sensitive to the surrounding transmission environment, the batteryless sensing can thus be achieved by exploring the RF-signals which carry the features of environmental factors; on the other hand, the inherent identification function of RFID system can be used to effectively distinguish multiple sensing signal sources, so as to realize the “marked” batteryless sensing. In this project, we will investigate the communication and sensing property of RFID, the topology and sensing feature of RFID tag array, as well as the model and methodology of the “marked” batteryless sensing, we will further leverage the spatiotemporal coupling to optimize the sensing performance. Finally, we will build up the experiment testbed, implement practical system prototypes, and realize the “marked” batteryless sensing towards the real complex environment.

无源感知技术通过从环境中获取能量支撑其计算、感知与通信，近年来已经成为感知计算领域的一项核心支撑技术。针对当前流行的非绑定式无源感知技术（如WiFi雷达感知）无法针对多个感知对象有效区分感知信号来源、缺乏“可标记性”的问题，本项目以RFID技术为切入点来研究可标记无源感知机制。一方面，基于RFID反向散射通信对传输环境的敏感性，利用标签反射信号中携带的各种环境因素的变化特征实现无源感知。另一方面，利用RFID能够被标识的特性可以有效区分多个感知信号来源，实现无源感知的“可标记性”。本项目将以多标签构建的感知阵列作为新型无源感知手段，重点研究如何基于RFID的通信与感知特性，有效构建RFID感知阵列的拓扑结构，挖掘和探索可标记的无源感知模型和感知方法，并结合多标签感知源的“时空协同”关系来优化感知效果，最终搭建实验平台，构建系统原型，面向真实复杂环境实现可标记的无源感知机制。

项目执行期间，研究团队针对当前流行的非绑定式无源感知技术无法针对多个感知对象有效区分感知信号来源、缺乏“可标记性”的问题，以RFID技术为切入点来研究可标记无源感知机制。一方面，基于RFID反向散射通信对传输环境的敏感性，利用标签反射信号中携带的各种环境因素的变化特征实现无源感知。另一方面，利用RFID能够被标识的特性可以有效区分多个感知信号来源，实现无源感知的“可标记性”。我们以多标签构建的感知阵列作为新型无源感知手段，重点研究如何基于RFID的通信与感知特性，有效构建RFID感知阵列的拓扑结构，挖掘和探索可标记的无源感知模型和感知方法，并结合多标签感知源的“时空协同”关系来优化感知效果，最终搭建实验平台，构建系统原型，面向真实复杂环境实现可标记的无源感知机制。.具体来说，项目执行期间，研究团队主要在基于RFID标签阵列的绑定式／非绑定式感知方面开展前沿理论和技术研究，在基于RFID的亚毫米级语音振动感知、基于RFID的毫米级滴速感知、基于心跳呼吸生命体征感知的身份鉴别机制、基于RFID的非接触转速感知等方面取得进展。在此基础上，发表／录用论文33篇，其中CCF A类会议／期刊论文19篇，包括ACM MOBICOM 1篇、IEEE INFOCOM 6篇、ACM UBICOMP 4篇、ACM MM 1篇以及IEEE Transaction on Mobile Computing(TMC) 7篇；CCF B类期刊论文6篇，包括IEEE ICDCS 1篇、ACM/IEEE IPSN 1篇、ACM/IEEE IWQoS 1篇、ACM Transactions on Sensor Networks 3篇；CCF C类期刊论文4篇；其他会议/期刊论文4篇等。申请国家发明专利10项。