面向开放式RFID系统的丢失标签检测技术的研究

As one of key sensing technologies in Internet of Things (IoT), Radio Frequency Identification (RFID) has attracted increasing interests in recent years due to its advantages of low cost, small size, and batch processing. When a tag is attached to a specific object, the RFID reader is able to sense whether or not the tagged object is missing, thereby guarding against theft in real time. To date, although some advanced research works on missing tag identification have been proposed, they are built on a strong assumption that there are no unknown tags in an ideal RFID system, which is hard to meet in reality. In light of this, this project removes this ideal assumption and studies how to efficiently identify missing tags in an open RFID system. First, for an uncertain RFID system, we study how to avoid the signal interference caused by unknown tags and lower the false negative ratio. Second, for a large RFID system, we study the collaborative inference based on multiple readers, establish the novel communication model tailored to networked tags, and design the state-free routing on top of the proposed communication model. Third, for a privacy-sensitive RFID system, we study how to deal with the irreconcilable collisions caused by blocker tags and design grouping-based schemes to improve the identification efficiency of missing tags.

射频识别（RFID）作为物联网的关键感知技术之一，凭借其低成本、体积小、批量读取等优势被广泛应用于各行各业之中。当RFID标签附着于某物体上，阅读器就可以通过收集标签信息来自动化感知该物体的存在与否，从而实现防盗监测、避免因盗窃造成的经济损失。然而，现有相关研究大都基于理想化的系统假设，难以应用于现实开放式RFID系统之中。针对于此，本项目移除这一不合理假设，从开放式RFID系统的不确定性、大规模性、及隐私敏感性三个特征出发，研究如何高效地识别丢失标签。首先，对于RFID系统的不确定性，研究如何避免未知标签对待监测标签的信号干扰，从而降低丢失标签的漏报率；其次，对于RFID系统的大规模性，研究多阅读器协同推理机制以及建立适用于网络标签的通信模型与无状态路由协议；最后，对于RFID系统的隐私敏感性，研究如何在阻塞响应机制下解决由阻塞标签导致的不可调和冲突以及设计分组策略加速丢失标签的识别。

射频识别（RFID）作为物联网的关键感知技术，被广泛应用于仓储物流、智慧医疗等行业。当RFID标签附着于物体上，阅读器就可以通过收集标签信息来自动化感知该物体的存在与否，从而实现防盗监测、避免因盗窃造成的经济损失。然而，现有相关研究大都基于理想化的系统假设，难以应用于现实开放式RFID系统之中。针对于此，本项目移除这一不合理假设，从开放式RFID系统的不确定性、大规模性、及隐私敏感性三个特征出发，研究高效的丢失标签识别协议。针对不确定性，研究了能够抵御未知标签干扰的丢失标签检测协议，从而降低漏报率；针对大规模性，研究了多阅读器协同推理机制以及建立了适用于网络标签的通信模型与无状态路由协议；针对隐私敏感性，研究解决了阻塞标签导致的不可调和冲突以及设计了分组策略加速丢失标签的识别。上述研究为真实场景下的RFID防盗应用奠定基础。