在2009年，中国首次提出“智慧地球”的概念，实现真正的物物相连。随后，国家加大对物联网相关研究的政策及资金扶持，中国物联网进入高速发展通道，新的物联网应用系统不断出现。物联网主要通过数据收集、信息传输以及信息处理管理和使用三个过程实现物理世界中的物体与虚拟网络世界的交互连接。其中，数据收集过程主要负责感知物理环境、收集实时物理数据并重构相应的通用概念，作为该过程核心技术之一的RFID技术则负责实际物体的识别，在物联网中具有非常重要的意义。 阅读器和多个标签组成射频识别RFID系统。系统工作时，阅读器发出查询命令，标签被动地响应。由于所有标签采用相同的无线信道，如果多于一个标签响应阅读器，并同时传输物品相关数据则会造成数据信号干扰，从而导致阅读器无法正确解码接收到的信号，造成标签碰撞问题。标签碰撞问题不仅增加识别延迟，还造成带宽浪费，对RFID系统管理操作具有很大的影响，通过设计高效的RFID标签防碰撞协议可以有效缓解这一问题提高标签识别的效率。目前有很多学者对RFID标签碰撞问题进行研究并提出了各种标签防碰撞协议，总体来说它们的系统效率还有待提高，特别是随着物联网行业的快速发展和广泛应用，各种各样的实际应用环境对防碰撞技术的要求也越来越高。例如在标签快速移动的应用环境中，新的标签不断加入阅读器的读取范围，造成阅读过程复杂，且由于标签的快速移动性，部分标签在没有被读取的情况下，移出阅读器的读取范围，造成标签漏读的情况，在这种情况下研究快速鲁棒性强的标签防碰撞协议是一项颇具挑战性的课题。 静态场景下的标签防碰撞协议假定：一轮标签识别过程中，没有新标签到达阅读器区域，所以通过前一轮识别结果，可以通过该历史信息，利用标签估计算法估计下一轮未识别的标签数量，据此来设计防碰撞协议并进行相关参数设置。在有标签移动的动态应用环境，标签可以随时加入每轮的识别过程，阅读器没有这些新到达标签的数量和ID信息，而对这些标签的数量估计又会受到其阅读范围内己识别标签的影响，因此在已有静态场景中的标签数量估计方法并不可行，从而这些静态场景下的协议也就无法在动态场景下正常运行，将可能出现严重的标签漏读问题。归纳这些并进一步分析可知，现有标签防碰撞协议不适应动态环境所引起的标签漏读问题的主要原因包括协议在确定其参数时没有考虑新到达标签以及它们的识别效率较低。本论文基于这两个原因，一方面从动态环境下标签估计角度，从密集场景下标签估计的准确性入手，提出改进标签拥塞问题的高效防碰撞算法。另一方面，提出了移动场景中针对停留标签分组快速识别，极大提高了系统效率。相应研究成果如下： (1)提出了一种计算复杂性低且更加精确的标签估计算法，研究了射频识别RFID系统在密集标签环境下识别性能下降问题。通过调整响应标签的数量，缓解标签拥塞问题，提出了优于PRQT的PMQT算法，更适合大量标签的碰撞问题解决和有更好的算法鲁棒性。 (2)提出了一种支持标签持续到达的动态场景下的防碰撞协议GSRA，该协议主要用于解决动态帧时隙ALOHA(DFSA)标签防碰撞协议在移动标签环境中的识别问题，它可以有效解决已到达标签对识别过程的影响，提高识别效率。为此，将识别过程分为两部分，对于新到达标签仍然采用现有识别算法，但对于已经识别的停留标签，在阅读器侧可利用其历史信息，对其进行分组快速识别，实现了识别过程的高效性。通过理论分析可知：其系统效率仅与标签迁移率及新到标签识别效率有关。 (3)研究汉明重量区分不同标签的能力，发现3位二进制能很好的区分碰撞过程中标签的前缀。在此基础上提出采用汉明重机制的防碰撞协议HWQT，它能够为标签准确分配识别时隙。该协议方法简单，以低前缀开销为代价，显著减少了时隙识别阶段碰撞帧时隙和空闲帧时隙数，从而将标签识别性能大幅提升，理论分析表明，其系统效率可达到75％。 关键词：射频识别；动态环境；标签防碰撞；标签估计；汉明重量

 The concept of “the wisdom of the earth” was put forward firstly. It will be realized the real object to be connected in 2009. Subsequently, the government increased the policy and financial support of the related research on the Internet of Things(IoT), and the Chinese IoT entered the high-speed development channel. The new IoT application system was constantly emerging. The IoT is mainly through data collection, information transmission and information processing management and the use of three processes to realize the interactive connection between objects in the physical world and the virtual network world. The data collection process is mainly responsible for sensing the physical environment, collecting real-time physical data and reconstructing the corresponding general concept, and the radio frequency identification(RFID) technology, which is one of the core technologies of the process, is responsible for the identification of the real object and has very important significance in the internet of things. The reader and the plurality of tags constitute a RFID system. When the system works, the reader issues a query command and the tag responds passively. Since all the tags use the same radio channel, if more than one tag responds to the reader and simultaneously transmits the item-related data, the data signal is disturbed, resulting in the reader being unable to correctly decode the received signal resulting in a tag collision problem. The problem of tag collision not only increases the identification delay, but also causes the waste of the bandwidth, has a great influence on the RFID system management operation, and can effectively alleviate the problem by designing the high-efficiency RFID tag anti-collision protocol and improve the efficiency of the tag identification. At present, many scholars have researched and put forward various anti-collision protocols for RFID tags, in general, their system efficiency is still to be improved, especially with the rapid development and wide application of the IoT industry. The requirement of anti-collision technology is also higher and higher for a wide variety of practical application environments. For example, in the application environment where the tag is fast moving, the new tag is continuously added to the reading range of the reader, causing the reading process to be complicated, and due to the rapid mobility of the tag, the part of the tag is moved out of the reading range of the reader without being read, causing the leak of tag. In this case, it is a challenging task to study the fast robust tag anti-collision protocol. The tag anti-collision protocol under the static scenario assumes that no new tag reaches the reader reading area during a round of tag identification, so that the number of the next unidentified tags can be estimated by the tag estimation algorithm through the history information through the previous round identification result. Based on this, the anti-collision protocol is designed and relevant parameters are set. In a dynamic application environment with a tag movement, the tag can join the identification process of each round at any time, the reader does not have the number and ID information of the newly arrived tags, Therefore, it is not feasible to estimate the number of tags in the existing static scenario, so that the protocols under these static scenarios can not run properly in the dynamic scenario, and the problem of serious tag leakage is likely to occur. These and further analyses show that the main reason for the problem of tag leakage reading caused by the existing tag anti-collision protocol is that the protocol does not take into account the new arrival tag and the recognition efficiency is low when determining its parameters. Based on the two reasons, on the one hand, from the perspective of the tag estimation under the dynamic environment and the accuracy of the tag estimation from the dense scene, an efficient anti-collision algorithm for improving the tag congestion problem is put forward. On the other hand, the fast identification of the arrived tag in the mobile scene is proposed, and the system efficiency is greatly improved. The corresponding research results are as follows: (1)A tag estimation algorithm with low computational complexity and more accurate is proposed, and the performance degradation problem of RFID system in a dense tag environment is studied. By adjusting the number of the response tags, the problem of tag congestion is alleviated, and the PMQT algorithm superior to the PRQT is proposed, which is more suitable for solving the collision problem of a large number of tags and has better algorithm robustness. (2)The anti-collision protocol (GSRA) under the dynamic scenario which supports the continuous arrival of the tag is proposed, which is mainly used to solve the problem of the identification of the dynamic frame time slot ALOHA (DFSA) tag anti-collision protocol in the mobile tag environment, and can effectively solve the influence of the arriving tag on the identification process, and the system efficiency is improved. For this purpose, the identification process is divided into two parts, and the existing identification algorithm is still adopted for the newly arrived tag, but for the resident tag which has been identified, its history information can be used on the reader side, the tag can be quickly identified, and the high efficiency of the identification process is realized. The theoretical analysis shows that the system efficiency is only related to the tag mobility and the new tag recognition efficiency. (3)To study the ability of the Hamming weight to distinguish the different tags, it is found that the 3-bit binary can well distinguish the prefix of the tag in the collision process. On this basis, a anti-collision protocol (HWQT) with Hamming-weight mechanism is proposed, which can accurately assign the identification time slot to the tag. The method is simple, at the cost of low prefix cost, the number of the collision frame time slot and the idle frame time slot in the time slot identification stage is obviously reduced, so that the identification performance of the tag is greatly improved, and the theoretical analysis shows that the system efficiency can reach 75 percent. Keywords: Radio Frequency Identification; Dynamic Environment; Tag Anti-collision; Tag estimation; Hamming Weight