【作者】 张浩；

【导师】 牛存良；

【作者基本信息】 河北工业大学 ， 计算机科学与技术， 2016， 硕士

【摘要】 物联网的目的就是连接世界上的所有物体,并对其进行识别,跟踪,管理。无线射频识别技术（Radio Frequency Identification,RFID）作为物联网应用的支撑关键技术,越来越受到人们的关注。RFID射频识别是通过射频载波信号从物品所携带的电子标签中获取识别信息,整个识别过程是自动化进行的,不需要人工干预。非接触性是RFID射频识别的最大特点,在RFID系统中,如果读写器的工作区域中存在多个标签并同时向读写器发送数据,标签的数据就发生了冲突,读写器无法正确识别标签数据。很明显,数据冲突是影响RFID系统识别效率的主要因素。因此,采用一个有效的防碰撞算法在RFID系统中极其重要。本文首先对当前典型的RFID防碰撞算法进行了研究,包括基于二进制的确定型算法和基于ALOHA的概率型算法。其次,本文分析了如今应用非常广泛的RFID防碰撞算法:Q算法。Q算法是由RFID标准化组织EPCglobal提出的,该算法是在DFSA算法基础上改进实现的。本文在分析了Q算法的性能及其缺陷之后,提出了Q-BS以及Fast-Q两种改进算法:1)Q算法是在DFSA（Dynamic Frame Slotted ALOHA）算法的基础上改进实现的。即使在标签数量较多的情况下,Q算法都有着稳定的系统识别效率。二进制分割算法（Binary Splitting,BS）是确定型的二进制防碰撞算法,在标签数量较小的场景下,BS算法有着相当好的识别效率。Q-BS算法将BS算法嵌入到Q算法中执行,充分利用了两种算法的优点,是一种混合的防碰撞算法。仿真结果表明:新算法能有效的提高RFID系统识别效率,减少所消耗的总时隙数和冲突时隙数,具有一定的实用价值。2)Q算法使用固定步长C来调整Q值是很不合理的。针对Q算法的这种不合理性,本文将提出了第二种改进算法Fast-Q算法。Fast-Q算法在冲突时隙和空闲时隙两种情况下使用不同的参数来调整Q值。如果为冲突时隙,使用参数C_coll调整Q值;如果为空闲时隙,使用参数C_idle调整Q值。随后本文将对于参数C_coll和C_idle的选取以及它们之间的比例关系做了详细的求解分析。仿真结果表明:新算法能有效的降低所消耗的总时隙数和冲突时隙数,缩短了识别标签所需的总时延。更多还原

【Abstract】 The purpose of the IOT（Internet of Things）is to connect all the objects in the world and to identify,track and manage them.As the key technology of IOT,RFID（Radio Frequency Identification）is achieving more and more attention.RFID is using the RF carrier signal from the goods carried by the electronic tags to obtain identification information.And The entire identification process is automated,without human intervention.Non-contact is the biggest characteristic of RFID.In the RFID system,if there are multiple tags in the work area of the RFID reader and send data to the reader at the same time,the data will conflict,the reader will not be able to correctly identify the data.Obviously,the data conflict is the main factor that affects the system identification efficiency of RFID.Therefore,the use of an effective anti-collision algorithm in the RFID system is extremely important.Firstly,in this paper,a typical RFID anti-collision algorithm is studied,including the binary-based deterministic algorithm and the ALOHA-based probabilistic algorithm.Secondly,this paper analyzes the widely used RFID anti-collision algorithm:Q algorithm.Q algorithm is proposed by the RFID standardization organization EPCglobal,the algorithm is improved on the basis of DFSA algorithm implementation.After analyzing the performance of Q algorithm and its defects,this paper proposes two improved algorithms,Q-BS algorithm and Fast-Q algorithm:1)Q algorithm is improved on the basis of DFSA algorithm.Even in the case of a large number of tags,Q algorithm has a stable system identification efficiency.BS（Binary Splitting）is a deterministic binary anti-collision algorithm,in the case of small number of tags,BS algorithm has a very good identifidation efficiency.The Q-BS algorithm embeds the BS algorithm into the Q algorithm,and makes full use of the advantages of the two algorithms.It is a hybrid anti-collision algorithm.The simulation results show that the new algorithm can effectively improve the efficiency of RFID system identification and reduce the number of required time slots and the number of collision slots,which has certain practical value.2)It is unreasonable that Q algorithm uses fixed step size C to adjust the Q value.Aiming at this kind of irrationality of Q algorithm,this paper proposes a second improved algorithm:Fast-Q algorithm.The Fast-Q algorithm uses different parameters to adjust the Q value in both collision slots and idle slots.If it is a collision slot,we will adjust the Q value using the parameter C_coll,if it is an idle slot,we will adjust the Q value using the parameter C_idle.In the following,we will analyze C_coll and C_idle in detail and their proportion relation.The simulation results show that the new algorithm can effectively reduce the number of total slots and collision slots,and shorten the total recognition delay.更多还原