
2014 年 6 月 Journal on Communications June 2014
第 35 卷第 6 期
通 信 学 报
Vol.35
No. 6
LRST:低冗余搜索树防碰撞算法
黄琼
1
,凌江涛
1
,张敏
2
,阳小龙
2
(1. 重庆邮电大学 移动通信技术重点实验室,重庆 400065;2. 北京科技大学 先进网络技术与新业务研究所,北京 100083)
摘 要:针对 RFID 标签防碰撞树型算法在识别过程中因询问命令过多、过长而产生大量冗余数据导致通信开销
过大的问题,在后退式动态搜索树算法的基础上提出一种低冗余搜索树防碰撞算法(LRST):为减少询问次数,
提出了“一问两答”询问方式,即碰撞标签根据最高碰撞位比特分别在第一个时隙或第二个时隙响应;为减小询
问命令的长度,用计数器替代标签中的前缀匹配电路,使算法不再需要前缀作为询问命令的标识参数;此外,提
出的预测识别和标签屏蔽机制规避了不必要的询问。理论分析和仿真结果表明,通信开销大大降低。
关键词:RFID;防碰撞;搜索树;低冗余
中图分类号:TP393 文献标识码:A 文章编号:1000-436X(2014)06-0110-07
LRST: searching tree anti-collision algorithm with low-redundancy
HUANG Qiong
1
, LING Jiang-tao
1
, ZHANG Min
2
, YANG Xiao-long
2
(1. Key Laboratory of Mobile Communication Technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, China;
2. Institute of Advanced Network Technology and New Services, University of Science and Technology Beijing, Beijing 100083, China)
Abstract: During the RFID tag identification process, the tree-based anti-collision algorithms usually incur large amount
of redundant data due to an excess of long query commands,which increases the communication overhead. To resolve
this problem, a searching tree anti-collision algorithm with low-redundancy on the basis of regressive-style dynamic
searching tree algorithm was proposed. In order to reduce the number of queries, a novel query mode was developed, i.e.,
single query with duo responses. Depending on the most significant collided bit, the collided tags respond in the first or
second slot separately. In order to reduce the length of query command, the prefix matching circuit in tag was replaced
with a counter, which eliminated the prefix as the parameter of query command. The predictive identification and block-
ing technique were also introduced to avoid unnecessary queries. Theoretical analysis and simulation results show that
the communication overhead is greatly reduced.
Key words: RFID; anti-collision; searching tree; low-redundancy
1 引言
射频识别(RFID, radio frequency identification)
系统中,多个标签同时响应阅读器请求时会发生碰
撞。为了正确识别标签,必须进行标签防碰撞处理。
目前,考虑到实现的复杂度和成本等因素,其中时
分多址(TDMA)是解决 RFID 系统碰撞问题的常用
方法之一。基于 TDMA 的防碰撞算法主要有两大
类
[1,2]
:一类是基于 ALOHA 协议的算法;另一类是
基于树型的算法。虽然基于 ALOHA 协议的算法对
标签硬件要求较低,但它们的吞吐率低(不超过
36.8%),并会出现“标签饥饿”问题(即某个标
签可能一直与其他标签碰撞而无法被识别)。然而
树型算法不仅吞吐率高,且不存在“标签饥饿”问
题。目前,树型算法主要有分裂树、查询树、搜索
树、逐位识别树等,其中搜索树算法因采用了曼彻
收稿日期:2013-05-06;修回日期:2013-12-06
基金项目:国家重点基础研究发展计划(“973”计划)基金资助项目(2012CB315905);国家自然科学基金资助项目(61171111
61102063);长江学者和创新团队发展计划基金资助项目(IRT1299);重庆市科委重点实验室专项经费基金资助项目
Foundation Items: The National Basic Research Program of China (973 Program) (2012CB315905); The Na
Foundation of China (61171111, 61102063); Changjiang Scholars and Innovative Research Team in University (IRT1299
Special Fund of Chongqing Key Laboratory
doi:10.3969/j.issn.1000-436x.2014.06.014