结合LSTM的分组比特时隙ALOHA改进算法  

LSTM⁃based improved algorithm of grouping based bit⁃slot ALOHA

作　　者：陈培东 李晓武[1] 李润鑫[1] 贾连印[1] CHEN Peidong;LI Xiaowu;LI Runxin;JIA Lianyin(Faculty of Information Engineering and Automation,Kunming University of Science and Technology,Kunming 650504,China)

机构地区：[1]昆明理工大学信息工程与自动化学院,云南昆明650504

出　　处：《现代电子技术》2023年第15期161-166,共6页Modern Electronics Technique

基　　金：昆明理工大学引进人才科研启动项目(KKSY201603016)。

摘　　要：单阅读器移动射频识别(RFID)系统是一种中小型区域RFID的标签识别系统,传统的标签防碰撞算法在单阅读器移动RFID系统中面临标签丢失、标签碰撞等问题。为改善上述问题,提出一种结合LSTM网络和GBSA(分组比特时隙ALOHA)算法的GBSA改进算法。GBSA算法使用分组策略使答应的标签数在最佳性能区间内,在单阅读器移动RFID系统中GBSA算法存在无法准确计算下一帧分组数的问题,导致阅读器花费大量时间与标签通信反复确认分组数,使识别效率下降,结合LSTM的GBSA改进算法使用LSTM网络解决该问题。LSTM网络使用阅读器识别标签的历史数据预测下一帧中GBSA算法的最佳分组数,通过对分组数的准确预测,提高标签识别率。仿真实验结果表明,结合LSTM的GBSA改进算法可以有效地减少标签丢失,提高识别效率。The single⁃reader mobile radio frequency identification(RFID)system is a tag identification system for RFID in small and medium⁃sized regions.In the use of the traditional tag anti⁃collision algorithm,tag loss and tag collision occur to the single⁃reader mobile RFID system.In view of this,a GBSA improved algorithm based on LSTM network and the GBSA(grouping based bit⁃slot ALOHA)algorithm is proposed.In the GBSA algorithm,a grouping strategy is used to keep the promised tag number within the optimal performance range.In the single⁃reader mobile RFID system,the GBSA algorithm fails to accurately calculate the number of grouping in the next frame,which causes the reader to spend a lot of time communicating with the tag to repeatedly confirm the number of grouping,which reduces the recognition efficiency.Therefore,LSTM network is used to solve the problem in the GBSA improved algorithm based on LSTM.The LSTM network uses the historical data of reader identification tags to predict the optimal grouping number of GBSA algorithm in the next frame.The tag identification rate is increased by accurate prediction of the grouping number.The results of simulation experiments indicate that the GBSA improved algorithm based on LSTM can effectively reduce tag loss and improve identification efficiency.

关 键 词：射频识别 中小型区域 单阅读器移动RFID系统 防碰撞算法 长短期记忆网络 分组预测 仿真实验 标签丢失率 

分 类 号：TN911.1-34[电子电信—通信与信息系统] TP391[电子电信—信息与通信工程]