一种基于贝叶斯推理的多标签数量估计方法  

A Method of Multi-Label Quantity Estimation based on Bayesian Inference

作　　者：刘婷 陈庆荣 杨弘峰 宋金圣 韦涛[3] LIU Ting;CHEN Qingrong;YANG Hongfeng;SONG Jinsheng;WEI Tao(Unit 93114 of PLA,Beijing 100085,China;University of Electronic Science and Technology of China,Chengdu Sichuan 611731,China;No.30 Institute of CETC,Chengdu Sichuan 610041,China)

机构地区：[1]中国人民解放军93114部队,北京100085 [2]电子科技大学,四川成都611731 [3]中国电子科技集团公司第三十研究所,四川成都610041

出　　处：《通信技术》2021年第5期1179-1183,共5页Communications Technology

基　　金：宜宾市科技项目(No.2018ZSF001);教育部-中国移动科研基金(No.MCM20180104)。

摘　　要：针对射频识别(Radio Frequency Identification,RFID)动态帧时隙防碰撞算法中,初始帧长选择比较敏感的问题,以及RFID动态帧时隙提前停止算法的问题进行了研究,提出了一种基于贝叶斯推断的标签数量估计算法。该算法可以有效利用历史时隙的状态信息,在每个时隙对标签数量进行一次估计。估计的精确度随着时隙数量的增多逐渐增加。仿真结果显示,该算法可以有效减少动态帧时隙,提前停止算法中观察时隙的数量,从而降低了初始帧长对系统效率的影响。In the RFID dynamic frame time slot anti-collision algorithm,the sensitive issue of the initial frame length selection and the early stop algorithm of the RFID dynamic frame time slot are explored,and an algorithm for estimating the number of tags based on Bayesian inference is proposed.The algorithm can effectively use the status information of historical time slots,and the number of tags is estimated once in each time slot.The accuracy of the estimation gradually increases as the number of time slots increases.Simulation results indicate that the algorithm can effectively reduce the number of observation time slots in the dynamic frame time slot early stop algorithm,thereby reducing the impact of the initial frame length on system efficiency.

关 键 词：贝叶斯推理 RFID防碰撞 动态帧时隙 算法复杂度 

分 类 号：TN929[电子电信—通信与信息系统]