摘要
基于动态帧时隙ALOHA算法,运用拟牛顿法,使用某种近似矩阵代替牛顿法中的Hessian矩阵,解决牛顿法中复杂度计算的问题,由此缩短计算时间并提高计算精度。统计出不同时隙数下可识别的标签数,进而建立专家库,使读写器根据标签数目更准确地设定时隙数,达到全局搜索的目的,进而缩短匹配识别时间。仿真结果表明,运用拟牛顿法明显提高了数据交换量和识别效率。
On the basis of dynamic frame slotted ALOHA algorithm,the Hessian matrix in the Newton method is replaced by a certain approximate matrix in quasi-Newton method to solve the problem of complexity calculation in Newton method,which can shorten the computation time and improve the calculation precision.The number of identifiable tags under different time slots is calculated to establish the expert library,so as to make the reader-writer set the number of time slots more accurately according to the number of tags,achieve the purpose of global search,and shorten the matching and recognition time.The simulation results show that the quasi-Newton method can improve the data exchange capacity and recognition efficiency effectively.
作者
陈夫桂
姜志峰
朱利娟
云中华
CHEN Fugui;JIANG Zhifeng;ZHU Lijuan;YUN Zhonghua(Office of Academic Affairs,Tibet University,Lhasa 850012,China;School of Engineering,Tibet University,Lhasa 850012,China;Tibetan Information Technology Research Center,Tibet University,Lhasa 850012,China)
出处
《现代电子技术》
北大核心
2018年第15期97-100,104,共5页
Modern Electronics Technique
基金
西藏自治区自然科学基金(2016ZR-15-8)
西藏自治区教育科学十三五规划2016年度课题(XZJKY416012)
西藏大学2017教学研究与改革项目(XZDXJXYJ201722)
西藏大学中青年教师队伍建设项目(2016教学能力提升)
西藏自治区高校青年教师创新支持计划项目(QCZ2016-26)~~
关键词
ALOHA
拟牛顿法
近似矩阵
专家库
全局搜索
识别效率
ALOHA
quasi-Newton method
approximate matrix
expert library
global search
recognition efficiency
