【作者】 李晓静； 谷峪； 吕雁飞； 王艳秋； 于戈；

【Author】 Li Xiaojing~1,Gu Yu~1,L(u|¨) Yanfei~2,Wang Yanqiu~1,and Yu Ge~1 1(College of Information Science and Engineering,Northeastern University,Shenyang 110004) 2(Software College,Northeastern University,Shenyang 110004)

【机构】 东北大学信息科学与工程学院； 东北大学软件学院；

【摘要】 由于RFID技术采用的无线射频信号极易受外界条件干扰,导致RFID数据采集过程中漏读现象频频发生,降低了RFID应用中查询结果的准确性.目前解决漏读问题的算法主要是基于数据层的,这种做法会填补许多与查询无关的数据,浪费系统资源.在逻辑区域层的基础上提出了两种基于动态事件概率模型的数据清洗算法,即贪婪算法和相似度算法.它们主要对已知的数据进行学习来预测未知的数据.大量实验证明了该数据清洗算法的有效性和高效性.更多还原

【Abstract】 Since wireless radio frequency signal adopted by RFID technology is easily interfered with environment,there are many missed readings on collecting RFID data,which reduce the accuracy of query results in RFID applications.Existing RFID data cleaning algorithms for solving reads-missing problems are mainly based on data level,and this solution may interpolate many data that the query doesn’t care about,inducing the waste of system resource.Proposed in this paper are two data cleaning methods on the basis of dynamic event probabilistic model on logic area level,namely rapacity algorithm and similar degree algorithm.They mainly study known data in order to forecast incoming data.Abundant experiments prove the effectiveness and efficiency of the proposed data cleaning algorithms.更多还原