RFID中间件数据过滤方法研究

Research on Data Filtering for RFID Middleware

【目的】探讨对RFID数据冗余、脏读以及漏读数据进行过滤,整理出上层应用所需要的有效事件的方法。【方法】运用基于时间和次数阀值、数据校验、TDS数据解析以及基于伯努利概率的动态时间窗等方法进行研究设计。【结果】提出3种数据过滤器:冗余过滤器用于处理多读数据,平滑过滤器用于处理脏读数据,基于伯努利的动态自适应平滑算法的漏读数据处理器处理漏读数据,将这3种过滤器依次组合,组成数据过滤模型,最终构成数据过滤方法。【结论】运用这3种过滤器组成的数据过滤模型,能够很好的对RFID原始数据进行过滤,获得上层应用所需要的有效事件。

[Objective]This paper illustrates a data filtering method to filter redundancy data, dirty read and missed readings of RFID middleware and sort out the effective incident for the upper layer system.[Method]A time-based and number-based filter, a CRC and TDS data analysis filter, and a dynamic time window filter based on Bernoulli probability are used to established the method.[Re-sult]Three filters was constructed, which are redundant data filter for more readings, smoothing fil-ter for dirty readings and dynamic time window filter based on Bernoulli model for missed readings. Finally, the three filters were combined and the RFID data filtering method was proposed.[Conclu-sion]This data filtering is effective in filtering of RFID raw data and obtaining effective incident for the upper layer system.