基于伪事件的RFID数据清洗方法

传统的基于事件驱动的滑动窗口清洗方法只能处理理想的匀速RFID数据流,然而在实际应用中处理非匀速数据流时却会产生较大的输出延迟.为了解决滑动窗口清洗方法的输出延迟和海量中间数据,提出了一种基于伪事件的数据清洗方法PSCleaning.该方法通过在滑动窗口中增加伪事件处理机制,有效地减小了数据的输出延迟;通过对错读和重复读进行的一次性处理,明显缩减了数据量,提高了数据清洗的效率.

RFID Data Cleaning Method Based on Pseudo Event

Traditional RFID data cleaning methods based on event-driven sliding window are clever at handling ideal contant-speed RFID data stream,but in reality the velocity of stream turns out to be unconstant.Thus,the traditional sliding-window methods may cause large time delay.Aiming at dealing with the shortcomings of the traditional sliding-window methods such as time delay and high volume of intermediate data,a data cleaning approach based on pseudo event,called PSCleaning,is proposed.In PSCleaning,time delay of data output is reduced effectively by introducing the notion of pseudo event into sliding window and the volume of output is obviously decreased by handling false positive and duplicate readings at the same time,thus the efficiency of data clean is improved.