Location:Home > Engineering science > Computer Science > Applied Computer Technology > BSpace: A Data Cleaning Approach for RFID Data Streams Based on Virtual Spatial Granularity

BSpace: A Data Cleaning Approach for RFID Data Streams Based on Virtual Spatial Granularity

Downloads: []
Tutor: SongBaoYan
School: Liaoning University
Course: Applied Computer Technology
Keywords: RFID,Data cleansing,Virtual spatial granularity,Bayesian estimation of
CLC: TP391.44
Type: Master's thesis
Year:  2009
Facebook Google+ Email Gmail Evernote LinkedIn Twitter Addthis

not access Image Error Other errors

With the advances in microelectronics , computer and network technology in recent years , the scope and depth of the RFID technology applications have been developed rapidly . RFID ( radio frequency identification ) technology is a non-contact automatic identification technology , this technology uses radiofrequency reader to capture moving items on the tag information , so as to achieve the purpose of identification and tracking items . But one of the main factors that affect the development of RFID technology is the unreliability of readers capture tag data ( raw data ) , such reliability can be categorized as follows: positive reading negative reading and redundant read . To effectively support the RFID the senior business logic processing , it is necessary to provide high-quality RFID data , thus cleaning RFID raw data is very necessary . RFID data cleaning techniques , in-depth research , a virtual space the granularity cleaning method bSpace specific work are as follows : for the sliding window as well as the limitations of traditional spatial granularity , the paper first proposed the concept of virtual spatial granularity given virtual space granularity matrix solving method ; in the cleaning process , according to the real-time observation on the reader to read the information , the proposed algorithm dynamically adjust the virtual space granularity . In order to achieve positive reading of the filter , the paper proposes a virtual spatial granularity , and constraint rules - based processing strategies ; To achieve negative reading to fill , label into the detection range of the reader , this paper proposes a strategy based on Bayesian probability estimates , effectively estimate the number of times the label is detected ; using different label on the same virtual space granularity period reads the characteristics of the same information , they do \allocation strategy , this paper presents the first intermediate , after both sides \the information read in the label are considered redundant filtered off. In this paper, experimental validation the bSpace the rationality and effectiveness of the method , the bSpace method effective RFID data cleaning .
Related Dissertations
Last updated
Sponsored Links
Home |About Us| Contact Us| Feedback| Privacy | copyright | Back to top