Research And Development Of Data Processing For Network RFID

Considering the multi-reader deployment in large scale RFID implementation, readers are not working independently but collaborate with each other with some logical relation in a certain context, thus they form a reader network. Data processing methods before cannot meet the requirements of distributed RFID network.This paper performs an in-depth analysis of RFID data processing, puts forward a component based processing solution. Firstly, we solve the problems of duplicate readings through redundant filter. Two kinds of redundant filter algorithms are introduced, they are designed for different situation, one is real-time redundancy filtering mechanism for out of order data cleaning online, and the other is delay redundancy filtering mechanism for total order. Secondly,data smoothing mechanism in three dimension is adopted to make sure the integrality of RFID data, they are time smooth for data from a physical reader, and space smooth for data from a logical reader, combing integrality-based restraint method to correct for lost and error reading. Lastly, real-time memory database is used to satisfy the demands of data cleaning online.The contents of this paper are organized as follows: Firstly, we summarize the technologies which will be used in RFID data processing; analyze some standards issued by international organizations, and architecture of the RFID middleware and the data processing solutions put forward by some big companies. Secondly, through analyzing the traits of RFID data processing as well as the issues encountered, we summarize the function architecture of the reader management modules in RFID middleware. And then we come up with RFID data cleaning algorithm and real-time memory database. Finally, we introduced our RFID data processing architecture and implementation solution.