tailieunhanh - Deploying RFID Challenges Solutions and Open Issues Part 2

Tham khảo tài liệu 'deploying rfid challenges solutions and open issues part 2', kỹ thuật - công nghệ, cơ khí - chế tạo máy phục vụ nhu cầu học tập, nghiên cứu và làm việc hiệu quả | The Challenges and Issues Facing the Deployment of RFID Technology 17 The P2P Collaboration method proposed by Peng Ji Luo Wong and Tan Peng et al. 2008 is an approach utilising Peer-to-Peer P2P networks within the RFID data set to detect and remove inaccurate readings. The system works by breaking the readings into detection nodes which are constantly sending and receiving messages. From these transmitted messages false negatives and false positives are able to be detected and corrected resulting in a cleaner data set. Ziekow and Ivantysynova have presented a method designed to correct RFID anomalies probabilistically by employing maximum likelihood operations Ziekow Ivantysynova 2008 . Their method utilises the position of a tag which may be determined by measuring properties associated with the Radio Frequency signal. The Cost-Conscious cleaning method is a cleaning algorithm which utilises a Bayesian Network to judge the likelihood that read tags correctly depict reality when based upon the previously read tags Gonzalez et al. 2007 . The Cost-Conscious cleaning approach houses several different cleaning algorithms and chooses the least costly algorithm which would offer the highest precision in correcting the raw data. A similar approach has also been proposed that utilises a Bayesian Network to judge the existence of tags scanned Floerkemeier 2004 . It lacks however the cost-saving analysis that would increase the speed of the clean. Data Mining Techniques refer to the use of mining past data to detect inaccuracies and possible solutions to raw RFID readings. A study which has used data mining techniques extensively to correct the entire data set table is the Deferred Rule Based Approach proposed in Rao et al. 2006 . The architecture of the system is reliant on the user defining rules which are utilised to determine anomalies in the data set and possibly to correct them. Probabilistic Inference refers to a process by which the in-coming data node will be .