Title: RFID data cleaning method in heterogeneous space based on linear probabilistic motion state model

Authors: Shuming Wang

Addresses: Hubei Province Company of China National Tobacco Corporation, Wuhan 430000, China

Abstract: In order to overcome the problems of low efficiency, low precision and high load of heterogeneous spatial data cleaning methods, this paper proposes a heterogeneous spatial RFID data cleaning method based on linear probabilistic motion state model. Combining with the infinity of RFID data stream, this method uses sliding window technology to smooth the spatial tag data, and effectively removes the noise in the data. The Bernoulli binomial distribution is used to model the RFID data stream. At the same time, a probabilistic motion model for RFID tags is introduced. The transformation relationship between RFID initial data and tag motion state information is established. The vulnerability data of heterogeneous space is filled by tag motion state information, and heterogeneous spatial data cleaning is realised. The experimental results show that the operation efficiency is always above 95%, the highest data cleaning accuracy is 99.01%, and the lowest operating load is 0.038%.

Keywords: RFID data; probabilistic motion model; heterogeneous space; data cleaning.

DOI: 10.1504/IJICT.2022.124330

International Journal of Information and Communication Technology, 2022 Vol.21 No.1, pp.79 - 92

Received: 21 Aug 2020
Accepted: 26 Oct 2020
Published online: 22 Jul 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article