TY - GEN
T1 - Accurate Indoor Localization Using Magnetic Sequence Fingerprints with Deep Learning
AU - Ding, Xuedong
AU - Zhu, Minghua
AU - Xiao, Bo
N1 - Publisher Copyright:
© 2022, Springer Nature Switzerland AG.
PY - 2022
Y1 - 2022
N2 - Magnetic field fingerprinting has been an interesting topic in indoor localization researches because of its advantages of being ubiquitous, energy-efficient and infrastructure-free. Most existing indoor magnetic field-based positioning methods use the raw three-dimensional magnetic field strength obtained by the magnetic sensor built in smartphones. However, they have to overcome the problem of ambiguity that originates from the nature of geomagnetic data, especially in the large-scale environment. In this paper, we first expand the dimension of magnetic data elements, and a sliding window mechanism is designed to construct magnetic sequence fingerprints to increase the distinguishability of magnetic field fingerprints. Moreover, an accurate indoor positioning model combining the advantages of one-dimensional convolutional neural network and long short-term memory network is designed to automatically learn the mapping between ground-truth positions and magnetic sequence fingerprints. To demonstrate the effectiveness of our proposed method, we perform a comprehensive experimental evaluation on three real-world datasets, and the results show that the proposed approach can remarkably improve positioning performance compared with other methods.
AB - Magnetic field fingerprinting has been an interesting topic in indoor localization researches because of its advantages of being ubiquitous, energy-efficient and infrastructure-free. Most existing indoor magnetic field-based positioning methods use the raw three-dimensional magnetic field strength obtained by the magnetic sensor built in smartphones. However, they have to overcome the problem of ambiguity that originates from the nature of geomagnetic data, especially in the large-scale environment. In this paper, we first expand the dimension of magnetic data elements, and a sliding window mechanism is designed to construct magnetic sequence fingerprints to increase the distinguishability of magnetic field fingerprints. Moreover, an accurate indoor positioning model combining the advantages of one-dimensional convolutional neural network and long short-term memory network is designed to automatically learn the mapping between ground-truth positions and magnetic sequence fingerprints. To demonstrate the effectiveness of our proposed method, we perform a comprehensive experimental evaluation on three real-world datasets, and the results show that the proposed approach can remarkably improve positioning performance compared with other methods.
KW - Deep learning
KW - Indoor localization
KW - Magnetic field
KW - Magnetic sequence fingerprints
KW - Smartphone
UR - https://www.scopus.com/pages/publications/85126217697
U2 - 10.1007/978-3-030-95384-3_5
DO - 10.1007/978-3-030-95384-3_5
M3 - 会议稿件
AN - SCOPUS:85126217697
SN - 9783030953836
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 65
EP - 84
BT - Algorithms and Architectures for Parallel Processing - 21st International Conference, ICA3PP 2021, Proceedings
A2 - Lai, Yongxuan
A2 - Wang, Tian
A2 - Jiang, Min
A2 - Xu, Guangquan
A2 - Liang, Wei
A2 - Castiglione, Aniello
PB - Springer Science and Business Media Deutschland GmbH
T2 - 21st International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 2021
Y2 - 3 December 2021 through 5 December 2021
ER -