TY - JOUR
T1 - Robust indoor localization based on multi-modal information fusion and multi-scale sequential feature extraction
AU - Wang, Qinghu
AU - Jia, Jie
AU - Chen, Jian
AU - Deng, Yansha
AU - Wang, Xingwei
AU - Aghvami, Abdol Hamid
N1 - Funding Information:
This work was supported in part by the Major Research Plan of National Natural Science Foundation of China under Grant No. 92167103 , in part by the National Natural Science Foundation of China under Grants No. 62172084 , 62132004 , 61972079 , in part by the LiaoNing Key Research and Development Program under Grant No. 2023JH2/101300196 , in part by the Central Government Guided Local Science and Technology Development Fund Project under Grant No. 2020ZY0003 , in part by the Fundamental Research Funds for the Central Universities under Grants No. N2324004-12 , N2216009 , N2216006 , N2116004 , in part by the LiaoNing Revitalization Talents Program under Grant No. XLYC2007162 , in part by the Science and Technology Plan Project of Inner Mongolia Autonomous Region of China under Grant No. 2020GG0189 , in part by the Natural Science Foundation of Inner Mongolia Autonomous Region under Grant No. 2022MS06029 , in part by the industry-University-Research-Application Innovation Team of Inner Mongolia Minzu University for Smart City and Village Construction Empowered by Information Technology .
Publisher Copyright:
© 2024
PY - 2024/6
Y1 - 2024/6
N2 - Magnetic-assisted indoor localization has attracted significant attention because of its commercial and social values. However, it is challenging to construct a robust and accurate system due to the severe feature ambiguity caused by different users, mobiles, attitudes, and moving speeds. In order to cope with this issue, we first propose to fuse magnetic with the multi-modal features, including Bluetooth low energy (BLE), and context information of continuous predictions, to improve the feature discrimination with more valuable localization clues. Then, we extract more orientation-insensitive magnetic features and remove the Direct Current (DC) component of the sequence can reduce the feature ambiguity caused by different holding attitudes, devices, and users. After that, we propose an online data augmentation algorithm to automatically generate a sufficient amount of various-speed sequences based on only one dense sampling benchmark sequence, thus reducing the influence of multi-scale sequences caused by different moving speeds. Finally, we propose a multi-branch and attention mechanism-based end-to-end localization model to extract and efficiently fuse the significant features of the multi-modal data for accurate localization. We evaluate the performance of the proposed localization system (DamLoc) in a typical indoor environment based on extensive experiments. Evaluation results showcase that DamLoc more robust for diverse heterogeneous factors, and can support about 63% improvement compared to state-of-the-art methods. It is worth pointing out that the BLE in our work can be replaced with other signals, such as Wireless Fidelity (Wi-Fi), which is more general than other fusion-based localization.
AB - Magnetic-assisted indoor localization has attracted significant attention because of its commercial and social values. However, it is challenging to construct a robust and accurate system due to the severe feature ambiguity caused by different users, mobiles, attitudes, and moving speeds. In order to cope with this issue, we first propose to fuse magnetic with the multi-modal features, including Bluetooth low energy (BLE), and context information of continuous predictions, to improve the feature discrimination with more valuable localization clues. Then, we extract more orientation-insensitive magnetic features and remove the Direct Current (DC) component of the sequence can reduce the feature ambiguity caused by different holding attitudes, devices, and users. After that, we propose an online data augmentation algorithm to automatically generate a sufficient amount of various-speed sequences based on only one dense sampling benchmark sequence, thus reducing the influence of multi-scale sequences caused by different moving speeds. Finally, we propose a multi-branch and attention mechanism-based end-to-end localization model to extract and efficiently fuse the significant features of the multi-modal data for accurate localization. We evaluate the performance of the proposed localization system (DamLoc) in a typical indoor environment based on extensive experiments. Evaluation results showcase that DamLoc more robust for diverse heterogeneous factors, and can support about 63% improvement compared to state-of-the-art methods. It is worth pointing out that the BLE in our work can be replaced with other signals, such as Wireless Fidelity (Wi-Fi), which is more general than other fusion-based localization.
KW - BLE
KW - Data augmentation
KW - Feature fusion
KW - Magnetic sequence
KW - Multi-branch localization
KW - Multi-modal information
UR - http://www.scopus.com/inward/record.url?scp=85185274554&partnerID=8YFLogxK
U2 - 10.1016/j.future.2024.02.002
DO - 10.1016/j.future.2024.02.002
M3 - Article
AN - SCOPUS:85185274554
SN - 0167-739X
VL - 155
SP - 164
EP - 178
JO - FUTURE GENERATION COMPUTER SYSTEMS
JF - FUTURE GENERATION COMPUTER SYSTEMS
ER -