Robust 3D indoor localization via graph optimization: A hybrid RSSD fingerprinting approach with fuzzy-correlation hyperplane constraints

Abstract In the era of ubiquitous wireless connectivity, precise positioning technology for unknown signal sources (USS) is playing an increasingly important role. Received signal strength difference (RSSD) can reduce the influence of common transmitter-side parameters, particularly unknown transmission power, and has therefore been widely applied to unknown signal source localization. However, measurement noise amplified during the RSSD calculation process can cause distortion in the mapping relationship between RSSD and spatial position, limiting positioning accuracy. Existing three-dimensional RSSD-based localization methods still face challenges associated with noise-amplified fingerprint matching, statistical dependence among pairwise RSSD observations, and the numerical stability of locally fitted observation constraints. The increase in dimensions not only introduces uncertainty in the height domain, but also causes the signal propagation model and computational complexity to increase non-linearly, making three-dimensional USS positioning both extremely challenging and highly valuable. By introducing graph optimization theory into the three-dimensional positioning scenario, this paper constructs a three-dimensional positioning model based on factor graph to achieve high-precision positioning of indoor USS. Firstly, a fuzzy weight mechanism is introduced to enhance the denoising of online RSSD data, and grey correlation analysis is employed to identify the reference points most correlated with the online measurement, thereby constructing the sub-positioning region. Secondly, an observation model for the hyperplane mapping between RSSD and three-dimensional spatial coordinates is derived and established, and a three-dimensional factor graph optimization framework is constructed. Finally, the belief propagation is carried out in the proposed positioning model using the sum-product algorithm, and the iterative update and fusion reasoning of the soft information are employed to ultimately calculate the three-dimensional location coordinates of the target. Simulation and physical experimental results demonstrate that the proposed method achieves improved localization accuracy compared with K-Nearest Neighbors (KNN), Support Vector Machine (SVM), and Attention Mechanism (ATT) under the evaluated conditions.

Authors

Institutions

Publication Details

Journal
Measurement Science and Technology
Published
2026-11-17
DOI
https://doi.org/10.1088/1361-6501/aec31c
Primary Topic
Indoor and Outdoor Localization Technologies
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Robust 3D indoor localization via graph optimization: A hybrid RSSD fingerprinting approach with fuzzy-correlation hyperplane constraints

Zhiyou Pan, Liyang Zhang, Rui Lin Gao, Kai Cheng et al.
3 links
Measurement Science and Technology
Indoor and Outdoor Localization Technologies
article

Robust 3D indoor localization via graph optimization: A hybrid RSSD fingerprinting approach with fuzzy-correlation hyperplane constraints

Zhiyou Pan, Liyang Zhang, Rui Lin Gao, Kai Cheng, Yan He, Keying Zhu
article en

Abstract

Abstract In the era of ubiquitous wireless connectivity, precise positioning technology for unknown signal sources (USS) is playing an increasingly important role. Received signal strength difference (RSSD) can reduce the influence of common transmitter-side parameters, particularly unknown transmission power, and has therefore been widely applied to unknown signal source localization. However, measurement noise amplified during the RSSD calculation process can cause distortion in the mapping relationship between RSSD and spatial position, limiting positioning accuracy. Existing three-dimensional RSSD-based localization methods still face challenges associated with noise-amplified fingerprint matching, statistical dependence among pairwise RSSD observations, and the numerical stability of locally fitted observation constraints. The increase in dimensions not only introduces uncertainty in the height domain, but also causes the signal propagation model and computational complexity to increase non-linearly, making three-dimensional USS positioning both extremely challenging and highly valuable. By introducing graph optimization theory into the three-dimensional positioning scenario, this paper constructs a three-dimensional positioning model based on factor graph to achieve high-precision positioning of indoor USS. Firstly, a fuzzy weight mechanism is introduced to enhance the denoising of online RSSD data, and grey correlation analysis is employed to identify the reference points most correlated with the online measurement, thereby constructing the sub-positioning region. Secondly, an observation model for the hyperplane mapping between RSSD and three-dimensional spatial coordinates is derived and established, and a three-dimensional factor graph optimization framework is constructed. Finally, the belief propagation is carried out in the proposed positioning model using the sum-product algorithm, and the iterative update and fusion reasoning of the soft information are employed to ultimately calculate the three-dimensional location coordinates of the target. Simulation and physical experimental results demonstrate that the proposed method achieves improved localization accuracy compared with K-Nearest Neighbors (KNN), Support Vector Machine (SVM), and Attention Mechanism (ATT) under the evaluated conditions.

Measurement Science and Technology
Tianjin Chengjian University (CN), China Automotive Technology and Research Center (CN)
National Natural Science Foundation of China
Indoor and Outdoor Localization Technologies
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon