WiFi fingerprint positioning method based on fusion of autoencoder and stacking mode

Guan, Junlin; Xin, Zhi; Wang, Huadeng; Yang, Lan

doi:10.1109/iccst50977.2020.00075

Cited by 14 publications

(9 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They train multiple weak learners (decision trees) by using the boosting or bagging method on the training set, and they integrate these learners into a strong learner for the final prediction. In recent years, deep learning (DL) has been increasingly used in the Wi-Fi-based positioning and navigation algorithms [ 14 , 15 , 16 , 17 , 18 , 19 , 32 ], such as DeepFi [ 32 ] and WiDeep [ 14 ], which adopt the channel state information (CSI) or the RSSI data from all subcarriers to train a neural network with more layers than the MLP network. Autoencoder (AE) is a popular network structure for feature extraction and is widely applied into these algorithms by developing different variants such as denoising AE.…”

Section: Related Workmentioning

confidence: 99%

“…In recent years, many machine learning (ML) algorithms [ 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 ] have been applied to the fingerprint-based positioning system. They could learn useful knowledge from multi-dimensional measured data with position labels to reduce the effect of RSSI fluctuation and improve fingerprinting accuracy and system robustness.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Wi-Fi-Based Indoor Localization and Navigation: A Robot-Aided Hybrid Deep Learning Approach

Lin

Gan

Jiang

et al. 2023

Sensors

View full text Add to dashboard Cite

Indoor localization and navigation have become an increasingly important problem in both industry and academia with the widespread use of mobile smart devices and the development of network techniques. The Wi-Fi-based technology shows great potential for applications due to the ubiquitous Wi-Fi infrastructure in public indoor environments. Most existing approaches use trilateration or machine learning methods to predict locations from a set of annotated Wi-Fi observations. However, annotated data are not always readily available. In this paper, we propose a robot-aided data collection strategy to obtain the limited but high-quality labeled data and a large amount of unlabeled data. Furthermore, we design two deep learning models based on a variational autoencoder for the localization and navigation tasks, respectively. To make full use of the collected data, a hybrid learning approach is developed to train the models by combining supervised, unsupervised and semi-supervised learning strategies. Extensive experiments suggest that our approach enables the models to learn effective knowledge from unlabeled data with incremental improvements, and it can achieve promising localization and navigation performance in a complex indoor environment with obstacles.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Wi-Fi-Based Indoor Localization and Navigation: A Robot-Aided Hybrid Deep Learning Approach

Lin

Gan

Jiang

et al. 2023

Sensors

View full text Add to dashboard Cite

show abstract

“…There is a trend in using DL models for fingerprint matching due to their modeling capabilities giving high localization performance. Several WiFi approaches used CNNs [69]- [73], [75], [77], where the RSS samples are either treated as images or time series. Ibrahim et al [71] proposed using consecutive WiFi RSS readings to form a time series processed using a CNN to exploit temporal dependencies.…”

Section: B Wifi Based Systemsmentioning

confidence: 99%

“…For instance, Kim et al [68] trained an SAE network and then replaced the decoder part with a fully-connected classifier network to estimate the floor and the location of the user sequentially. JunLin et al [69] used SAE with stacking model fusion of ELM, CNN, XGBoost, and SVM to perform floor localization. Another way to tackle dimensionality reduction is to use Principal Component Analysis (PCA), such as the proposed system in [115] by Qi et al, where PCA is used with an ensemble of ELM models for floor localization.…”

Section: B Wifi Based Systemsmentioning

confidence: 99%

A Survey of Indoor Localization Systems in Multi-Floor Environments

Mostafa¹,

Harras²,

Youssef³

2022

Preprint

View full text Add to dashboard Cite

This work presents a survey for the indoor localization systems handling multi-floor environments, including floor localization, and estimating the floor and the 2D location of the user. It focuses on approaches targetting users and user devices. The survey comprehensively covers various techniques and technologies focusing especially on the recent solutions to give insights to the researchers for further development in the domain of indoor localization. Finally, several challenges are addressed and possible solutions are proposed pointing to future research directions. No data was collected for this work.

show abstract

“…Generally, AEs are composed of two main submodels: the encoder, which is responsible for dimensionality reduction or data compression, and the decoder, which attempts to reconstruct the input data from the compressed or encoded representation. For instance, in [22], the authors used a selfencoder to reduce the dimensionality of IEEE 802.11 Wireless LAN (Wi-Fi) radio maps as well as a stacked model with the random-forest algorithm in order to reduce the positioning error. Consequently, the authors were able to improve the floor hit rate by more than 3% in comparison with the baselines.…”

mentioning

confidence: 99%

Autoencoder Extreme Learning Machine for Fingerprint-Based Positioning: A Good Weight Initialization is Decisive

Quezada-Gaibor

Klus

et al. 2023

J. Ind. Sea. Pos. Nav.

View full text Add to dashboard Cite

Indoor positioning based on machine-learning (ML) models has attracted widespread interest in the last few years, given its high performance and usability. Supervised, semisupervised, and unsupervised models have thus been widely used in this field, not only to estimate the user position, but also to compress, clean, and denoise fingerprinting datasets. Some scholars have focused on developing, improving, and optimizing ML models to provide accurate solutions to the end user. This article introduces a novel method to initialize the input weights in autoencoder extreme learning machine (AE-ELM), namely factorized input data (FID), which is based on the normalized form of the orthogonal component of the input data. AE-ELM with FID weight initialization is used to efficiently reduce the radio map. Once the dimensionality of the dataset is reduced, we use k-nearest neighbors to perform the position estimation. This research work includes a comparative analysis with several traditional ways to initialize the input weights in AE-ELM, showing that FID provide a significantly better reconstruction error. Finally, we perform an assessment with 13 indoor positioning datasets collected from different buildings and in different countries. We show that the dimensionality of the datasets can be reduced more than 11 times on average, while the positioning error suffers only a small increment of 15% (on average) in comparison to the baseline.

show abstract

WiFi fingerprint positioning method based on fusion of autoencoder and stacking mode

Cited by 14 publications

References 13 publications

Wi-Fi-Based Indoor Localization and Navigation: A Robot-Aided Hybrid Deep Learning Approach

Wi-Fi-Based Indoor Localization and Navigation: A Robot-Aided Hybrid Deep Learning Approach

A Survey of Indoor Localization Systems in Multi-Floor Environments

Autoencoder Extreme Learning Machine for Fingerprint-Based Positioning: A Good Weight Initialization is Decisive

Contact Info

Product

Resources

About