Semi-supervised Variational Autoencoder for WiFi Indoor Localization

We address the problem of indoor localization based on WiFi signal strengths. We develop a semi-supervised deep learning method able to train a prediction model from a small set of annotated WiFi observations and a massive set of non-annotated ones. Our method is based on the variational au-toencoder deep network. We complement the network with an additional component of structural projection able to further improve the localization accuracy in a complex, multi-building and multi-floor environment. We consider several different network compositions which combine the classification and regression sub-tasks to achieve optimal performance. We evaluate our method on the public UJI-IndoorLoc dataset and show that the proposed method allows to maintain the state of the art localization accuracy with a very limited amount of annotated data.

[1]  Michal R. Nowicki,et al.  Low-effort place recognition with WiFi fingerprints using deep learning , 2016, AUTOMATION.

[2]  Doina Precup,et al.  Real-Time Indoor Localization in Smart Homes Using Semi-Supervised Learning , 2017, AAAI.

[3]  Max Welling,et al.  Semi-supervised Learning with Deep Generative Models , 2014, NIPS.

[4]  Dongsoo Han,et al.  Unsupervised Learning for Crowdsourced Indoor Localization in Wireless Networks , 2016, IEEE Transactions on Mobile Computing.

[5]  Kin K. Leung,et al.  A Survey of Indoor Localization Systems and Technologies , 2017, IEEE Communications Surveys & Tutorials.

[6]  Adolfo Martínez Usó,et al.  UJIIndoorLoc: A new multi-building and multi-floor database for WLAN fingerprint-based indoor localization problems , 2014, 2014 International Conference on Indoor Positioning and Indoor Navigation (IPIN).

[7]  Hichem Snoussi,et al.  SVM-based indoor localization in Wireless Sensor Networks , 2017, 2017 13th International Wireless Communications and Mobile Computing Conference (IWCMC).

[8]  Yinfeng Wu,et al.  A Radio-Map Automatic Construction Algorithm Based on Crowdsourcing , 2016, Sensors.

[9]  Suk-Hoon Jung,et al.  Automated Construction and Maintenance of Wi-Fi Radio Maps for Crowdsourcing-Based Indoor Positioning Systems , 2018, IEEE Access.

[10]  Meng Sun,et al.  Fast Radio Map Construction by using Adaptive Path Loss Model Interpolation in Large-Scale Building , 2019, Sensors.

[11]  Hsin-Piao Lin,et al.  Applying Deep Neural Network (DNN) for Robust Indoor Localization in Multi-Building Environment , 2018, Applied Sciences.

[12]  Ayhan Demiriz,et al.  Semi-Supervised Support Vector Machines , 1998, NIPS.

[13]  Taehoon Lee,et al.  Semi-supervised Learning with Deep Generative Models for Asset Failure Prediction , 2017, ArXiv.

[14]  Hamed Haddadi,et al.  Deep Learning in Mobile and Wireless Networking: A Survey , 2018, IEEE Communications Surveys & Tutorials.

[15]  Yi Liu,et al.  Indoor Fingerprint Positioning Based on Wi-Fi: An Overview , 2017, ISPRS Int. J. Geo Inf..

[16]  Mohsen Guizani,et al.  Semisupervised Deep Reinforcement Learning in Support of IoT and Smart City Services , 2018, IEEE Internet of Things Journal.

[17]  Karl Henrik Johansson,et al.  Semi-supervised learning for mobile robot localization using wireless signal strengths , 2017, 2017 International Conference on Indoor Positioning and Indoor Navigation (IPIN).

[18]  Ye Yuan,et al.  Efficient WiFi fingerprint training using semi-supervised learning , 2014, 2014 Ubiquitous Positioning Indoor Navigation and Location Based Service (UPINLBS).

[19]  Teemu Roos,et al.  Semi-supervised Learning for WLAN Positioning , 2011, ICANN.

[20]  Günther Retscher,et al.  Location determination using WiFi fingerprinting versus WiFi trilateration , 2007, J. Locat. Based Serv..