Gradient Boost Decision Tree Fingerprint Algorithm for Wi-Fi Localization

Location Based Services (LBS) is require the indoor and outdoor seamless positioning that providing real-time, stable and high-accuracy localization and navigation. Mobile devices can be positioning by using Wi-Fi signals based on correlation between Wi-Fi signals strength and coordinates. And Wi-Fi signals are common in modern buildings, so there needn’t deploy equipment. But there are still some drawbacks, such as poorly positioning accuracy and too long online computing time during using Wi-Fi signals to localization. For this reason, we proposed a Gradient Boosting Decision Tree (GBDT) fingerprint algorithm for Wi-Fi localization, this algorithm adopt a linear combination of multiple decision trees to obtain an approximate model of the coordinates and received signal strength (RSS). Experiment shows that about 13% increases in positioning accuracy and 65% reduces in online computation time compares with AdaBoost-based algorithm.

[1]  Qin Xiao,et al.  Improved AdaBoost-based fingerprint algorithm for WiFi indoor localization , 2014, 2014 IEEE 7th Joint International Information Technology and Artificial Intelligence Conference.

[2]  Yuanyuan Yang,et al.  A fine-grained indoor fingerprinting localization based on magnetic field strength and channel state information , 2017, Pervasive Mob. Comput..

[3]  Xiangliang Zhang,et al.  An up-to-date comparison of state-of-the-art classification algorithms , 2017, Expert Syst. Appl..

[4]  Paramvir Bahl,et al.  RADAR: an in-building RF-based user location and tracking system , 2000, Proceedings IEEE INFOCOM 2000. Conference on Computer Communications. Nineteenth Annual Joint Conference of the IEEE Computer and Communications Societies (Cat. No.00CH37064).

[5]  Adam C. Winstanley,et al.  Indoor location based services challenges, requirements and usability of current solutions , 2017, Comput. Sci. Rev..

[6]  Kwan-Wu Chin,et al.  A comparison of deterministic and probabilistic methods for indoor localization , 2011, J. Syst. Softw..

[7]  J. Friedman.  Greedy function approximation: A gradient boosting machine. , 2001 .

[8]  Holger Claussen,et al.  Wireless RSSI fingerprinting localization , 2017, Signal Process..

[9]  John P. Morgan,et al.  Universally optimal designs with blocksize $p\times 2$ and correlated observations , 1997 .