Home // UBICOMM 2014, The Eighth International Conference on Mobile Ubiquitous Computing, Systems, Services and Technologies // View article

CGSIL: A Viable Training-Free Wi-Fi Localization

Authors:
Han Dinh Nguyen
Thong Minh Doan
Nam Tuan Nguyen

Keywords: search engine; clustering; regional relationship

Abstract:
Localization for indoor environment normally does not use GPS signals since it cannot penetrate through walls and buildings. Instead, many works have focused on using Wi-Fi signals as the mean to locate the position of the mobile devices. However, most of these approaches require a training step to build a Wi-Fi’s map for each location. This requirement practically prevents these approaches from being realistic, since the training step is extremely time-consuming (hundreds of labor hours). Recently, ISIL has been proposed as the first Wi-Fi-based technique that is training-free, in which the localization can be done instantly at any location without the need of training and building Wi-Fi map. ISIL collects from the web the related information of all observable access points and infers the current position based on that. As the first search-based Wi-Fi localization, ISIL removes the unacceptable time-consuming training step. However, it still does not provide adequate accuracy due to the lack of exploiting regional correlation of information returned by the search engine. In this paper, we proposed CGSIL, another kind of search-based Wi-Fi localization that provides the accuracy level of nearly twice as much as ISIL by collaborative filtering and clustering geographic information collected from the search engines. Through experiment results, CGSIL proves to be a feasible replacement for future indoor localization due to its high accuracy and reasonable cost.

Pages: 268 to 274

Copyright: Copyright (c) IARIA, 2014

Publication date: August 24, 2014

Published in: conference

ISSN: 2308-4278

ISBN: 978-1-61208-353-7

Location: Rome, Italy

Dates: from August 24, 2014 to August 28, 2014