A low-complexity trajectory privacy preservation approach for indoor fingerprinting positioning systems

Sazdar, Amir Mahdi, Ghorashi, Seyed Ali ORCID: https://orcid.org/0000-0002-2910-9208, Moghtadaiee, Vahideh, Khonsari, Ahmad and Windridge, David ORCID: https://orcid.org/0000-0001-5507-8516 (2020) A low-complexity trajectory privacy preservation approach for indoor fingerprinting positioning systems. Journal of Information Security and Applications, 53 , 102515. ISSN 2214-2126 [Article] (doi:10.1016/j.jisa.2020.102515)

[img]
Preview
PDF - Final accepted version (with author's formatting)
Available under License Creative Commons Attribution-NonCommercial-NoDerivatives 4.0.

Download (4MB) | Preview
[img] PDF - Draft pre-submission version (with author's formatting)
Restricted to Repository staff and depositor only

Download (4MB)

Abstract

Location fingerprinting is a technique employed when Global Positioning System (GPS) positioning breaks down within indoor environments. Since Location Service Providers (LSPs) would implicitly have access to such information, preserving user privacy has become a challenging issue in location estimation systems. This paper proposes a low-complexity k-anonymity approach for preserving the privacy of user location and trajectory, in which real location/trajectory data is hidden within k fake locations/trajectories held by the LSP, without degrading overall localization accuracy. To this end, three novel location privacy preserving methods and a trajectory privacy preserving algorithm are outlined. The fake trajectories are generated so as to exhibit characteristics of the user’s real trajectory. In the proposed method, no initial knowledge of the environment or location of the Access Points (APs) is required in order for the user to generate the fake location/trajectory. Moreover, the LSP is able to preserve privacy of the fingerprinting database from the users. The proposed approaches are evaluated in both simulation and experimental testing, with the proposed methods outperforming other well-known k-anonymity methods. The method further exhibits a lower implementation complexity and higher movement similarity (of up to 88%) between the real and fake trajectories.

Item Type: Article
Keywords (uncontrolled): Location privacy-preserving, Trajectory privacy-preserving, Fingerprinting positioning, k-anonymity
Research Areas: A. > School of Science and Technology > Computer Science
Item ID: 33252
Notes on copyright: © 2020. This manuscript version is made available under the CC-BY-NC-ND 4.0 license.
Useful Links:
Depositing User: David Windridge
Date Deposited: 14 May 2021 08:17
Last Modified: 15 Jul 2021 02:26
URI: https://eprints.mdx.ac.uk/id/eprint/33252

Actions (login required)

View Item View Item

Statistics

Downloads
Activity Overview
7Downloads
27Hits

Additional statistics are available via IRStats2.