Premium
Search me if you can: Multiple mix zones with location privacy protection for mapping services
Author(s) -
Memon Imran,
Arain Qasim Ali,
Memon Muhammad Hammad,
Mangi Farman Ali,
Akhtar Rizwan
Publication year - 2017
Publication title -
international journal of communication systems
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.344
H-Index - 49
eISSN - 1099-1131
pISSN - 1074-5351
DOI - 10.1002/dac.3312
Subject(s) - computer science , location based service , point of interest , usability , service (business) , mobile device , computer security , popularity , global positioning system , world wide web , computer network , telecommunications , human–computer interaction , business , artificial intelligence , psychology , social psychology , marketing
Summary The mobile vehicle is gaining popularity nowadays using map services like Google Maps and other mapping services. However, map services users have to expose sensitive information like geographic locations (GPS coordinates) or address to personal privacy concerns as users share their locations and queries to obtain desired services. Existing mix zones location privacy protection methods are most general purposed and theoretical value while not applicable when applied to provide location privacy for map service users. In this paper, we present new (multiple mix zones location privacy protection) MMLPP method specially designed for map services on mobile vehicles over the road network. This method enables mobile vehicle users to query a route between 2 endpoints on the map, without revealing any confidential location and queries information. The basic idea is to strategically endpoints to nearby ones, such that (1) the semantic meanings encoded in these endpoints (eg, their GPS coordinates) change much, ie, location privacy is protected; (2) the routes returned by map services little change, ie, services usability are maintained. Specifically, a mobile client first privately retrieves point of interest close to the original endpoints, and then selects 2 points of interest as the shifted endpoints satisfying the property of geoindistinguishability. We evaluate our MMLPP approach road network application for GTMobiSim on different scales of map services and conduct experiments with real traces. Results show that MMLPP strikes a good balance between location privacy and service usability.