Date of Award

Fall 2008

Degree Type

Thesis - Restricted

Degree Name

Master of Science (MS)

Department

Mathematics, Statistics and Computer Science

First Advisor

Ahamed, Sheikh I.

Second Advisor

Harris, Douglas

Third Advisor

Madiraju, Praveen

Abstract

With the proliferation of mobile handheld devices and widespread deployment of wireless technologies, Location-based Services have become an integral part of our day to day life. The easy availability of these services enables people to locate special points of interests such as restaurants, shopping malls in map on their handheld devices. Yet, the increasing popularity is marred by the privacy threat arising from the usage. The threat stems from the disclosure of location information by the request issuer when he or she submits the Location based Query to the service provider. The adversary with ill intent may be able to correlate the information obtained from a query with the help of additional external information such as publicly available directories (WhitePages) and thus obtain a location trace of the individuals. Therefore, the goal of the research is to protect the privacy of such Location based Query in such a way that it thwarts any re-identification attempt by the adversary and presents the query result I in an efficient manner, considering the resource constraints of the handheld devices. In this thesis, we examine and address the privacy threats arising from the usage of the Location-based Services that use a Location Anonymizer (in order to cloak or obfuscate the location information of the user with several other users' locations). While the existing privacy protection mechanisms in the literature succumb to attacks by the adversary, our proposed approach happens to be attack resistant and at the same time demonstrates efficiency in terms of optimizing the candidate result set. The experimental evaluation provides comparative analysis among different privacy protection mechanisms and proves the claim that our approach is efficient and very appropriate for the mobile environment compared to other approaches.

Share

COinS