Research On Privacy Protected Reverse Nearest Neighbor Query Processing For Spatio-Temporal Data

In recently years, along with the development of wireless communication technology and intelligent device popularization, Location Based Service (LBS) plays increasingly important position in people’s lives. As the basic of LBS, Spatio-Temporal Database (STDB) plays a very important role in information retrieval and recommendation system applications. STDB includes both temporal and spatial elements. It is mainly used for storaging and managing all kinds of space objects whose positions or shapes change over time. STDB involves the spatio-temporal objects expression, spatio-temporal data modeling, indexing and query, in which the most important two are indexing and query. The main purpose of spatio-temporal data indexing is establishing index system of time and space data so as to access them efficiently and accurately. Spatial data query has more applications in today’s services, which refers to getting the results with satisfied efficiency and accuracy. The common queries in STDB are Nearest Neighbor(NN) Query, Range Query and Reverse Nearest Neighbor(RNN) Query. The RNN query was purposed in1999at the first time, and developed rapidly in the last decade. Some works expand the area of RNN research to uncertain dataset, moving objects, continue query and queried object visible or not, such scenario and so on. RNN query has widely application in recommendation system and decision system. Therefore it becomes the typical inquires in LBS.LBS bring more convenience for people’s life. While in the same time, attackers can get people’s location information by attacking the LBS server.. Thus, privacy protection is an important issue when LBSs are used.The traditional spatio-temporal data operation is usually based on the Euclidean Space. However, with the popularity of LBS, the importance of Road Network was promoted rapidly in STDB. In Euclidean space, theoretical arguments are intuitional, while road network is more close to the real life. People always follow necessary path on the ground when they from one place to another, not in Euclidean space straight line. Therefore, both Euclidean space and road network are significant distance matrix.In this thesis, the authors focus on LBS platform to realize privacy-preserving reverse nearest neighbor query processing in spatio-temporal database. First of all, the paper considers both Euclidean space and road network distance matrixes. The two kinds of model have separately strengths and can collaborate with each other to work advantage. Secondly, the paper has considered user location privacy protection. In Euclidean space and road network models, discrete anonymous, continuous anonymous and X-Star are used to protect user’s location privacy against the attacking from adversaries. Finally, the authors conducted the experimental evaluation based on the above algorithms. The experiments proved efficiency and accuracy of the algorithm proposed in this thesis. These contributions will help the further popularization of LBS and provide more satisfactory and security service for user.