LBS Privacy Protection Based On Centroid Shift Clustering Algorithm

In recent years, with the rapid development of mobile internet, LBS (Location Based Services)is becoming increasingly important in people’s daily life. LBS provide high-quality services at thesame time, there are still security risks. In order to make sustained and healthy development, LBSprivacy protection is becoming a hot topic in the field of information security reserching.Firstly, this thesis introduces the LBS reserch history, current research status and faced problemin privacy protection field. Then it analyzes three existing mainstream protection models: fakedposition, spatial anonymous, anonymous space-time protection model. This thesis chooses theanonymous space of K-anonymity model as a research direction, analysis the strengths andweaknesses of the DHLib anonymity algorithm implemented on the existing K-anonymity model. itproposed the CSACA algorithm based on the centroid shift clustering anonymous algorithm, thecore idea of the CSACA algorithm is to use the grid partition theory, divided the area whichcoveraged by the anonymous server into grids. Then, calculate centroid of the anonymous box fromthe target user’s grid; join the grid in anonymous box which nearest to centroid, until the mass of theanonymous box is not more than K. Synthesis anonymous request information in accordance withthe anonymous box and original request. Another innovation in this thesis is introducing the conceptof security levels, adaptive matched K values according to the request content; protect the requestsin different grading according to different security levels.Finally, it verified the feasibility of the CSACA algorithm by experiments, proved thesuperiority of the new algorithm By comparison with the DHLib algorithm.