Location-based services are popular server modes. However, location privacy problem in the Location-based services becomes a potential concern for service providers and users. In addition, existing obfuscation schemes downplay the effect of the privacy area and the practical road distribution in the areas. They mainly consider how to generate k dummies using either specific locations or randomly generated fake locations in circle areas or grid areas. However, circle areas or grid areas are not practical in a real world. We propose a context-aware dummy generation algorithm in convex areas. The convex areas are obtained according to the Voronoi diagram of Point of Interests (POIs). The algorithm controls the distribution of the dummy positions considering background information of each position to resist inference attack. The proposed algorithm obtains accurate query results without sending real query location and exclude much redundancy query results by filtering the same POIs. Empirical simulation shows our algorithm can improve location privacy at a low cost of communication overheads.
2016 IEEE First International Conference on Data Science in Cyberspace (DSC)
Voronoi based dummy generation algorithm,privacy aware location,service providers,obfuscation schemes,road distribution,generated fake locations,circle areas,grid areas,context-aware dummy generation algorithm,convex areas,Voronoi diagram,Point of Interests,POI,inference attack
Data mining,Computer science,Server,Algorithm,Location-based service,Service provider,Redundancy (engineering),Voronoi diagram,Inference attack,Obfuscation,Grid