Although many techniques have been proposed to deal with location privacy problem, which is one of popular research issues in location based services, some limitations still remain and hence they cannot be applied to the real world. One of the most typical proposed techniques is obfuscation that preserves location privacy by degrading the quality of user’s location information. But the less exact information, the more secure and the less effective services can be supported. Thus the goal of obfuscated techniques is balancing between privacy and quality of services. However, most of obfuscated techniques are separated from database level, leading to other security and performance issues. In this paper, we introduce a new approach to protect location privacy at database level, called Semantic Bob-tree, an index structure that is based on Bdual-tree and Bob-tree and contains semantic aware information in its nodes. It can achieve high level of privacy and keep up the quality of services.