In recent years, most location-based service privacy protection turns to peer-to-peer based on subscriber cooperation. However, the biggest drawback of that mode lies in the integrity of collaborative users. If there are vicious users in ambient collaborative users, the information of collaborative users may be accessed illegally. So this paper proposes a method of preventing the location privacy from being revealed to vicious users via using third-party trust institutions to detect the abnormal network behaviors for mobile users. In the meanwhile, we broaden the zone of collaborative users in the mode of peer to peer through binary tree. The experimental result shows the method can effectively prevent unfaithful users from joining in the anonymous group and reduce the probability of the information leakage.