In a location query process the most expensive procedure is maintaining the location information of mobiles. To reduce the expensive cost, we adopt the profile-based approach to design a behavior-based strategy (BBS) based on the moving behavior of each mobile generated by long-term collection of its moving history. We use a data mining technique to mine the moving behavior of each mobile and then estimate the probability that each mobile stays in each location at each time region given the last known location from its moving behavior. To reduce unnecessary computation, we consider the location tracking and computational cost and then derive a cost model. A greedy heuristic is proposed to minimize the cost model through finding the appropriate checkpoints. The experimental results show our strategy outperforms fixed paging area strategy currently used in GSM or IS-54 system and time-based strategy for highly regular moving mobiles.