Abstract: Because the existing target association and positioning method separates association and positioning into two independent processes，the target positioning accuracy is seriously limited by the correct association in this kind of step-by-step processing when taking the local optimal association and positioning method.Therefore，a joint multi-station target association and positioning method based on the divide-and-conquer and greedy thoughts is proposed. The sets of real intersection points on each bearing line are selected according to the divide-and-conquer thought and the minimal distance principle，and all the sets of intersection points are merged according to the greedy thought. In the merging process of sets，the target association and positioning accuracies are guaranteed through repeated verification between the positioning and association processes.The mutually exclusive target measurement sets are combined according to the association relationship，and the multi-target measurement set with the maximum joint association probability is selected as the final output.The simulated results show that the association accuracy of the method remains above 90% when there is clutter，and still above 70% when there is both false alarm and missing detection. The proposed method is to use the greedy thought to reduce the calculation complexity and overhead.The simulated and experimental results both verify that the proposed method has high target association and positioning performance，and is suitable for underwater environments with low detection limit and high false alarms.

Key words: bearingcross, targetassociationandpositioning, divide-and-conquerandgreedythoughts, underwatersensornetwork