Abstract: With the gradual deepening of human understanding of the ocean, location information of underwater targets is an indispensable source of underwater information, and traditional underwater localization methods can no longer meet the application requirements of different scenarios. There are limitations in the fusion localization of diversified multiple information or in the joint localization of insufficient time-domain information. In this regard, a single beacon localization algorithm based on a probability graph model is proposed. In the cases where beacons are sparse or beacon communication information is missing, the localization can be achieved by combining all the measured information at other location moments. In addition, the impacts of model resolution and ranging error on localization results are analyzed, and the effectiveness of the probability graph method is verified through lake trials. It is proved that the proposed method can estimate the position of the target at each moment and describe the trajectory of the target in the single beacon localization and tracking process, which is impossible to do so for traditional localization methods. And the average location accuracy reaches 1.203 5 m.