This work deals with the problem of simultaneous localization and mapping (SLAM). Classical methods for solving the SLAM problem are based on the Extended Kalman Filter (EKF-SLAM) or particle filter (FastSLAM). These kinds of algorithms allow on-line solving but could be inconsistent. In this report, the above-mentioned algorithms are not studied but global ones. Global approaches need all measurements from the initial step to the final step in order to compute the trajectory of the robot and the location of the landmarks. Even if global approaches do not allow on-line solving, they can be more interesting than EKF-SLAM or FastSLAM since they are less sensitive to inconsistencies. Two algorithms are studied: the GraphSLAM, a probabilistic method based on gaussian hypothesis, and the "interval SLAM" which is a deterministic approach. A comparison of the algorithms is made in simulation, for the bearing-only case. Landmarks are 3D points from which we measure the bearing and elevation angles. The results show the consistency of both algorithms when the errors are centered. In this case, if we look the size of the belief areas provided by the algorithms, GraphSLAM delivers better results than interval SLAM. Finally, the GraphSLAM algorithm becomes inconsistent when input data are biased. In the latter case, interval SLAM gives good and consistent results.