This paper presents the results of a control and guidance strategy for homing and docking tasks using an autonomous underwater vehicle. An optimal high-order sliding mode control via state-dependent Riccati equation approach is introduced providing a robustness of motion control including elimination of chattering effect for decoupled systems of an AUV. Motion planning for a docking is introduced. The average vector field based on an artificial potential field method gives a desired trajectory using existing information from ocean network sensors. It provides a guidance for an AUV to follow the path to a required position with final desired orientation. A Line-of-Sight method is used for an AUV to follow the predefined path. In order to improve a docking manoeuver, a switched weight technique is proposed for controlling a vehicle's path and final stage docking.