The highly non-linear and coupled dynamics of Autonomous Underwater Vehicles (AUVs), added with modeling errors, parametric uncertainties and payload variations pose a major challenge towards autonomous control of AUVs for various application requirements. Environmental hazards such as ocean currents sometimes dominate and make the control of underwater systems even more complicated. The proposed control technique addresses the design of a robust controller for reasonably accurate path tracking of AUVs incorporating the effects of above uncertain paradigms within some known bounds. It is well-known that measurement noise, which is associated with the navigational sensors, degrades the performance of the controller leading to substantial deviation from the reference path. Incorporation of sensor fusion technique, which is driven by sensors error characteristics, is necessary to improve the controller performance. Performance of the controller is verified using the real-life parameters an AUV, developed at CSIR-CMERI, Durgapur, India considering a few uncertainties.