Navigation of autonomous mobile robots in unpredictable and dynamic environments is restricted by uncertainty, complexity, and unreliability issues of robots and their environments. In this context a navigation system for an autonomous mobile robot using intelligent fuzzy logic technique will be presented. Fuzzy logic control is able to emulate human experience about how best to control a system without needing accurate model equations, and can handle any perturbation in the system. A hierarchical behavior based control strategy has been devised in which four different reactive behaviors are combined by means of a fuzzy supervisor. For testing and validation, many simulations have been proposed which focus on: moving towards static or movable goal; escaping from local minima whenever it's detected; and traversing a cluttered environment with unknown static and dynamic obstacles.