Abstarct: The proper control for tracking the desired path is one of the most important military superiority and maneuverability of a fighter aircraft. The fighter aircraft control system is usually automatic. Automatic control systems for aircraft guidance realize via the various control parameters such as angle of attack, pitch angle, flight angle and altitude. Therefore, it is important to use an intelligent and online method for an efficient track of complex and nonlinear maneuver. In this thesis, kernel least mean square (KLMS) baxsed method is proposed for the intelligent control and online tracking of a desirable nonlinear maneuver. The kernel least mean square algorithm enhances the present understanding of the LMS algorithm with a machine learning perspective. Moreover, training the RBF networks with the KLMS is different from conventional RBF networks. For conventional RBF training, the kernel centers and their number have to be chosen heuristically or through a complex algorithm. While in the KLMS method, the centers and its number are automatically chosen during the learning, and there do not need to be determined beforehand. Therefore, this method performs as a generalized simple neural network with radial type basis functions. Furthermore, for updating the kernel function, an adaptive learning algorithm is also presented. Finally, the efficiency of the proposed approach is successfully investigated through the nonlinear maneuver tracking of a fighter aircraft.