The accuracy of navigation, particularly in metropolitan areas,depends on aplication of appropriate kinematic models and filters.It is needed to filter the perturbing effects of phenomena like maneuvers, sudden accelarations and irregular stops of vehicles. Inthis respect, the success of Kalman Filtering is closely related toforming noise models which play a key role in representing the pertubations. With appropriate kinematic and noise modeling, KalmanFilter can be implemented optimally and navigation solutions canbe improved to a greater extent. In this study, a simulation of a real time Kalman FilteringApplication with NMEA-formatted data, the most common formatin navigation and tracking applications, is carried out, and constantvelocity, constant acceleration, time-correlated velocity and timecorrelated acceleration models are investigated in view of boththeory and practice