Real-time location-based tracking applications require of continuous GPS calculations and transmissions, which consume a considerable amount of the phone's battery. As a result, methods have been devised to reduce the amount of GPS calculations and transmissions without sacrificing the tracking capabilities of the applications. One of these methods is based on a state machine that dynamically changes the frequency of GPS updates according to the user direction, speed, received signal strength, and other factors. However, the state machine, although efficient in terms of energy savings, still presents one major problem: it does not take into account the presence of noise in GPS data. In order to distinguish between actual GPS data and noise, three versions of the Kalman filter have been implemented within the state machine. These modified Kalman filters remove noisy GPS fixes with little to no input from the user in a very efficient manner. The filters are discussed in detail and tested against one another to determine which one removes GPS noise better and which one reduces the energy consumption in the cellular phone more with no loss of valuable tracking data. Experiments conducted show the Adaptive Kalman Filter as the best performer. No loss of valuable tracking data is seen while it introduces a significant decrease in the number of “asleep” fixes. The Adaptive Robust Kalman Filter is the second best performer of the three filters. It shows no loss of tracking data, while a slightly less decrease in “sleep” fixes. Testing shows that the Robust Kalman Filter is the worst performer of the three. This is because the Robust Kalman Filter is the slowest version to “wake up” and make transitions to a “sleep” state.