Kalman-Takens filtering in communication systems.

We apply a model-free predictor based on the Kalman Filter to signal-to-noise ratio (SNR) data from a mobile communication system experiment. The experiment consist of collecting performance indicators on a mobile device during the trajectory of a city bus. In particular, we analyze the SNR measured by the mobile, which is collected every second via an application. Since some mechanisms in a mobile network depend on the SNR, like power control and handoff processes, our results show that it is possible to use prediction models to improve several procedures in mobile communications systems. [ABSTRACT FROM AUTHOR]