A low order model of sleep stage II EEG based on the gamma kernel

In this paper we propose and analyze a model based on the gamma kernel to simulate sleep stage II EEG signals. The only input required to build the model is a segment of sleep stage II signal. The model is then excited by Gaussian white noise to produce the simulated signal. Some comparisons between a real signal and a simulated signal are presented that show that a good matching can be achieved with a low order model.