Sampling rate selection for real-time de-noising systems

With the continuously increasing capabilities of the data acquisition hardware, signals can be sampled at a much higher rate than necessary. In a real-time control application, oversampling noisy signals makes them more difficult to de-noise, particularly when the de-noising must be done in real-time, on embedded hardware, as is the case when de-noising signals in feedback control loops. This work investigates the effect of sampling rate on signal de-noising and shows that appropriately decimating an oversampled noisy signal before it is de-noised can improve the performance of the de-noising process. The work provides a methodology to determine whether or not a signal is oversampled; if it is, the methodology is used to determine the appropriate decimation factor. Two example signals are used. The improvement of the de-noising performance shows the effectiveness of the decimation technique.