Hello everybody
I am doing an application for Low Frequency Noise Measurement. I am getting the Power Spectrum (PS) of N samples acquired from signal noise every T time at a fixed sampling frequency Fs.
I am confused in the next data processing: Assuming that during long-time of data acquisition, I have obtained 100 PS related to a continuous noise signal observation during 100*T time. Is it correct trying to obtain only one PS that represents all the 100 PS? In the case that it is correct: How can I do it?
I have made some experiments in order to answer this questions but I am not sure about the results obtained or at least I can not understand well it is meaning?
Thanks in advance by the answers.
Thank you very much Mr Lerch. I will do some little experiments and I will let you know. Thanks.
Good weekend also to you!!
Hi,
I may add some comments. First of all Philippe is right. You problem is based on a system model which should be stable. The theoretical concept is g rounded in the ergodic hypothesis which is well defined in Physics and in your case in Signal and System Theory or statistics. If you take a look in this topic you will see what is to do to define whether you model is ergodic or not and what you can/should do regarding Ergodicity.
- Badges
- Science topic