Signal Processing Stack Exchange is a question and answer site for practitioners of the art and science of signal, image and video processing. It only takes a minute to sign up.
Sign up to join this community
Suppose you have the ability to inject any arbitrary waveform into a piece of analog rf hardware and collect and digitize the output for analysis.
If you wanted to characterize/estimate the transfer function of the hardware, what waveforms would you use?
You can use adaptive filters for this task. Feed the hardware with uniform white noise and collect the output. Then feed the adaptive filter with exactly the same noise sequence and use the output of the hardware you recorded as the desired response of the adaptive filter. The filter can then be trained with standard algorithm such as LMS. Of course, you will not get the exact transfer function, but the trained adaptive filter's transfer function should be close enough to serve as model of the hardware in question.
My first guess would be: Use a swept sinus signal and signal analyzer.
For the sake of completeness: you could also feed a Dirac impulse into your hardware and sample the impulse response at its output. Then apply a DFT to the measured impulse response in order to get the transfer function.
Ideal Dirac impulses cannot be generated, though. So in practice a very narrow impulse will have to do. I would therefore prefer white noise in practice.
