Comparing two signals with different sampling rates

Hello,

I am currently trying to compare two signals to evaluate how similar they are. I would like to put a quantitative criterion on this evaluation by using a comparison technique such as the mean square error for instance.

The problem is, the two signals cannot be compared directly via this method because they have different sampling rates. I cannot use functions such as upsample or downsample because the ratio between the two sampling rates is not an integer.

What I have come up with so far would be to use to use the vector of the signal with the highest sampling rate as my "base vector" and then interpolate a value for the other signal for each missing time step. I would then be able to compare both signals, but this technique doesn't seems appropriate considering it would introduce an additional error.

Any suggestions will be welcomed!

Thanks,

Answer 1

[Image Analyst]

Seems fine to me. It's done all the time. How would you recognize if the comparison is inaccurate or not accurate enough for you? In other words, try it and see. If it's no good (which I doubt) then you'll know it. If you don't notice any problem then, of course, it's good enough for your purposes.

Answer 2

[Haitham Hussien]

Dynamic Time Wraping