🚀 go-pugleaf

liuc@madrone.eecs.ucdavis.edu (Chia-Liang Liu) writes:

> >1.  The energy of a continuous time signal is just the integral of
> >the square of the signal.  For a digitized signal, is it true that
> >the energy contained is just the sum of the squares of each point?
> 
> Yes, if you assume that the sampling interval, Ts, is 1.0; otherwise
> you need to multiply it by Ts to get the energy.
> 

Isn't this only an approximation? For a countinuous signal the power is
evaluated as the Integral of the function over a given period divided by
the period. Isn't the above algorithm akin to evaulating the integration
with Riemann sums. Simpson's rule could be used to increase the accuracy.
I would think that the error gets significant as the the frequency of
the function approaches half the sampling frequency.
What would be an exact solution ?

Greg Basile
greg@gmp.lonestar.org