Need Complex samples

Hi,

    I am digitizing an analog signal. Now I have real samples of the
analog signal. I need to calculate the power of the signal with these
samples. 
    Is this possible to calculate the power of the signal? 
    Does it need complex samples to calculate? 
    If complex samples are needed to calcuate the signal power How can I
get complex samples from the real samples?

On Aug 21, 7:14&#4294967295;am, "gomatam" <ranganad...@rediffmail.com> wrote:
> Hi,
>
> &#4294967295; &#4294967295; I am digitizing an analog signal. Now I have real samples of the
> analog signal. I need to calculate the power of the signal with these
> samples.
> &#4294967295; &#4294967295; Is this possible to calculate the power of the signal?
> &#4294967295; &#4294967295; Does it need complex samples to calculate?
> &#4294967295; &#4294967295; If complex samples are needed to calcuate the signal power How can I
> get complex samples from the real samples?

How about simple squaring and averaging, and then
normalizing to your sampling rate?

On 21 Aug, 14:14, "gomatam" <ranganad...@rediffmail.com> wrote:
> Hi,
>
> &#4294967295; &#4294967295; I am digitizing an analog signal. Now I have real samples of the
> analog signal. I need to calculate the power of the signal with these
> samples.
> &#4294967295; &#4294967295; Is this possible to calculate the power of the signal?

Sure. Square each sample, sum them, and divide by sampling period.

> &#4294967295; &#4294967295; Does it need complex samples to calculate?

No. One often writes recipes where the samples are complex-valued
because one wants to save time and effort. Most of the time
one needs the real-valued variation, and one sets the imaginary
part of the computations to 0.

However, in the few cases one needs the complex-variations one
already has everything ready and needs not do the derivations
again.

Rune

On Aug 21, 7:30&#4294967295;am, julius <juli...@gmail.com> wrote:
> On Aug 21, 7:14&#4294967295;am, "gomatam" <ranganad...@rediffmail.com> wrote:
>
> > Hi,
>
> > &#4294967295; &#4294967295; I am digitizing an analog signal. Now I have real samples of the
> > analog signal. I need to calculate the power of the signal with these
> > samples.
> > &#4294967295; &#4294967295; Is this possible to calculate the power of the signal?
> > &#4294967295; &#4294967295; Does it need complex samples to calculate?
> > &#4294967295; &#4294967295; If complex samples are needed to calcuate the signal power How can I
> > get complex samples from the real samples?
>
> How about simple squaring and averaging, and then
> normalizing to your sampling rate?

Gotam, please keep your questions on this forum.