ok first can some one please please please tell me how to do fourier analysis on a simple square random waveform.


i take it that for a random waveform you only consider pulses,, and its got something to do with a probabilty weighting of 0 and 1 signal...


ok

(wfm = waveform)

now for a special digital pulse....

originally i got

random square wfm: 0 1 0 0 1 1 0


now i divided frequency by 3.........
and i get a new signal the rule for which

for 0 i get 010
for 1 i get 100
for no change (from previous state) 000

hence using the original
random square wfm: 0 1 0 0 1 1 0

using the rule to get new sig.
i get new wfm: 010 100 010 000 100 000 010


so in other words,, i can only have 010 or 100 or 000 but never can i have 110 or 111 or 101 or 011

so how do i do the fourier analysis on this piece of rubbish (and probably utterly useless wfm)?

please please please
pretty please with sugar on top

(can some tell me already ? :o :o :confused:

Numerical Recipes is the best place to go.

http://www.library.cornell.edu/nr/bookcpdf.html

You'll be especially interested in Chapter 12.

- Warren

hey thanks,, i will check it out tonight (espacially cos i can't find the ref of Phd, my teacher gave me, cos stupid library,,,) but i am a bit confused,, i didn't think FFT deals with random pulses does it?

Originally posted by news
hey thanks,, i will check it out tonight (espacially cos i can't find the ref of Phd, my teacher gave me, cos stupid library,,,) but i am a bit confused,, i didn't think FFT deals with random pulses does it?
Neither the term 'random pulse,' nor the stuff you typed in your first post, have anything to do with math... so I'm not sure exactly what you mean.

The FFT can be used to compute the Fourier transform on discretely sampled data.. which means basically any piece of data representing a waveform of some kind can be passed through it.

- Warren

hey thanks for that,,,,,,

you just reminded me,,,,

what i was talking about is used as somewhat,, in the theory of transmission link regenration filter design


now i didn't think that was useful earlier cos my signal has to do with optical transmission,, but thanks,,,, i think the discussion of that random pulse signal is useful,, so i have to look at that, and then come to you..

the reason why i need the new signal from the original is the question i asked my teacher when he gave my assignment and he said,, answer lies in fourier,,

anyway, i get finding my comm lecture notes again

i post that up tonight and see what you think....

thanks,,,

FFT is for actual data,,,

but what i am talking about is theoretical discussion wise analysis,,,,,

but still i got some graphs about that filter i was talking about a minute that are suppose to show something

ok ok enuff talk got to study,,

don't worry i post it, cos i need help with understanding that probably as well anyway,,

The NR book also has a decent introduction to analytical Fourier transforms as well, which is what I think you mean by "theoretical discussion wise analysis."

- Warren

yep your right,,,

i need to be talking about fourier transform


exactly what is done on page 1 of the attached pdf file,, but thats only for a normal random pulse where the signal is either 1 or 0 , which means probability of a 1/2 for either 1 or 0 (hence the inside sinc bracket the fucntion wtp/ 2 )

but for my new signal for this example,, it is still random

and the only signal i can get is 100, or 010 or 000.

in other words probability of last bit being 1 is 0 because its always 0, probabiltiy of first bit being 1 is 1/3,, and probability of second bit being 1 is also 1/3

i think thats what my teacher told me as an introduction to it,, then he said something about weighing probabilities and mentioned fourier analysis,,,,

so looking at notes attached,,, it is briefly covered in page 1 only,, but
yep your right i need to be talking about fourier transform,,

thanks man,, your a great help so far,, cos as you can guess i got no expertise here whatsoever...

Well, I've helped you all I can.

I don't know what a "normal random pulse" is, and you seem to be unable to express anything in proper language. You don't really make much sense to me.

I have given you a link on how to carry out an analytical Fourier transform, and how to perform a numerical Fast Fourier Transform. That doesn't seem to be enough for you, but I really can't understand what else you want.

- Warren