Prev: fft question
Next: Birthday Wisdom from my Mother
From: commengr on 31 Jul 2010 12:57

Hello, please help me learn the way to write equations for a block diagram,
that I have for a communication system.

The block diagram is shown in the figure

http://i29.tinypic.com/2yyppnr.gif


I believe the model is quite simple but I am willing to explain the model,
if it is required.


Actually, I do not know how to represent the process of "symbol repetition
onto parallel streams", or "serial to parallel conversion" or "parallel to
serial conversion" (used here for time combining) as equations. Spreading
or IDFT are simple enough.


Thanks.

From: commengr on 1 Aug 2010 06:04
Any suggestions or references?

>
>Hello, please help me learn the way to write equations for a block
diagram,
>that I have for a communication system.
>
>The block diagram is shown in the figure
>
> http://i29.tinypic.com/2yyppnr.gif
>
>
>I believe the model is quite simple but I am willing to explain the
model,
>if it is required.
>
>
>Actually, I do not know how to represent the process of "symbol
repetition
>onto parallel streams", or "serial to parallel conversion" or "parallel
to
>serial conversion" (used here for time combining) as equations. Spreading
>or IDFT are simple enough.
>
>
>Thanks.
>
>
From: Fred Marshall on 3 Aug 2010 14:55
commengr wrote:
> Any suggestions or references?
>

It looks like homework.

The diagram appears to be so full of underlying assumptions that I'd
hesitate to try it myself. I'm not sure what "n" implies here (except
the obvious connotation) in terms of the "solution space".
What is assumed or expected in that context? But then, I'm no expert in
this BPSK application space.

But, your question is an interesting one because I've *never* seen a
reference on this sort of thing. It's always been an implicit skill
developed by understanding a bit about calculus notation and some
physical thing. Then, one applies what they know about algebra and
calculus, to model or describe the physical thing that they understand,
into mathematical expressions. It's like the "word problems" we did in
math classes.

Dealing with discrete sequences is almost the same thing except the
integrals become sums, etc. Then you pay close attention to things like
indices - how are they to be represented, etc.

I've found that large pieces of paper with sequences shown
sample-for-sample are often necessary to derive more compact notation.
That way, time shifts, etc. can be more easily visualized and taken into
account.
Did that on a frequency-domain beamformer and a reverberation generator
called REVGEN.
The more you practice, the more you might get straight to the compact
notation - which I think are the expressions you're looking to develop.

After a little practice, it's more about careful hard work and less
about skill.

Fred
 | 
Pages: 1
Prev: fft question
Next: Birthday Wisdom from my Mother