Hello Everyone,

Thank you for the replies. I'm still a little

confused though. May be my fundamental understanding

of OFDM is completely wrong.

Here is my knowledge thus far:

1. A serial stream of bits is broken down into "N"

parallel streams. "N" being the number of

sub-carriers.

2. The binary data in each of these streams is mapped

using any digital modulation scheme.

3. So at this point, we have complex numbers on each

of the parallel streams. The data in all these streams

will have to be modulated with "N" orthogonal

carriers, which I thought was achieved by the IFFT.

4. So, at this point, isn't the output of IFFT a

composite signal, representing the sum of all

carriers.

I have attached my test model. I have tried to

generate an OFDM signal with 4 sub-carriers. Could you

please review this and let me know if I am doing

anything wrong?

Thank you so much.

Regards,

Rohith

--- Amit Pathania wrote:

> Rohit,

>

> When you perform an IFFT you are essentially

> converting your overlapping subcarriers(if they

> indeed are overlapping to begin with) into a time

> domain signal.

> Rohith Ramnath wrote:

> Hello Everyone,

>

> I started working recently on OFDM. I am trying to

> use simulink to

> build a simple system with four sub-carriers.

> However, I am unable to

> comprehend the results I am getting. Here is what I

> am doing:

>

> 1. I am using a Bernoulli Binary data Generator

> 2. It's output is QPSK modulated.

> 3. After separating the real and complex parts I

> concatenate the

> matrix and zero pad it to increase the length to 16.

> 4. I then do an IFFT on this.

> 5. If the plot the result using

> plot(fftshift(abs(result))), I do not

> see any overlapping sinusoids.

>

> So my question is:

>

> a. Is this approach correct?

> b. What am I doing wrong? Could you please explain?

> c. Does anyone have a basic simulink model for a

> simple 4

> sub-carrier (no guard interval) transmitter?

>

> Thank you,

> Rohith

>

>

> Amit Pathania