OFDM and Wimax data mapping

[gpi]

Hi all,

I am trying to simulate a data flow according to the wimax standard.
Could someone please help to understand how to manage the data
flow dimensions after encoding before mapping to ofdm symbols?
More precisely:
I generate, for example, a message of 72 words having in mind a 16 QAM modulation, with an overall coding rate 3/4. This, after rs and cc coding, interleaving and puncturing, leads to 192 symbols to be transmitted.
I note that this number is exactly the number of used data subcarriers for the 256 point ifft.
The questions are:
1. can I use, for example, the 1024 points ifft?
2. considering the predetermined allocation of null carriers and pilot carriers, how do i fill the (720-192)=528 frequencies left?

Maybe I am missing something

Many thanks

Gian Paolo

[charan langton]

Upsample your data by 4 to do a 1024 size FFT.
Charan

[gpi]

Aug 6, 2008 3:07:24 GMT -5 charan langton said:

Upsample your data by 4 to do a 1024 size FFT.
Charan

Hi Charan,

thanks for the reply.
Anyway, I see that factor 4 seems not enough: in fact, I would get 192*4 =768 symbols to be mapped to data-carriers, but for 1024 ifft the Wimax standard prescribes 720 data carriers only. Thus RS/CC coding and puncturing, at least in the form as in IEEE Std 802.16-2004 document don't apply straightforward.

I am wondering if the RS code in that document, precisely Table 215 in Section 8.3.3.2.1, applies to the case of 256 points ifft only.

what do you think?

many thanks

Gian Paolo

[gpi]

Hi,

maybe the night brought the correct answer:
I was considering PUSC prescriptions instead of FUSC.
In this case one resembles the correct number of
data carriers and multipying by 4 is fine.

Many thanks for the discussion!

Gian Paolo