| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
|
Vafa,
Our interpretation of Table 311b is that the DC subcarrier is included in the equation for DL-FUSC. Thus, you would not add 1 to the indices greater than FFTSize/2. As a result, the index for your case below falls at index 469.
However, for the multiple antenna case (NumTxAnt = 4), this becomes an issue for Antennas 2 and 3 on every other odd and even symbols because the indices are defined by VariableSet#1 + 1 (Sec 8.4.8.2.2 in P802.16-2004/Cor1/D5).
Does anyone else have a clarification to the interpretation of Table 311b above or the multiple antenna case ?
Best Regards,
Dave Magee Texas Instruments
From: Vafa Ghazi
[mailto:vafa@ieee.org]
Hi,
I have a question regarding the 512-FFT OFDMA DL FUSC mode in the 802.16e amendment. It seems that for odd valued FUSC Symbol Numbers, the last pilot from the Variable Set #1 will overlap the first guard subcarrier on the right side of the spectrum.
Based on Table 311b on page 427 of IEEE P802.16e/D12, the last pilot from the Variable Set #1 is at location index 420. Accounting for the 43 left guard subcarriers and the DC subcarrier, the actual location of this last pilot subcarrier is index 464 (in the 0:511 range of 512 total subcarriers). Based on Equation 110 on page 565 of IEEE 802.16-2004, the location of Variable Set pilot subcarriers is shifted an additional 6 positions to the right for odd FUSC Symbol Numbers:
PilotsLocation = VariableSet#x + 6*(FUSC_SymbolNumber mod 2)
Therefore the location of last pilot of Variable Set #1 for odd FUSC Symbol Number counts is 470. But based on Table 311b, there are 42 right guard subcarriers, which are allocated on subcarriers 470:511. Therefore the last Variable Set #1 pilot overlaps the first guard subcarriers in the right subcarrier.
Is this is an error or am I missing something here? Any input will be highly appreciated.
Regards, Vafa Ghazi CoWare, Inc.
|