[STDS-802-16] Fwd: Question regarding 802.16-2004 standard
This note is from ERNEST KURNIAWAN
<ernest_kr@pmail.ntu.edu.sg>. Feel free to
respond.
Roger
>Dear Marks,
>
>I have a question regarding 802.16-2004
>standards. I tried to build the system
>simulation based on the specification on the
>802.16-2004 standards. And I tried putting in
>the test Input in the example given in the
>standard (pp. 618-700 section 8.4.9.4.4). But
>somehow, I get different output on one of its
>building block, so I need to clarify whether my
>block design is wrong or if there are some steps
>I missed out. The following are the detail of my
>block processing:
>· For the Randomizer block, I managed to
>process the "Input Data" to "Randomized Data" by
>using PRBS constructed as in Figure 253, and
>initializing the randomizer with the OFDMA
>Symbol Offset (=35 in our example) for the first
>10 digits (b14..b5) and Sub-channel offset (=6
>in our example) for the last 5 digits (b4..b0).
>· For the FEC block, I managed to
>process the "Randomized Data" to "Convolutional
>encoded Data" by using convolutional encoder
>with rate=½ as depicted in Figure 255, and
>applying tail biting algorithm for the
>initialization value.
>· Now, my problem starts on the
>Interleaver block. I used the equation (130) and
>(131) on page 615 to compute the interleaved bit
>index with the following values as the constants:
>o Ncbps = 192 (which is the number of
>bits contained in the message to be interleaved)
>o d = 16
>o Ncpc = 2 (which is the number of bits per symbol in QPSK modulation)
>o s = 1 (which is half the value of Ncpc)
>But the interleaved value I obtain is "6D B7 EF
>FD B8 68 38 C9 27 9E D4 A3 AE FC 15 DE FE B9 68
>AD 68 56 19 9A". Interestingly, when I set Ncpc
>to 4 (by assuming that we use 16-QAM instead), I
>managed to get the value as in the example. >>
>So is there anything I missed out or any
>parameter I wrongly set here?
>· From this point, I assumed that the
>information I get as the interleaved data is the
>same as what the example give, and keep
>maintaining the QPSK modulation configuration.
>· For the Constellation Mapper block, I
>managed to get the output right by applying the
>constellation depicted in Figure 263 for QPSK,
>and taking the Interleaved Data given in the
>example as its input.
>· Now, performing sub-channel allocation
>is the biggest confusion that I have here. In
>the example, it is stated that the logical slot
>6 is mapped onto physical sub-channel 16 in the
>first time slot and physical sub-channel 17 in
>the second time slot. So, by applying equation
>(113) on page 569, with parameter s set to 16
>and 17 (sub-channel number), and parameter
>UL_IDcell set to 5 (IDcell number), I managed to
>obtain the tiles corresponding to the two
>physical sub-channels to be used. For the first
>sub-channel, the six tiles spans across OFDMA
>symbol index 35-37, and sub-carrier range of
>448-451; 512-515; 984-987; 1189-1192; 1505-1508;
>1753-1753 correspondingly. The second
>sub-channel, the six tiles spans across OFDMA
>symbol index 38-40, and sub-carrier range of
>232-235; 704-707; 908-911; 1225-1228; 1473-1476;
>1813-1816 correspondingly. Taking the values
>depicted in the example given, I rewrite it on
>the table as can be seen on the attached file.
>There are two doubts that I have in this
>scenario:
> >> On the second sub-channel, it is noted that
>the pilot location does not conform to the
>structure depicted in figure 236.
> >> Applying the reordering sequence as
>explained in page 570 and equation (114), I
>could not manage to obtain the same mapping as
>what is given in the example. Here, I used
>Nsubcarriers = 48 (as there are 48 data
>sub-carriers in each sub-channel), and parameter
>s set to 16 and 17 (for the sub-channel 16 and
>17 correspondingly). I would appreciate if you
>could elaborate more on the reordering process
>and equation (114) in page 570.
>
>Please let me know if you need any further
>details on how I obtained the above value. Your
>help is greatly appreciated. Thank you very much
>in advance.
>
>Regards,
>Ernest.