Bin-Chul,
Guard
bands are defined to be subcarriers with zero modulation. If you modulate
something
on
subcarriers belonging in guard bands, they are not guard bands anymore.
You will have
to go
back to D4 and redefine guard bands.
If
you come back with analyses of comprehensive spectrum mask studies and
decide that
there are
too many guard bands currently defined in D4 and the spectrum mask
requirement
meets
emission standards all over the world, we should go back to D4 and redo the
guard band
requirements. These extra subcarriers can be better used for
data than for a reference signal.
Better
yet, you can also choose to use the DC subcarrier for reference signal if you
don't
mind
voiding DCR devices.
Jason
Dear all,
This e-mail is from LGE.
Our contribution proposes the reference
signal supporting preamble search.
The reference signal is inserted in the
available unused subcarriers in FUSC and PUSC while the reference
signal is not included in O-FUSC and AMC.
That is, # of subcarriers
in PUSC or FUSC + # of reference signals subcarriers = # of subcarriers in
O-FUSC or AMC.
The reference signal does not break the
spectral mask.
This case is depicted in Fig.3 of
our contribution "Reference Signals for Fast Preamble Search in
OFDMA".
As a result, the
reference signal is NOT included in guardband
prohibited to carry signal.
For your understanding, refer to our
contribution.
The reference signal contains the information
of remaining symbols to the next preamble.
SS which obtains the information from the
reference signal can stop the exhaustive preamble searching process about the
duration of remaining symbols. This makes SS save
power considerably.
Please, Do not misunderstand the usage of
guardband.
The reference signal does
not use guardband.
Best Regards
Bin-Chul
Ihm
|