Terrestrial Transmission of Digital Television Signals (DVB-T) - Digital Video and Audio Broadcasting Technology - page 336

for a 2046-points IFFT and 8K stands for an 8192-points IFFT. As is al-
ready known from the chapter on COFDM, the number of COFDM sub-
carriers must be a power of two. In DVB-T, it was decided to use symbols
with a length of about 250 Μs (2K mode) or 1 ms (8K mode). Depending
on requirements, one or the other mode can be selected. The 2K mode has
greater subcarrier spacing of about 4 kHz but the symbol period is much
shorter. Compared with the 8K mode with a subcarrier spacing of about 1
kHz, it is much less susceptible to spreading in the frequency domain
caused by doppler effects due to mobile reception and multiple echoes but
much more susceptible to greater echo delays. In single frequency net-
works, for example, the 8K mode will always be selected because of the
greater transmitter spacing possible. In mobile reception, the 2K mode is
better because of the greater subcarrier spacing. The DVB-T standard al-
lows for flexible control of the transmission parameters.
Apart from the symbol length, which is a result of the use of 2K or 8K
mode, the guard interval can also be adjusted within a range of 1/4 to 1/32
of the symbol length. It is possible to select the type of modulation (QPSK,
16QAM or 64QAM)). The error protection (FEC) is designed to be the
same as in the DVB-S satellite standard. The DVB-T transmission can be
adapted to the respective requirements with regard to robustness or net
data rates by adjusting the code rate (1/2 ... 7/8).
In addition, the DVB-T standard provides for hierarchical coding as an
option. In hierarchical coding, the modulator has two transport stream in-
puts and two independently configurable but identical FECs. The idea is to
apply a large amount of error correction to a transport stream with a low
data rate and then to transmit it with a very robust type of modulation.
This transport stream path is then called the high priority (HP) path. The
second transport stream has a higher data rate and is transmitted with less
error correction and, e.g. 64QAM modulation and the path is called the
low priority (LP) path. It would then be possible, e.g. to subject the identi-
cal program packet to MPEG-2 coding, once at the higher data rate and
once at the lower data rate, and to combine the two packets in two multi-
plex packets transported in independent transport streams. Higher data rate
automatically means better (picture) quality. The data stream with the
lower data rate and correspondingly lower picture quality is fed to the high
priority path and that with the higher data rate is supplied to the low prior-
ity path. At the receiving end, the high priority signal is demodulated more
easily than the low priority one. Depending on the conditions of reception,
the HP path or the LP path will be selected at the receiving end. If the re-
ception is poor, there will at least still be reception due to the lower data
rate and higher compression, even if the quality of the picture and sound is
inferior.