Terrestrial Transmission of Digital Television Signals (DVB-T) - Digital Video and Audio Broadcasting Technology - page 361

standard. The de-mapped data pass from the de-mapper into the symbol
and bit de-interleaver where they are resorted and fed into the Viterbi de-
coder. At the locations where bits have been punctured, dummy bits are
inserted again. These are dealt with similarly to errored bits by the Viterbi
decoder which then attempts to correct the first errors in accordance with
methods known from the trellis decoder.

MPEG2
TS

Video

DVBT

MPEG2

Tuner

TP

A/D

SAW

decoder

demod.

Audio

I2C bus

Micro
processor

Keyboard /
remote
control

Fig. 20.14. Block diagram of a DVB-T set-top box

The Viterbi decoder is followed by the convolutional deinterleaver
which breaks up error bursts by undoing the interleaving. This makes it
easier for the Reed Solomon decoder to correct bit errors. The Reed Solo-
mon decoder corrects up to 8 bit errors per packet with the aid of the 16 er-
ror control bytes. If there are more than 8 errors per packet, the 'transport
error indicator' is set to one and then this transport stream packet cannot be
processed further in the MPEG-2 decoder and error masking must be car-
ried out. As well, the energy dispersal must then be undone. This stage is
synchronized by the inverted sync bytes and this sync byte inversion must
also be undone, after which the MPEG-2 transport stream is available
again.
A practical DVB-T receiver (Fig. 20.14.) has only a few discrete com-
ponents such as the tuner, SAW filter, the mixing oscillator for the 2nd IF
and the low-pass filter. These are followed by a DVB-T demodulator chip
which contains all modules of the DVB demodulator after the A/D con-
verter. The transport stream coming out of the DVB-T demodulator is fed
into the downstream MPEG-2 decoder where it is decoded back into video