Itu t j.83-b: coding and mapping for cable – Atec Rohde-Schwarz-SFQ Series User Manual
Page 9
TV Test Transmitter R&S
®
SFQ
9
whose depth can be selected specifically
for each layer. Delay adjustment is also
assigned to the time interleaver in order
to compensate for different delays in the
paths.
Subsequent frequency interleaving
scrambles the data in an OFDM symbol,
i.e. in the frequency domain. First an
inter-segment interleaver is applied bet-
ween the OFDM segments that have the
same modulation, followed by an intra-
segment interleaver that rotates the data
in a segment. Finally, the data passes
through an intra-segment randomizer
that shifts the data in a segment to quasi-
random positions.
The next step is OFDM framing. Frames
are formed from 204 OFDM symbols by
adding pilot carriers. Depending on the
mode and the selected modulation, pilot
carriers are inserted into the data stream
at different positions. Moreover, TMCC
(transmission and multiplexing configura-
tion control) carriers and AC (auxiliary
channel) carriers are added.
The data generated in this way under-
goes inverse fast Fourier transform (IFFT)
to transfer it from the frequency domain
to the time domain as is usual with OFDM
modulation. The length of IFFT depends
on the selected ISDB-T mode and can be
2K, 4K or 8K.
IFFT is followed by the insertion of the
guard interval. This guard interval extends
the OFDM symbols by a specific factor
(1/4, 1/8, 1/16 or 1/32). This measure has
a positive effect on the receiving charac-
teristics in the case of multipath propaga-
tion and mobile reception.
ITU T J.83-B: coding and
mapping for cable
The symbol rate of the coder and conse-
quently the bandwidth of the output sig-
nal can be varied over a wide range of
±10% of the standard symbol rate.
Larger variations of the symbol rate can be
made in the TS parallel mode, where the
symbol rate of the coder immediately fol-
lows the coder input data rate. However,
conformance with specifications cannot
be warranted outside the range ±10%.
The data signal applied to the R&S SFQ
can be replaced with an internal test
sequence (NULL TS PACKETS, NULL PRBS
PACKETS, SYNC PRBS), which is helpful
for BER measurements.
Coding: The coder expects
an MPEG-coded input
data stream packetized to
standard with a packet
length of 188 bytes. The
data is divided into pack-
ets by means of a sync
byte (47 hex) in the trans-
port stream, the sync byte
also being used for
receiver synchronization.
In the J83-B cable transmission system,
additional error control is introduced at
the transport stream level by means of a
sliding checksum, calculated for the
transport stream packets, and substituted
for the sync byte. This check sum byte
allows the receiver to synchronize to the
packets and to check for errored packets.
The J83-B FEC layer, which is next,
accepts and transports data without any
restrictions imposed by the protocol, i.e.
checksum generation and FEC coding are
completely independent processes.
FEC in the J83-B system is implemented
in the four following stages to ensure reli-
able data transmission via cable:
◆
Reed-Solomon coding (128, 122) for
outer error correction, allowing up to
three symbols in a Reed-Solomon
block to be corrected
◆
A convolutional interleaver distribut-
ing consecutive symbols uniformly
across the data stream, thus protect-
ing the signal from burst-type impair-
ments
◆
A randomizer to give a uniform power
density in the channel
◆
Trellis coding for inner error correc-
tion, involving convolutional coding of
data and adding defined redundant
information to the symbols, thus en-
abling the receiver to detect and cor-
rect any sporadic impairments on the
transmission path by means of soft-
decision methods
The randomizer, interleaver and Reed-
Solomon coder can be switched off,
which is very useful when receivers are
being developed.
J.83B