This system transmits an MPEG-2 or MPEG-4 family digital audio/video stream using a QPSK modulation with channel coding through satellite 11/12 GHz channels.
Fig.1 DVB-S Transmitter
The input signal is a sequence of standard MPEG transport stream packets. Each packet consist of 288 bytes. First each packet is scrambled for energy dispersion and packet synchronisation is modified. After that it goes through a Reed-Solomon encoder and 16 bytes are added to each packet for error protection. The individual packet length becomes 304 bytes.
Next, modified packets go through a convolutional interleaver with depth 12 and a convolutional error correction coder. Convolutional code rate is ½ with constant length 7. The code rate ½ is modified by puncturing that is done after the convolutional coder. After puncturing, the coding rate may take values 1/2, 2/3, 3/4, 5/6 and 7/8. Puncturing is combined with mapping to two dimensional symbols with I and Q components. Since DVB-S uses QPSK modulation, the I and Q components are binary signals.
Next, the I and Q binary components of each symbol are shaped by a square-root raised cosine filter (SRRC) with the roll-off factor α = 0.35. Finally, the shaped I and Q are modulated to the carrier frequency.
Fig 2 DVB-S Receiver
The receiver performs the opposite sequence of operations. First, the RF baseband signal is amplified and level adjusted, then downconverted and demodulated. For demodulation, carrier and timing synchronization is done. After that, the signal goes through a matched filter.
Next, demapping of the symbols to bytes combined with depuncturing is done. The depuncturer inserts zero bits for the bits punctured in the transmitter. After that the convolutional decoder is applied based on the Viterbi algorithm that corrects some errors caused by channel noise and distortions. Next, MPEG sync bytes are detected in the sync detector to provide synchronization information for the deinterleaving. The sync detector is also able to distinguish the ambiguity of the QPSK modulation that is not detectable by the Viterbi decoder.
Two stages of error correction coding allows the creation of a Quasi-Error-Free (QEF) environment that is characterised by an error rate from 10-10 to 10-11 in the output MPEG transport stream. The error rate after the Viterbi decoder is 2∙10-4. Puncturing improves the modulated signal bandwidth utilization efficiency but degrades the error correction performance at the output of the convolutional decoder.
|Inner code rate||Required Eb/No for QEF at the receiver output|
Eb –energy per transmitted bit, No – spectral power density of the channel noise