DVB-S2 is the second-generation specification for satellite broadcasting developed by DVB in 2003. It uses more advanced channel coding (LDPC codes) and modulation formats (QPSK, 8PSK, 16APSK, 32APSK) for a 30% increase in transmission capacity over DVB-S. DVB-S2 allows for adaptive coding and modulation to optimize transmission for each user. It is designed for broadcast, interactive, and professional applications with flexibility to handle different transponder characteristics and content formats.

1. DIGITAL VIDEO BROADCASTING
DVB-S2 — ready for lift off
Alberto Morello and Vittoria Mignone
RAI, Radiotelevisione Italiana
DVB-S2 is the second-generation specification for satellite broadcasting – developed
by the DVB (Digital Video Broadcasting) Project in 2003. It benefits from more recent
developments in channel coding (LDPC codes) combined with a variety of
modulation formats (QPSK, 8PSK, 16APSK and 32APSK). When used for interactive
applications, such as Internet navigation, it may implement Adaptive Coding &
Modulation (ACM), thus optimizing the transmission parameters for each individual
user, dependant on path conditions. Backwards-compatible modes are available,
allowing existing DVB-S set-top-boxes to continue working during any transitional
period.
This article, based on a presentation given at IBC-2004, gives a variety of examples of
DVB-S2 – focusing on television applications including broadcasting, contribution
TV links and DSNG, and DTT signal distribution to transmitters – which illustrate the
advantages of the new system over DVB-S and, indeed, DVB-DSNG.
The DVB-S2 system has been designed for several satellite broadband applications:
broadcast services for standard definition TV and HDTV;
interactive services, including Internet access, for consumer applications;
professional applications, such as Digital TV contribution and News Gathering, TV distribution
to terrestrial VHF/UHF transmitters;
data content distribution and Internet trunking.
It is based on a “tool-kit” approach which allows us to cover all the application areas while still
keeping the single-chip decoder at reasonable complexity levels, thus enabling the use of mass
market products also for professional applications.
The DVB-S2 standard has been specified around three key concepts: (i) best transmission perform-
ance, (ii) total flexibility and (iii) reasonable receiver complexity. To achieve the best performance-
complexity trade-off, quantifiable in about 30% capacity gain over DVB-S, DVB-S2 benefits from
more recent developments in channel coding and modulation. For interactive point-to-point applica-
tions such as IP unicasting, the adoption of the Adaptive Coding & Modulation (ACM) functionality
allows us to optimize the transmission parameters for each individual user on a frame-by-frame
basis, dependant on path conditions, under closed-loop control via a return channel (terrestrial or by
satellite): the result is an even greater gain of DVB-S2 over DVB-S.
DVB-S2 is so flexible that it can cope with any existing satellite transponder characteristics, with a
large variety of spectrum efficiencies and associated C/N requirements. Furthermore, it is not
EBU TECHNICAL REVIEW – October 2004 1 / 10
A. Morello and V. Mignone

2. DIGITAL VIDEO BROADCASTING
limited to MPEG-2 video and audio coding, but it is designed to handle a variety of advanced audio-
video formats which the DVB Project is currently defining. DVB-S2 accommodates any input stream
format, including single or multiple MPEG Transport Streams, continuous bit-streams, IP as well as
ATM packets.
Forward Error Correction (FEC) and modulation
The DVB-S2 detailed system block diagram is described in [1] and [2]. The FEC is the key
subsystem to achieve excellent performance by satellite, in the presence of high levels of noise and
interference. The selection process, based on computer simulations, compared seven proposals –
parallel or serially concatenated convolutional codes, product codes, low density parity check codes
(LDPC) – all using “turbo” (i.e. recursive) decoding techniques. The winning system, based on
LDPC, offered the minimum distance from the Shannon limit on the linear AWGN channel, under the
constraint of maximum decoder complexity of 14 mm2 of silicon (0.13 µm technology).
The selected LDPC codes [2] use very large block lengths (64800 bits for applications not too critical
for delays, and 16200 bits). Code rates of 1/4, 1/3, 2/5, 1/2, 3/5, 2/3, 3/4, 4/5, 5/6, 8/9 and 9/10 are
available, depending on the selected modulation and the system requirements. Coding rates 1/4, 1/3
and 2/5 have been introduced to operate, in combination with QPSK, under exceptionally poor link
conditions, where the signal level is below the noise level. Concatenated BCH outer codes are intro-
duced to avoid error floors at low bit error rates (BER).
Four modulation modes can be
Q
selected for the transmitted Q I=MSB
Q=LSB 100
payload (see Fig. 1). 10 110
00 000
ρ=1 ρ=1
QPSK and 8PSK are typically
010 φ=π/4
proposed for broadcast applica- I I
001
tions, since they are virtually
constant envelope modulations 011
and can be used in non-linear 11 01 101
111
satellite transponders driven
Q
near saturation. The 16APSK
Q 01101
and 32APSK modes, mainly 11101 01001
1010 1000
targeted at professional appli- 01100 R 3 00101 11001
0010 R2 0000 00001
cations, can also be used for MSB
R2
LSB 11100 00100 00000 01000
broadcasting, but these require
0110 R1 1100 0100
a higher level of available C/N 1110 10100 10101 R1 10001 10000
11000
11110
and the adoption of advanced I I
0111 10110 10111 10010
pre-distortion methods in the 1111 1101 0101 10011
01110
up-link station to minimize the 00110 00010 11010
0011 0001
effect of transponder non-line- 11111
00111 00011
01010
arity. 1011 1001
01111 11011
01011
Whilst these modes are not as
Figure 1
power-efficient as the other
The four possible DVB-S2 constellations before physical layer
modes, the spectrum efficiency scrambling
is much greater. The 16APSK
and 32APSK constellations
have been optimized to operate over a non-linear transponder by placing the points on circles.
Nevertheless their performance on a linear channel are comparable with those of 16QAM and
32QAM respectively.
By selecting the modulation constellation and code rates, spectrum efficiencies from 0.5 to 4.5 bit
per symbol are available and can be chosen dependant on the capabilities and restrictions of the
satellite transponder used.
EBU TECHNICAL REVIEW – October 2004 2 / 10
A. Morello and V. Mignone

3. DIGITAL VIDEO BROADCASTING
DVB-S2 has three “roll-off factor” choices to determine spectrum shape: 0.35 as in DVB-S, 0.25 and
0.20 for tighter bandwidth restrictions.
Framing structure
Two levels of framing structures have been designed:
the first at the physical level (PL), carrying few highly-protected signalling bits;
the second at base-band level, carrying a variety of signalling bits, to allow maximum flexibility
on the input signal adaptation.
FEC Useful
redundancy data
Type of channel
coding and modulation
adopted in the wagon
PL FRAME
H FEC FRAME H FEC FRAME H FEC FRAME
8PSK 5/6 QPSK 2/3 16APSK 3/4
Figure 2
Pictorial representation of the physical-layer framing structure
The first level of framing structure has been designed to provide robust synchronization and signal-
ling at the physical layer [2]. Thus a receiver may synchronize (carrier and phase recovery, frame
synchronization) and detect the modulation and coding parameters before demodulation and FEC
decoding.
With reference to Fig. 2, the DVB-S2 physical layer signal is composed of a regular sequence of
“lorries” (frames): within a lorry, the modulation and coding scheme is homogeneous, but may
change (Adaptive Coding & Modulation) in adjacent lorries. Every frame is composed of a payload
of 64800 bits (or 16200 bits), corresponding to a code block of the concatenated LDPC/BCH FEC,
and a Header (90 binary modulation symbols), containing synchronization and signalling informa-
tion. Since the PL Header is the first entity to be decoded by the receiver, it could not be protected
by the powerful LDPC/BCH FEC scheme.
On the other hand, it had to be perfectly decodable under the worst-case link conditions. Therefore
the system designers selected a very low-rate 7/64 block code to protect it, suitable for soft-decision
correlation decoding, and minimized the number of signalling bits to reduce decoding complexity
and global efficiency loss.
The second level of framing structure, the “baseband frame”, allows a more complete signalling
functionality to configure the receiver according to the application scenarios: single or multiple input
streams, generic or transport stream, CCM (Constant Coding & Modulation) or ACM (Adaptive
Coding & Modulation), and many other configuration details. Thanks to the LDPC/BCH protection
and the wide length of the FEC frame, the Baseband (BB) Header may contain many signalling bits
(80) without losing transmission efficiency or ruggedness against noise.
EBU TECHNICAL REVIEW – October 2004 3 / 10
A. Morello and V. Mignone

4. DIGITAL VIDEO BROADCASTING
Backwards-compatible modes
The large number of DVB-S receivers already installed makes it very difficult for many established
broadcasters to think of an abrupt change of technology in favour of DVB-S2 – especially where
there is a receiver subsidy and for free-to-air public services. In such scenarios, backwards-compat-
ibility may be required in the migration period, allowing legacy DVB-S receivers to continue oper-
ating, while providing additional capacity and services to new, advanced receivers. At the end of the
migration process, when the complete receiver population has migrated to DVB-S2, the transmitted
signal could be modified to the non-backward compatible mode, thus exploiting the full potential of
DVB-S2.
Optional backwards-compatible (BC) modes have therefore been defined in DVB-S2, intended to
send two Transport Streams on a single satellite channel. The first (High Priority, HP) stream is
compatible with DVB-S receivers (according to EN 300 421 [3]) as well as with DVB-S2 receivers,
while the second (Low Priority, LP) stream is compatible with DVB-S2 receivers only [2].
Backwards compatibility can be implemented by hierarchical modulation [4], where the two HP and
LP Transport Streams are synchronously combined at modulation symbol level on a non-uniform
8PSK constellation. The LP DVB-S2-compliant signal is BCH and LDPC encoded, with LDPC code
rates 1/4, 1/3, 1/2 or 3/5. Then the hierarchical mapper generates the non-uniform 8PSK constella-
tion: the two HP DVB-S bits define a QPSK constellation point, while the single bit from the DVB-S2
LDPC encoder sets an additional rotation ±θ before transmission. Since the resulting signal has a
quasi-constant envelope, it can be transmitted on a single transponder driven near saturation.
System performance
Dependant on the selected code rate and modulation constellation, the system can operate at
carrier-to-noise ratios from –2.4 dB (using QPSK 1/4) to 16 dB (using 32APSK 9/10), assuming an
AWGN channel and ideal demodulator (see Fig. 3). These results have been obtained by computer
Spectrum efficiency versus required C/N on AWGN channel
4.5
32APSK
Ru [bit/s] per unit Symbol Rate Rs
4.0 = modulation-constrained Shannon limit
16APSK
3.5
3.0
2.5 8PSK
2.0 DVB-DSNG
QPSK
1.5
DVB-S
1.0
0.5
0.0
-3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
C/N [dB] in Rs
Figure 3
Required C/N versus spectrum efficiency on the AWGN channel (ideal demodulation), where C/N refers to
the average power
EBU TECHNICAL REVIEW – October 2004 4 / 10
A. Morello and V. Mignone

5. DIGITAL VIDEO BROADCASTING
simulations for a Packet Error Rate of 10-7 (one erroneous Transport Stream Packet per transmis-
sion hour in a 5 Mbit/s video service). The distance from the Shannon limit ranges from 0.7 to
1.2 dB. On AWGN, the result is typically a 20 – 35 percent capacity increase over DVB-S and DVB-
DSNG under the same transmission conditions and 2 – 2.5 dB more robust reception for the same
spectrum efficiency.
The DVB-S2 system may be used in “single-carrier-per-transponder” or in “multi-carriers-per-trans-
ponder” (FDM) configurations. Fig. 3 also indicates examples of the useful bitrate capacity Ru
achievable by the system in the different modulation/coding configurations, assuming unit symbol
rate RS. The symbol rate RS corresponds to the –3dB bandwidth of the modulated signal, while
RS(1+α) corresponds to the theoretical total signal bandwidth after the modulator, with α repre-
senting the roll-off factor of the modulation. The use of the narrower roll-off α = 0.25 and α = 0.20
may allow a transmission capacity increase, but may also produce larger non-linear degradations by
satellite for single-carrier operation.
When DVB-S2 is transmitted by satellite, quasi-constant envelope modulations such as QPSK and
8PSK are power efficient in the single-carrier-per-transponder configuration, since they can operate
on transponders driven near saturation. 16APSK and 32APSK, which are inherently more sensitive
to non-linear distortions and would require quasi-linear transponders (i.e., with larger Output Back-
Off, OBO) may be improved in terms of power efficiency by using non-linear compensation tech-
niques in the up-link station [2]. In FDM configurations, where multiple carriers occupy the same
transponder, this latter must be kept in the quasi-linear operating region (i.e., with large OBO) to
avoid excessive inter-modulation interference between signals. In this case, the AWGN perform-
ance figures may be adopted for link budget computations.
Table 1 shows, for the single-carrier-per-transponder configuration, the simulated C/N degradation
using the satellite channel models and phase noise mask given in [1] (non-linearised TWTA, phase
noise relevant to consumer LNBs), at the optimum operating TWTA point (computer simulations by
ESA). CSAT is the un-modulated carrier power at HPA saturation, OBO is the measured power ratio
(dB) between the un-modulated carrier at saturation and the modulated carrier (after OMUX). The
figures show the large advantage offered by the use of dynamic pre-distortion for 16APSK and
32APSK. The large phase noise degradations quoted for APSK, and in particular for 32APSK, can
be considered as pessimistic, since they refer to consumer-type LNBs while, for professional appli-
cations, better front-ends may be adopted at negligible additional cost.
Table 1
CSAT/N loss [dB] on the satellite channel
Transmission No pre-distortion Dynamic pre-distortion Dynamic pre-distortion
Mode No Phase Noise No Phase Noise Phase Noise
QPSK 1/2 0.6 (OBO=0.4) 0.5 (IBO=0; OBO=0.4) 0.6
8PSK 2/3 1.0 (OBO=0.3) 0.6 (IBO=0; OBO=0.4) 0.9
16APSK 3/4 3.2 (OBO=1.7) 1.5 (IBO=1.0; OBO=1.1) 1.8
32APSK 4/5 6.2 (OBO=3.8) 2.8 (IBO=3.6; OBO=2.0) 3.5
Examples of possible uses of the system
Some examples may better clarify the functionalities and flexibility of DVB-S2. Starting from TV
broadcasting using constant coding and modulation, and variable coding and modulation, some
examples are given in the following to cover professional TV applications such as DSNG and DTT
distribution to transmitters. Examples on IP unicast services to consumers are instead available in
[2]. For broadcasting services, only down-link parameters have been evaluated, for satellite EIRPs
of 51 and 53.7 dBW at the service area contour. Conversely, for professional applications, link
EBU TECHNICAL REVIEW – October 2004 5 / 10
A. Morello and V. Mignone

6. DIGITAL VIDEO BROADCASTING
budget evaluations have been carried out for a typical Ku-band 36 MHz satellite with European-wide
up-link and down-link coverage, following the simplified analysis method described in [5]. Ideal
target carrier-to-noise ratios are derived from Fig. 3; implementation margins are included, as
derived from [3] for DVB-S and [6] for DVB-DSNG. The following link characteristics have been
adopted:
Up-link: ITU climatic zone L; frequency: 14.29 GHz; Atmospheric loss and rain attenuation for
99.9% of average year (a.y.): 0.2 + 5.6 dB.
Satellite: G/T(dB/°K): 4.3; transmitted EIRP at saturation: 46.5 dBW.
Down-link: ITU climatic zone K; frequency: 10.99 GHz; antenna efficiency: 60%; coupling loss:
0.5 dB, pointing loss: 0.5 dB; LNB noise figure: 1.1 dB; Atmospheric loss and rain attenuation
for 99.9% a.y.: 0.1 + 2.4 dB.
SDTV and HDTV broadcasting (CCM and VCM)
Table 2 compares DVB-S2 and DVB-S broadcasting services via 36 MHz (at –3 dB) satellite trans-
ponders in Europe, using 60 cm receiving antenna diameters. The example video coding bitrates
are: 4.4 Mbit/s (SDTV) and 18 Mbit/s (HDTV) using traditional MPEG-2 coding, or 2.2 Mbit/s (SDTV)
and 9 Mbit/s (HDTV) using advanced video coding (AVC) systems which the DVB Project is
currently defining for future applications.
The required C/N of the two systems, DVB-S and DVB-S2, have been balanced by exploiting
different transmission modes and by fine tuning the DVB-S2 roll-off factor and symbol rate. The
results confirm the capacity gain of DVB-S2 versus DVB-S, exceeding 30%. Furthermore, by
combining DVB-S2 and AVC coding, an impressive number of 21 to 26 SDTV channels per trans-
ponder are obtained, thus dramatically reducing the cost per channel of the satellite capacity. The
combination of DVB-S2 and new AVC coding schemes can favour the introduction of new HDTV
services, with an adequate number of programmes per transponder (e.g. 5 to 6), reducing the satel-
lite capacity cost increase with respect to current SDTV services.
Table 2
Example comparison between DVB-S and DVB-S2 for TV broadcasting
Satellite EIRP (dBW) 51 53.7
System DVB-S DVB-S2 DVB-S DVB-S2
Modulation & coding QPSK 2/3 QPSK 3/4 QPSK 7/8 8PSK 2/3
Symbol rate (Mbaud) 27.5 (α = 0.35) 30.9 (α = 0.0) 27.5 (α = 0.35) 29.7 (α = 0.25)
C/N (in 27. 5 MHz) (dB) 5.1 5.1 7.8 7.8
Useful bitrate (Mbit/s) 33.8 46 (gain = 36%) 44.4 58.8 (gain = 32%)
Number of SDTV 7 MPEG-2 10 MPEG-2 10 MPEG-2 13 MPEG-2
programmes 15 AVC 21 AVC 20 AVC 26 AVC
Number of HDTV 1-2 MPEG-2 2 MPEG-2 2 MPEG-2 3 MPEG-2
programmes
3 - 4 AVC 5 AVC 5 AVC 6 AVC
The DVB-S2 system may also deliver broadcasting services over multiple Transport Streams,
providing differentiated error protection per multiplex (VCM). A typical application is broadcasting of
a highly protected multiplex for SDTV, and of a less protected multiplex for HDTV. Assuming we
transmit a symbol rate of 27.5 Mbaud and use 8PSK 3/4 and QPSK 2/3 modulation, 40 Mbit/s could
be available for two HDTV programmes and 12 Mbit/s for two to three SDTV programmes, with a
difference in C/N requirements of around 5 dB.
EBU TECHNICAL REVIEW – October 2004 6 / 10
A. Morello and V. Mignone

7. DIGITAL VIDEO BROADCASTING
Distribution of multiple MPEG multiplexes to DTT transmitters
Digital Terrestrial Television (DTT) is being introduced in many countries around the world. One of
the possible solutions for distributing the MPEG streams to the digital terrestrial transmitters is via
satellite. Current systems are based on DVB-S, allowing the transmission of a single MPEG multi-
plex per signal. The result is that, for the distribution of n MPEG multiplexes, n carriers per trans-
ponder should be transmitted, requiring a large satellite HPA OBO (alternatively, n transponders are
to be used). The adoption of DVB-S2 allows the distribution of multiple MPEG multiplexes using a
single-carrier-per-transponder configuration, thus optimizing the power efficiency by saturating the
satellite HPA. Adaptive coding and modulation is not considered, since multiple stations have to
receive the same signals.
For example, a DVB-S2 signal at the symbol rate of 30 Mbaud may be transmitted on a 36 MHz
transponder using α = 0.20. Thus to transmit two DTT MUXes over DVB-S2 at 24 Mbit/s each, a
spectrum efficiency of 1.6 [bit/s/Hz] is required, corresponding to QPSK 5/6. The required C/N
would be around 6 dB in 30 MHz bandwidth. The target link availability (99.9% a.y.) could be
achieved with a 3 m up-link antenna (EIRP of 64 dBW), near-saturated transponder in clear sky, and
1.2 m receiving antennas at the terrestrial transmitter sites. Using DVB-DSNG with 8PSK 2/3 and
allocating two FDM carriers in 36 MHz at a symbol rate of 13.3 Mbaud, the required C/N would be
9 dB in the receiver bandwidth. To guarantee 99.9% a.y. link availability, the transmitted up-link
EIRP was set to 75 dBW, the transponder gain setting adjusted to achieve OBO = 5.5 dB per carrier
in clear sky and the receiving antenna size could not be less than 2 m. Therefore DVB-S2 would
allow significantly smaller receiving antennas (nearly halved diameters) and cheaper up-link
stations.
Adaptive Coding and Modulation for one-to-one services
When DVB-S2 is used for inter-
active point-to-point applica-
tions like IP unicasting, its gain
over DVB-S is even greater, if
Adaptive Coding and Modula-
tion (ACM) schemes are used.
In fact ACM allows us to
recover the so called “clear sky
margin” (4 to 8 dB of power),
typically wasted in conventional Gateway High bitrate
“constant coding and modula- ACM DVB-S2 forward link
tion” satellite links, thus modulator
doubling or even tripling the
average satellite throughput Modulation & coding
selection C/N+I & MODCOD
and reducing dramatically the signalling
service cost [2]. The ACM gain Source rate Satellite
versus CCM increases for crit- control (*) terminal
Return
ical propagation conditions: channel
therefore ACM is fundamental Info
for the higher frequency bands SOURCE(s) C/N+I measurement
(e.g. Ka band) and for tropical
climatic zones. (*) Source rate control may be directly applied to source(s) or locally at the GTW input,
or via Network Traffic Control
Fig. 4 shows the scheme of an
Figure 4
ACM satellite link, composed of
Block diagram of a DVB-S2 ACM link
an ACM Gateway (GW), the
DVB-S2 ACM modulator, the
up-link station, the Satellite and the Satellite receiving Terminal (ST) connected to the ACM GW via
a return channel. The DVB-S2 ACM modulator operates at constant symbol rate, since the available
EBU TECHNICAL REVIEW – October 2004 7 / 10
A. Morello and V. Mignone

8. DIGITAL VIDEO BROADCASTING
transponder bandwidth is assumed to be constant. ACM is implemented by the DVB-S2 modulator
by transmitting a TDM sequence of frames, where coding and modulation format may change
frame-by-frame. Therefore service continuity is achieved, during rain fades, by reducing user bits
while increasing at the same time the FEC redundancy and/or modulation ruggedness.
Physical layer adaptation is achieved as follows:
1) Each ST measures the channel status (available C/N+I) and reports it via the return channel to
the Gateway (GW);
2) The ST reports are taken into account by the GW while selecting the assigned protection level
for data packets addressed to the ST;
3) In order to avoid information overflow during fades, a user bitrate control mechanism should in
principle be implemented, adapting the offered traffic to the available channel capacity. This can
be implemented in various ways, according to the specific service requirements and network
architecture. The GW imposes error protection, applied to a given portion of user data via suit-
able interfacing mechanisms. With respect to one-to-one services (e.g., DSNG), IP unicast links
using DVB-S2 ACM must adapt the error protection on a user-per-user basis: the number of
users may be very large (e.g. up to hundreds of thousands). Furthermore, direct source rate
control may be impossible, since information sources (IP information providers) are far from the
satellite GW. Further details on this application scenario is given in [2].
A crucial issue in ACM systems is the physical layer adaptation loop delay, as it is strictly linked to
the system capability of tracking channel variations. If loop adaptation is fast, service continuity
may be guaranteed even during fast rain fades while, at the same time, keeping low C/N transmis-
sion margins to maximize the overall system throughput. Since maximum C/N+I variation rates at
Ka band have been estimated to be of about 0.5 dB per second during heavy rain fades [2], and
since the C/N distance between two adjacent DVB-S2 protection levels is around 1 dB, control loop
delays smaller than 1 second should allow minimization of transmission packet losses.
TV contribution and DSNG services using ACM
ACM techniques look very promising to improve the performance of point-to-point and point-to-
multipoint TV contribution links (e.g. DSNG), where a single TS is sent to a unique or multiple (few)
Abbreviations
ACM Adaptive Coding and Modulation FDM Frequency Division Multiplex
AHG Ad hoc Group FEC Forward Error Correction
APSK Amplitude Phase-Shift Keying HP High-Priority
ATM Asynchronous Transfer Mode HPA High Power Amplifier
AVC (MPEG-4) Advanced Video Coding IBO Input Back-Off
AWGN Additive White Gaussian Noise IP Internet Protocol
BCH Bose-Chaudhuri-Hocquenghem (code) LDPC Low Density Parity Check
BER Bit-Error Ratio LNB Low-Noise Block
C/N Carrier-to-Noise ratio LP Low-Priority
CCM Constant Coding and Modulation MUX Multiplex / multiplexer / multiplexing
DSNG Digital Satellite News Gathering OBO Output Back-Off
DTT Digital Terrestrial Television OMUX Output Multiplexer
DVB Digital Video Broadcasting PER Packet Error Rate
DVB-S DVB - Satellite PSK Phase-Shift Keying
DVB-S2 DVB - Satellite, version 2 QAM Quadrature Amplitude Modulation
EIRP Effective Isotropic Radiated Power QPSK Quadrature (Quaternary) Phase-Shift Keying
ESA European Space Agency TWTA Travelling-Wave-Tube Amplifier
ETSI European Telecommunication Standards VCM Variable Coding and Modulation
Institute
EBU TECHNICAL REVIEW – October 2004 8 / 10
A. Morello and V. Mignone

9. DIGITAL VIDEO BROADCASTING
receiving stations. In this case, the TS packets protection must follow the C/N+I variations on the
satellite channel in the receiving location. Constant Transport Stream bitrate and end-to-end delay,
as required by MPEG, may be guaranteed by using DVB-S2 stream adaptation tools which are
described in detail in [2]. In order to avoid stream overflow when the channel throughput is reduced,
a variable bitrate (VBR) video encoder has to be adopted.
Let us consider first the examples of TV contribution services using large transmitting and receiving
stations to access a typical 36 MHz transponder with four FDMA signals. Using the DVB-DSNG
standard and 16QAM 3/4 mode, four TV contribution signals at 18.5 Mbit/s may be allocated in the
transponder (76 dBW up-link EIRP, 14.3 global IBO and 7 m transmitting/receiving antenna diame-
ters, 99.9% a.y. link availability). Using DVB-S2, 16APSK 5/6 and roll-off = 0.2, the information rate
of each link can be increased to 24.75 Mbit/s, thus confirming a bitrate gain of DVB-S2 over DVB-
DSNG of more than 30%. On the other side, keeping the same bitrate as DVB-DSNG, the better
performance of DVB-S2 may be used to significantly decrease the dimension of the transmitting/
receiving antennas down to 4.5 m (8PSK 5/6, 74 dBW up-link EIRP, 13 dB global IBO). With the
same 4.5 m antennas, but adding the ACM functionality of DVB-S2, the useful bitrate would be
increased again to 24.75 Mbit/s, at least under clear sky conditions.
The advantages of DVB-S2 and ACM are also evident for DSNG services. For example, in a 9 MHz
satellite bandwidth slot, a DSNG van with 1.2m antenna (61 dBW up-link EIRP) may transmit
19.8 Mbit/s in clear sky conditions (16APSK 2/3, roll-off = 0.2) and switch to 14.85 Mbit/s under
heavy fading (8PSK 2/3). For sake of comparison, DVB-DSNG with QPSK 7/8 would allow the
transmission of 10.7 Mbit/s only.
As a last example, let us consider a fly-away emergency DSNG station, with a 90 cm antenna and
only 12 W HPA power (99.9% a.y., 49 dBW up-link EIRP, 12 dB global IBO, 4 m receiving antenna,
four signals per transponder). Using DVB-S2 and ACM, 9.9 Mbit/s (QPSK 2/3, roll-off = 0.2) would
be available in clear sky, 8.9 Mbit/s (QPSK 3/5) under typical propagation conditions and 3.68 Mbit/s
(QPSK 1/4) under critical link conditions. This would offer a good picture quality using MPEG-2
video coding, and an excellent quality using new AVC encoders. DVB-S (QPSK 1/2) would require a
5 dB more powerful station to offer a constant bitrate of 6.1 Mbit/s.
Conclusions
The DVB Project does not see DVB-S2 replacing DVB-S in the short term for conventional TV
broadcasting applications. Millions of DVB-S decoders are already operating reliably and contrib-
uting to successful digital satellite businesses around the world. New applications are being envis-
aged for satellite environments such as the delivery of consumer HDTV and the delivery of IP-based
services. Two examples can highlight the revolution we have in front of us. Combining DVB-S2 and
new video and audio coding schemes (e.g. H.264), some 20 – 25 SDTV or 5 – 6 HDTV programmes
may be broadcast in a conventional 36 MHz transponder. In the area of professional TV applica-
tions, the ACM tool may offer very large benefits for one-to-one connections, such as fly-away small
DSNG stations. In these new application areas, DVB-S2 will do what DVB-S could never have
done.
Acknowledgements
The authors wish to thank the members of the “DVB-S2 AHG” of the DVB Technical Module for their
excellent contributions during the system design and the provision of many simulation results
presented in this article. In particular we wish to acknowledge Hughes Network Systems (HNS) and
the European Space Agency (ESA).
EBU TECHNICAL REVIEW – October 2004 9 / 10
A. Morello and V. Mignone

10. DIGITAL VIDEO BROADCASTING
Alberto Morello graduated in Electronic Engineering from Turin Polytechnic Univer-
sity in 1982 and took his doctorate degree in 1987. He joined the Research and
Technical Innovation Centre of RAI-Radiotelevisione Italiana in 1984 where he is
now Director. He is engaged in research on digital modulation and coding tech-
niques for video, audio and data transmission and broadcasting – via terrestrial and
satellite channels.
Dr Morello is a member of several international EBU and DVB groups and has par-
ticipated in various EUREKA, ACTS and IST projects. He was Chairman of the DVB
ad-hoc groups which defined the technical specifications for the DVB-S and DVB-
DSNG systems, and now for DVB-S2 system. He is the author of various technical
and scientific articles and has presented numerous papers, relevant to his studies and research work, at
national and international events.
Vittoria Mignone was born in 1966 and received the “Laurea in Ingegneria Elet-
tronica” degree from Turin Polytechnic in 1990.
In 1991, in co-operation with the Electronics Department of Turin Polytechnic, Ms
Mignone was engaged in studies on satellite broadcasting, on behalf of the National
Research Council. Since 1992, she has been with the RAI Research and Technical
Innovation Centre in Turin, involved in studies to define the ETSI Standards for dig-
ital television broadcasting via satellite, cable and terrestrial channels, and for
DSNG. Her current activities are in the field of advanced digital modulation and
channel coding techniques for satellite and terrestrial transmissions. She is the
author of various technical papers.
References
[1] ETSI: Draft EN 302 307: Digital Video Broadcasting (DVB); Second generation framing
structure, channel coding and modulation systems for Broadcasting, Interactive Serv-
ices, News Gathering and other broadband satellite applications (DVB-S2).
[2] The DVB-S2 standard for broadband satellite systems
Special Issue of the International Journal of Satellite Communications Networks, 2004.
[3] ETSI: EN 300 421: Digital Video Broadcasting (DVB); Framing structure, channel coding
and modulation for 11/12 GHz satellite services
[4] Chris Weck : Hierarchical modulation — the transmission of two independent DVB-T
multiplexes on a single TV frequency
EBU Technical Review No. 294, April 2003
[5] A. Morello and V. Mignone: New DVB standard for DSNG — and contribution satellite links
EBU Technical Review No. 277, Autumn 1998
[6] ETSI: EN 301 210: DVB: Framing structure, channel coding and modulation for DSNG
and other contribution applications by satellite.
EBU TECHNICAL REVIEW – October 2004 10 / 10
A. Morello and V. Mignone