EN RI TN A IM NM NT T ENTE RTTEA EEN

DIRECT-TO-HOME TV:
TRANSMISSION AND RECEPTION
Direct-to-home (DTH) TV is all set to revolutionise the Indian entertainment scenario. It promises quality TV reception even in shadow areas and many value-added services such as the Internet and interactive multimedia. Find out how the DTH system works, followed by different architectures for DTH reception
GP. CAPT. K.C. BHASIN

n direct-to-home (DTH) telecast, TV channels/programmes are directly distributed via satellite to the subscribers homes without the intervention of a cable operator. The signals are transmitted in Ku band (10.7 GHz to 18 GHz) and are received by the subscribers through a small dish antenna (about 45cm in dia.) and a set-top box (or an integrated receiver decoder). The DTH system can also provide many value-added services such as the Internet, e-mail, data casting, e-commerce, and interactive multimedia. It has the provision for a subscriber management system similar to the one for conditional access system (CAS). The current means of broadcasting in India dont provide quality reception in shadow areas, particularly in the north-eastern region. The DTH can fill this void easily. All in all, DTH offers immense opportunities to both the broadcasters and the viewers. The detailed guidelines for starting the DTH service in India were issued by the

SESAT satellite from Eutelsat connecting East and West

DTH receiver from Eutelsat

government on March 15, 2001, followed by guidelines in March 2003 for uplinking of foreign-owned news channels. So far only three companies have applied for starting the DTH broadcasting service. These are Space TV, A.S.C. Enterprises, and Essel Shyam Communications. Essel Shyam hopes to uplink around 250 channels from its teleport facility at Noida in two to three years. Doordarshan is also planning to launch DTH television satellite with Prasar Bharati allocating Rs 5000 million for the project, says Ravi Shankar Prasad, Information & Broadcasting Minister.
AUGUST 2003 ELECTRONICS FOR YOU

ENTERTAINMENT

Fig. 1: Functional block diagram of a DVB-S channel

The DVB core system

Generally, a digital video broadcasting (DVB) system valid for all media carries a flexible combination of MPEG-2 video, audio, and data using the common MPEG-2 transport stream multiplex. The common service information system gives details of the programmes. Modulation and additional channel coding system, if any, are chosen to meet the requirements of different transmission media. A common scrambling system and a common conditional access interface are available. DVB-S is the DVB standard for satellite delivery. It is an extension of MPEG-2 standard with specific instructions for implementing the satellite links. This standard is widely adopted.

DVB-S transmission
Many of the formats and transmission aspects of satellite DVB services are standardised by international bodies such as the International Organisation for Standardisation (ISO) and the International Telecommunications Union (ITU). The standard developed by the European Telecommunication Standard Organisation (part of ITU) applies to Kuband satellites operating at 11/12 GHz. It is designed to provide quasi error-free (QEF) service at bit error rates (BERs) of 10-10 to 10-11. By using a fairly robust error-prevention scheme, which can be

varied depending on the channel environment, it can provide this QEF rate to channels with non-corrected error rates of 10-1 to 10-2. The functional block diagram of a DVB-S channel is shown in Fig. 1. The MPEG encoder unit can take in several compressed video channels (including the programme audio and other digital data). All the data is compressed to produce a single MPEG data block of 188 bytes. Latest MPEG encoders can compress together up to ten regular video channels. The whole DVB-S system operates in the time-division multiplexing (TDM) mode. The input data must be in 188-byte blocks with 1-byte sync word at the beginning. Energy dispersal. The MPEG blocks are shuffled to improve the output spectrum. Data coding. A Reed Solomon code (204/188) is applied to the data. This coding can correct up to eight errors. In it, 16 bytes of overhead are added to the 188 bytes from the MPEG encoder. On the receive side, the Reed Solomon decoder can take in data coming at a BER of approximately 10-4 and convert it into a BER of 10-10 or lower. Interleaving. The data is then optionally Fornay interleaved (convolutional interleaving with depth 12). It is delimited by occasional sync packets. On the receive side, the interleaver provides a gain of approximately 3 dB. This enhances the ability to correct burst errors that have been missed by the inner convolutional

decoder. Inner code. The data is then convolutionally coded depending on the transponder size and channel quality desired. (Increasing the code rate reduces the redundancy from the base rate. Increasing the code rate increases the information rate and hence the error rate but reduces Eb/N0 requirements.) The basic code rate is with K=7. But this rate can be increased by puncturing the code at code rates of 2/3, 3/4, 5/6, 7/8, and others. Each code rate is tried and then locked using the sync data. On the receive side, the convolutional decoder can take in a service quality of 10-2 and improve it to an error rate of 10-4. Baseband pulse shaping. Baseband pulses are then gray-coded and root-raised cosine filtered. The roll-off rate is 0.35. QPSK modulation. This single carrier is now quaternary phase shift keying (QPSK) modulated. The table shown below provides data rates and transponder sizes. This is an example, and the parameters for individual systems may vary. The code rate is dynamically variable. So when the link is clean, the transmitter may be transmitting at a high code rate (less overhead). But if the link deteriorates, say, due to rainfall, the transmitter switches to a higher coder rate to provide the same BER. This means that the system must be designed to operate in the worst condition. To guarantee a certain link quality, the system must provide the highest Eb/N0 listed in the table.

Data Rates and Transponder Sizes

Transponder bandwidth (MHz) 24-27 QPSK Coded symbol rate bitrate (MS/s) (Mbps) 19.5 39 Convolutional code rate Reed Solomon code rate 188/204 (=0.922) Information bitrate (Mbps) 16.30 18.00 21.60 24.40 27.00 28.80 30.00 31.40 Eb/N0 (dB)

Characteristics of DVB-S carriers

Satellite access modes. The multiple-channel per carrier (MCPC) mode of operation is used for transmission of a high-rate multiplex (typically, 38Mbps) comprising four to eight standard definition digital television programmes. Usually, such transmissions use a dedicated (single-carrier) transponder. The single-channel per carrier (SCPC)

5/11=0.45 1/2=0.50 3/5=0.60 2/3=0.67 3/4=0.75 4/5=0.80 5/6=0.83 7/8=0.88

4.00 4.00 4.50 4.80 5.00 5.60 6.20 7.00

ELECTRONICS FOR YOU

AUGUST 2003

ENTERTAINMENT
mode of operation is used for transmission of a medium-rate multiplex (typically, 4-8Mbps), which usually comprises a single digital television programme. Such transmissions use part of the satellite transponders bandwidth and power. In other words, the transponder is accessed in the multi-carrier mode, i.e. the transponder resources are shared amongst several users (uplink carriers). The modulation, coding, and multiplexing structures of MCPC and SCPC carriers are identical and are as specified in the DVB-S standard. However, SCPC and MCPC transmissions differ in transmission (symcan have a symbol rate as low as 3 MS/s, depending upon the amount of information carried, for example, one or more digital TV programmes. The upper limit on the symbol rate extends beyond the lower limit for MCPC, which means that the equipment capable of receiving both SCPC and MCPC carriers would need to be able to process symbol rates over the full 3-30MS/s range. Radiated power levels. Effective isotropic radiated power (EIRP) is a measure of the signal strength that a satellite transmits towards the earth below. The EIRP is highest at the centre of the beam and de-

The Bandwidth Available and the Symbol Rate

The carrier on the satellite is made up of a sequence of pulses joined together to make a continuous signal. Each pulse is a symbol. According to the modulation method, each symbol represents 1, 2, 3, etc bits of transmission rate data. In phase shift keying (PSK) modulation, each pulse is a burst of carrier signal with its sinewave-zero crossing point timing adjusted forwards or backwards in time to constitute a phase shift. Phase shifts of 180 apply in BPSK and 90 in QPSK. A phase shift of 90 represents a time shift of 1/4 of a full cycle of the sine wave. The closer the spacing phase shifts, the more difficult the distinction between them at the receive end. So for each higher-order PSK scheme, a higher carrier-to-noise ratio is required. As a general rule, if you have bandwidth to spare, use a lower-order modulation or a lowrate FEC (like 1/2 or 2/3) to spread out the signal. If you have power to spare, use a higherorder modulation and/or a higher-rate FEC (like 3/4 or 7/8). Ideally, you would want to use all of the available bandwidth and power simultaneously. If you use larger receive dishes, you will always be able to increase the system capacity. If you are doing a point-to-point link, it is worth using larger dishes. If you have thousands of receive dishes, the aggregate cost of these is significant and you will want to allow smaller sizes even though this reduces system capacity and increases space segment costs. FEC is applied to the customers information data at the transmit end, so transmission data rate = customer information rate x 1/FEC rate FEC rate is typically 0.5 to 0.9, so the transmission data rate is always significantly higher than the customer information rate. The symbol rate is related to other quantities as per the following relationship: DR SR = m x CRv x CRrs where SR is the symbol rate, DR is the data rate (or the customer information rate), CRv is the Viterbi FEC code rate (typically, 1/2, 2/3, 3/4, 5/6, or 7/8), CRrs is the Reed Soloman FEC code rate (typically, 168/204), and m is the modulation factor or transmission rate bits per symbol (BPSK=1, QPSK=2, 8PSK=3, etc). On a spectrum analyser, the 3dB bandwidth is approximately the same as the symbol rate.

bol) rates; radiated power levels; frequency accuracy, frequency stability, and phase noise performance of reception equipment; and channel hopping response times. These may also employ different degrees of forward error correction (FEC). Symbol rates (refer the box). Typically, MCPC transmissions have a symbol rate of 20 to 30 symbols per second (S/ s), depending upon the bandwidth available for transmission. A symbol rate of 27.5 MS/s is compatible with 33MHz satellite transponder bandwidth and is the most commonly used. A symbol rate of 30 MS/s is compatible with 36MHz transponder bandwidth. SCPC transmissions

creases at angles away from the bore-sight. MCPC transmissions essentially employ all of the available satellite transponder power with little or no transponder back-off. (Transponder back-off means operation of the satellites highpower amplifier below its maximum output level in order to reduce the adverse effects of channel non-linearities on the transmission quality.) SCPC transmissions share the satellite transponder with other SCPC transmissions and consequently the EIRP assigned to each carrier is significantly less than the transponders maximum EIRP capability. The back-off per carrier is at least 5 dB
AUGUST 2003 ELECTRONICS FOR YOU

ENTERTAINMENT

Fig. 2: Functional block diagram of a typical Ku-band satellite transmit terminal

Fig. 3: Antenna for the transmit terminal

and is frequently much higher, depending upon the number of SCPC transmissions that share the transponder. Consequently, for a given satellite coverage (EIRP), a larger antenna may be required to receive an SCPC transmission than is required to receive a MCPC transmission. It is for this reason that MCPC access is preferred for DTH applications. Forward error correction. Both MCPC and SCPC transmissions employ the full range of FEC rates specified in the DVB-S standard (1/2, 2/3, 3/4, 5/ 6, and 7/8). MCPC transmissions frequently employ an FEC rate of 3/4 but are not constrained to do so. SCPC signals are transmitted with a lower power and consequently these often employ a higher FEC rate PanAmSats Castle Rock Teleport ground facility
ELECTRONICS FOR YOU AUGUST 2003

(2/3 or 1/2) in order to minimise the size of the antenna needed for service reception. Frequency stability and phase noise considerations. The frequency stability and phase noise performance of outdoor reception systems designed for FM TV services may be adequate for reception of MCPC digital TV transmissions. However, use of a digital-ready low-noise block converter (LNB) will guarantee reception of all MCPC transmissions. When choosing a digital-ready LNB for SCPC reception, the most important parameters to be considered are the initial frequency accuracy and the temperature stability. The device should be selected for the best initial frequency precision and the best tempera-

ture stability (<1 to 2 MHz over the operating temperature range). Channel hopping considerations. The response time during channel hopping can be acceptably short with MCPC access, so long as the service information is delivered at an adequate rate. The maximum delay is likely to occur when switching from one multiplex to another multiplex, which requires retuning of the receiver to the new carrier frequency. With SCPC access, the response time can be as high as 5 seconds. This is partly because the data transmission rate is much lower than for MCPC transmissions, leading to a lower rate of service information transfer for the same degree of overhead (percentage of the capacity allocated to the service information). Switching between different multiplexes, and hence receiver retuning, will also occur more frequently, as SCPC transmissions will generally carry only one or very few digital TV programmes. Spectrum inversion. The DVB-S standard specifies the use of QPSK for transmission via satellite and stipulates the correct way to map bits from the in-phase (I) and quadrature (Q) baseband bitstreams onto the four phase states. Whilst most transmissions comply with this mapping, some transmissions are made with the I and Q data streams interchanged. The result is spectrum inversion, which affects both broadcasters and receiver manufacturers. Spectrum inversion can be present regardless of the ac-

ENTERTAINMENT

(a) Single-feed system

(c) Switched dual-feed system

(e) Mono-block system

(b) Dual-feed system Fig. 4: DTH receiver architecture

(d) Loop-through dual-feed system

cess mode (MCPC or SCPC). Fig. 2 shows the functional block diagram of a typical Ku-band satellite transmit terminal. The antenna for the transmit terminal is shown in Fig. 3.

DVB-S reception
Reception system architecture. In general, reception systems are of two types, namely, individual reception systems and collective reception systems. Individual reception systems are simply referred to as DTH systems, while collective reception systems are often referred to as satellite master antenna television (SMATV) systems. The installed reception systems should have the maximum configuration flexibility, regardless of their architecture. Given that the television services of general interest exist at more than one orbital position of a satellite, to provide the consumers with the maximum choice of programmes, DTH reception systems

should be capable of receiving the telecast from at least two orbital positions of the satellites telecasting the TV programmes. Various possible architectures are discussed below. Fig. 4(a) shows a single-feed system for reception. It utilises a single universal LNB to receive all signals in the 10.712.75GHz range on any one of the two linear polarisations. Frequency band switching and polarisation switching are performed within the LNB. Figs 4(b)-(e) show systems for reception from two orbital positions via a dualfeed assembly. Single-antenna architectures such as these are recommended for DTH systems receiving from two orbital positions in order to avoid the cost and environmental impact associated with installing two separate antenna sub-systems. Systems with a single mechanically steerable antenna, though having a longer response time when switching between orbital positions, would also suit this pur-

pose. Their architecture differs in cabling and signalling requirements for the connection between the IRD and the antenna sub-system. In Fig. 4(b), two dual-band LNBs share a common reflector. The system is intended for simultaneous reception from the satellites normal orbital location and a closeby orbital location (typically within 6). Frequency band switching and polarisation switching are performed independently within the LNB for each orbital position. Two cables are required for connection to the IRD (indoor unit), which must be equipped with two LNB inputs that can be independently controlled. Fig. 4(c) shows an alternative dualfeed architecture that additionally allows switching between the two LNBs (orbital positions) to be performed at the antenna under the control of the IRD. Only one cable is required for interconnecting the antenna sub-system and the IRD, and the IRD needs to be equipped with only a single LNB input. Fig. 4(d) shows a dual-feed architecture that is logically and functionally similar to the architecture in Fig. 4(c), but uses internal switching in one of the LNBs to provide an LNB loop-through facility. It also requires a single cable for interconnecting the antenna sub-system and the IRD. Switching within the LNB eliminates the need for a separate switchbox, leading to potential cost and weight savings and reducing the number of connections that require waterproofing. Fig. 4(e) shows a mono-block universal LNB for reception from two satellites spaced 6 apart. Reception is possible across the full 10.7-12.75GHz frequency
AUGUST 2003 ELECTRONICS FOR YOU

ENTERTAINMENT
communication, so that the IRD can obtain feedback on the settings of remotely configured equipment, and as a means for signaling between the outdoor and indoor units. For the purpose, a proprietary communication protocol named DiSEqC (developed for European Telecommunications Satellite Organisation) is available. The same can be requested from the Website http:// www.eutelsat.com. The output from the LNB is brought to the IRD receiver for processing. The functional block diagram of the DVB-S IRD receiver is shown in Fig. 5. Here, after RF amplification, the signal undergoes QPSK demodulation followed by error correction. The demodulated output is passed through a CAS controlled decryption stage. This stage is followed by an MPEG-2 transport stream demultiplexer to feed the MPEG-2 based video and audio decoders as well as a data interface stage. The audio and video outputs from the respective decoders are processed by a PAL TV encoder for connection to a TV set, VCR, or Hi-Fi audio system, as appropriate. The data interface is used for extracting any teletext and electronic programme guide (EPG) information for use in TV/VCR. EPG is a combination of hardware and software to enable viewers to easily navigate the large number of channels and select the desired service. The data stream is also interfaced, via modem, to a PC for e-mail, the Internet, and other applications.

Fig. 5: Functional block diagram of DVB-S IRD receiver

band and on both polarisations from each orbital position. Such devices are simpler to install than equivalent systems employing two physically separate LNBs (see Figs 4(b)-(d)) to achieve the same purpose. These allow existing antennae receiving from a single satellite position to be easily upgraded for reception from two orbital positions. A single cable is required for interconnecting the LNB and the IRD. This is the most preferred solution. All the architectures shown in Figs 4(a)-(e) require signaling from the IRD so that the appropriate frequency band and polarisation can be selected at the antenna sub-system. Depending upon the architecture, this can be done using either conventional tone signalling methods or by means of a compatible communications protocol. Conventional signalling (control) method. Receiving systems, regardless of their architecture, should operate over the full 10.7-12.75GHz frequency range and on both linear polarisations (horizontal and vertical) in order to receive all of the services available from one or more orbital positions of the programme telecasting satellite. The simplest form of the receiver architecture consists of an antenna equipped with a single feed (LNB) and connected by a single cable to the IRD (refer the DTH reception architecture of Fig. 4(a)). A single LNB should work over the full frequency range and on both polarisations. As the bandwidth employed by the satellite broadcasting system is too large
ELECTRONICS FOR YOU AUGUST 2003

to be directly implemented in the LNB, the channel is generally received in one of the two frequency bands, i.e. in the low frequency band (10.711.7 GHz) or in the high frequency band (11.712.75 GHz). The LNB is switched to receive in a frequency band that is appropriate for the desired satellite channel. The LNB also receives on one of the two linear polarisations, i.e. horizontal (X) or vertical (Y). It is switched to receive in a polarisation that is appropriate for the desired satellite channel. Consequently, the simplest reception architecture requires four signaling states to be communicated from the IRD to the antenna sub-system: 1. Low-band, horizontal polarisation 2. Low-band, vertical polarisation 3. High-band, horizontal polarisation 4. High-band, vertical polarisation The conventional solution is to use continuous-tone signaling and variation of the LNB power supply voltage for this purpose. The cable that conveys the received satellite signals to the IRD also carries these control signals. The dual-feed DTH system that receives from two different orbital positions needs to select the appropriate feed (LNB) as well as the frequency band and polarisation. Additional signaling states are therefore required. The number of additional states depends upon the complexity of the receive system. In general, SMATV systems require more complex signaling than single-user DTH systems. It is also desirable to have two-way

Challenges ahead
A study by European Space Agency on next-generation broadcasting satellites says, ..Today, two-way satellite Internet service, compared to cable or DSL, is an inferior service despite being overpriced and those providing two-way satellite connections are losing money. The main focus for future developments in satellite broadcasting, both for the home and professional user services, should be the continuous reduction of the cost and improvement of the quality of service of two-way interactive satellite connections. The first priority is to prepare the advent in the 2005-07 time frame of service bouquets bundling TV and light Internet services. To this end, efforts should be made to reduce the cost of low-rate interactive satellite terminals.