Skip to main content


This two-part article provides an overview of new and innovative solutions for delivering local content using advanced, multiplexing technology to bridge the satellite and digital terrestrial broadcasting worlds.

When designing next-generation DVB-T2 networks, terrestrial operators are driven by the desire to achieve maximum coverage in the most cost-effective manner, while at the same time delivering new high-definition TV services and local content using an infrastructure that is prepared for future multi-device services.

When building these next-generation networks, there is usually a requirement to cover a minimum of 90 percent to 95 percent of the country with digital terrestrial services. Reaching the entire population has often been viewed as not practical due to difficult terrain and the costs involved in overcoming topographical challenges. By using DTH satellite infrastructure and advanced, deterministic multiplexing technology, it is now possible to use the same DTH satellite signal to feed the terrestrial transmitters, as a gap-filler, in areas where terrestrial coverage is difficult.

The same deterministic, multiplexing technology can be used to feed multiple SFN regions, each with its own unique local content, where each local signal is generated from a common feed. (See Figure 1.)

Deterministic processing can also be utilized in IP networks to feed terrestrial transmitters. This increased flexibility provides operators with much greater choice in how best to design and deploy new DVB-T2 distribution networks.

Ultimately, the successful deployment of the next-generation DVB network will be determined by viewer satisfaction. Delivery of new HD channels, uninterrupted TV services, local content, and a correct and comprehensive Program Guide (EPG) are key to achieving this. However, there are challenges to terrestrial multiplexing.


SFNs enable the highly-efficient use of broadcast capacity. SFN bands can be used in multiple adjacent areas and are capable of covering a large geographical footprint. Another advantage of SFNs is that they greatly simplify the addition of fill-in transmitters to cover shadow areas in valleys or city centers. By using SFN technology, it is possible to build a network with a number of cost-effective, low-power transmitters instead of a single, large transmitter per region. Making use of low-power, simple transmitters, located in good positions, means it's possible to cover the region's receiver population for a much lower cost. DVB-T2 has made SFN networks even more attractive as the increased number of carriers in the COFDM signal offers extension of the guard interval, providing significant gains in the size of an SFN area compared to DVB-T operation.

There is, however, a trade-off. SFN operation significantly impacts the flexibility to modify signals. A common requirement in terrestrial systems is the insertion of local content into a national multiplex. For robust SFN operation, it is not possible to add a local program using straight-forward multiplexing. Additional processing is required for SFNs in order to insert the local content in such a way that the transmitted signal is identical — bit-for-bit — at all sites within the SFN region. This is normally solved by distributing the signal from a local processing center in the SFN region. An alternative procedure is to distribute a FAT multiplex signal to all sites and carry out exactly the same processing at all transmitter sites so that the output is SFN compatible. To achieve this, several alternatives aimed at producing identical DVB-T2 streams are available.

Bridging signals

Where there is an existing DTH satellite network, this can be used by the terrestrial operator as a cost-effective way to transport a multiplex to both satellite subscribers and DVB-T/T2 transmitters in one or more SFN regions.

A deterministic multiplexing solution converts the DVB-S2 satellite signal to an SFN-compatible DVB-T2 MI stream. The solution is described as “deterministic” because the way in which the final DVB-T or T2 multiplexes are rebuilt is pre-determined when the initial satellite multiplex is created at the headend/uplink site. It is also possible to implement the same deterministic process in an IP network feeding SFN transmitters.

At the headend/uplink site, a TS is preprocessed before it is scrambled and modulated into a DVB-S2 signal. The DVB-S2 output signal is fully compatible with S2 set-top boxes and is used for DTH purposes. At the transmitter site, the DVB-S2 signal is received, descrambled and processed to generate the DVB-T2 multiplexes. This processing consists of: splitting the satellite signal into DVB-T2 multiplexes, rate adaptation, synchronization and generation of the T2-MI stream. The correct Program Specific Information/Service Information (PSI/SI) tables are inserted by Packet Identifier (PID) remapping.

Deterministic remultiplexing

The deterministic remultiplexing solution feeds the DVB-T2 transmitter over C- and Ku-band satellite platforms. It is compatible with DVB-S and DVB-S2 satellite modulation. The deterministic remultiplexing solution is agnostic to video and audio compression technologies and requires the content to be transported in MPEG-2 TS in accordance with the ISO/IEC 13818-1 International Standard. Deterministic remultiplexing offers full freedom in the composition of the SFN signal.

The TS is preprocessed at the uplink headend by inserting special TS packets with absolute time information — so-called Time Marker Packets (TMP). The regional adapter receives the stream via satellite and extracts the relevant services and SI tables, and performs PID remapping and Program Clock Reference (PCR) re-stamping. TMP packets ensure that the TS packets are aligned in an identical manner in the T2-MI or DVB-T stream. The TMP packet is a normal TS packet containing a PCR value and absolute time stamp in UTC or International Atomic Time (TAI). The resulting remultiplexed transport stream will be identical for all transmitters in the SFN region. The timing packets do not alter the condition of the TS and do not influence the reception of the DTH signal on DVB-S set-top boxes.

The Master stream, distributed over satellite, is a normal TS without any encapsulation layer and is compatible with standard the DVB-S2 TVs and S2-STBs available in the market.

The core of the system is the Deterministic SFN Master, which works a multi-region virtual SFN adaptor. The signal may be distributed over satellite (or an IP cloud) and is compatible with standard DVB-S or S2-STBs in the market. The signal is received at the transmitter sites where the relevant TS packets are filtered and remapped, and PCR is restamped. This processing is assisted by special timer packets in such a way that the new stream can be rebuilt in the exact same manner at all transmitters within an SFN region. The distribution system may be as shown here with separate feeds for national and regional content or using the “FAT multiplex” principle.

A FAT multiplex contains all the national programs plus a number of services that carry local content. In addition, it includes a number of timer packets containing the ideal time and the offset at which the packet was sent. The local TS is built by extracting the relevant TS packets and dropping the others. The PSI/SI tables in the FAT pipe are no longer correct. Therefore, correct local SI must be generated by PID remapping. The updated local PSI/SI, which has been transported as unsignalled PIDs, now replaces the obsolete PSI/SI tables. This PID drop/remap process ensures that the order of the packets will always be the same at all transmitter sites in the same region. (See Figure 2.) However, this process will change the relative timing between packets, and the PCR values will be in error. On the basis of the precise time stamps in special timer packets, it is possible to recalculate the correct values for the PCR. This will ensure that the transport stream is SFN-compliant.

Regional content insertion

Today's digital terrestrial networks typically contain between three and seven multiplexes, each carrying around a dozen television programs and perhaps some additional radio services. A unique characteristic of terrestrial networks is the large number of regional content variations. In some cases, up to 50 percent of TV services include some local content. Regional services may range from full-time local channels to local news and local advertising. Today many of these channels are only transmitted in analog format. With the Analog Switch Off (ASO) coming up in the EU at the end of 2012, all these local channels need to convert to digital in order to continue broadcasting.

The regional content insertion issue is solved either through a centralized solution or by regional multiplexing. With a centralized approach, the national programs and all regional programs and matching SI are sent to all of the regional sites where the appropriate local programming and corresponding data is extracted. This is commonly known as the FAT multiplex approach as the bit rate ranges from 70Mb/s to 100Mb/s. The benefit of this solution is that complex processing such as conditional access and statistical multiplexing is performed at the national headend, while the processing at the regional sites is limited to service drop and regeneration of PSI and SI tables.

As mentioned earlier, SFN remultiplexing poses additional challenges in the area of regional content insertion. In DVB-T2 networks, there are two methods available to address regional content insertion that ensures the delivery of identical T2 MI signals at the transmitters. One is to utilize deterministic remultiplexing processes, and the other is to make use of T2 Physical Layer Pipe technology.

Deterministic processing may be used to merge content coming from different sites such as central and regional headends. Deterministic, remultiplexing processes will support insertion of local content, which then may be transported either in-band over satellite, out-of-band over other satellites or over local IP/fiber distribution networks.

The simplest approach is, in principle, the same as normal add/drop multiplexing that has been performed since the introduction of MPEG-2 digital broadcasting in 1995. However, the service filtering from a FAT multiplex has been enhanced by using logical and mathematical operations to construct the new transport stream in such a way that the PCRs are correctly calculated, and all transmitters in the SFN region are fed with identical T2 MI signals. Local content may be included in the FAT DVB-S2 multiplex or distributed to the regional transmitters over a regional IP network. In the case of regional insertion of local content, the regional TS must also include the TMP. This ensures correct packet order and PCR value. Following steps ensures correct SFN operation.

Next month, part 2 describes how layer pipes can be used to add regional content.

Dr. Ing Helge Stephansen is CTO at T-VIPS. B. Tommy Jensen, is senior engineer at T-VIPS.