Expanding DTV coverage Part II

Single-frequency networks have been in development for years and only recently has the FCC approved their use with DTV. What is a single-frequency network and how can it improve reception, especially for mobile DTV

Single-frequency networks (SFN) have been thought of as a way to fill in where the main transmitter’s signal fails to reach, but when used with DTV, its uses can go beyond simple gap filling. A full-power analog TV station could rely on the local cable operator to carry its single program, but that same cable company does not have to carry all of your digital channels. It’s up to your over-the-air signal to deliver all your DTV channels to the viewers, whatever the terrain.

Mobile TV can bring in more viewers, with little screens, but only if they can reliably receive the signal. And because these viewers are mobile, you never know where they will be within your service area. These areas in the past may not have been seen as important, such as highways, shopping malls and even downtown at street level, but they are just the places where those new viewers may be.

Figure 1: Shadow areas and overlap caused by boosters  Click to enlarge

Using an SFN can also reduce the amount of interference in the areas surrounding your service area while increasing the station’s coverage within it. (See Figure 1.)

Whatever the reason a station might consider using a SFN, you need to know how one is put together and how it works. The FCC calls these systems distributed transmission systems (DTS), while the ATSC refers to them as SFNs, but they all share several things in common: two or more transmitters use the same frequency to cover a service area, the precise timing of the transport streams from each transmitter is carefully controlled and these transport streams are identical down to the last bit.

Standards and rules

It all starts with the ATSC and its A/110 “Synchronization Standard for Distributed Transmission,” which is also referred to as DTxN. This lays out the specification for a complete system to implement a SFN. A/110 follows the work of several companies that have carried out development and testing over the last several years on SFN. There also is a companion document called RP/111 that spells out how to design a SFN.

The FCC initially approved SFN operation under Part 73 on a case-by-case basis. Clarification of these rules was published in an Order and Notice of Proposed Rulemaking released by the FCC Nov. 4, 2005 (FCC 05-192). The FCC has now adopted the rules for SFN as announced in Public Notice DA 09-528 Feb. 27, 2009, and they have updated Forms 301 and 340 to accommodate the use of SFNs.

The FCC lays out its rules for using DTS in 73 FR 74047, but it does not require the use of the ATSC A/110 standard. It leaves open the question of how to synchronize the various transmitters to allow for future innovations in this area. Several large-scale tests have been carried out over the last several years with great success.

The reason for SFN

Terrain shielding is the main reason for using SFNs. That one big transmitter putting out anywhere from 50kW to 1000kW from a single location can have its signal blocked by hills, valleys and buildings. UHF television works on line of sight, and if there are any shadows in the coverage area, reception will be compromised. Broadcasters have tried to get around this by increasing power levels and hoping that reflections will fill in those shadows, but this does not always work. With analog television, people would put up bigger directional antennas, something not aimed at the transmitting antenna, but at a point of reflection. Now with most of the audience out of the habit of putting up rooftop antennas, and the easier it is to receive a DTV signal, the more likely you are to keep that viewer, especially for the DTV channels that the local cable company does not carry.

Those reflections mentioned above cause their own problems with 8-VSB. With analog, a reflection or ghost will cause a faintly shifted picture on the screen, whereas 8-VSB can prevent the reception of the signal with a resulting loss of picture. Being able to control the reflections that cause multipath disturbances is another reason for the use of SFN. In a total SFN installation, multipath issues can be controlled and minimized to a great extent, increasing a station’s coverage.

ATSC-M/H is another reason for SFN in difficult to reach areas that are now of prime importance. Just like FM radio’s prime audience is mobile, the same mobile audience will be increasingly important to mobile TV. And like FM radio, if a station’s reception is weak or non-existent in some areas, the viewer is quick to change channels to a station with better reception.

The receiver’s role

The development of the DTV receiver has played a big role in the evolution of SFN. The ability to pick out a particular signal and lock to it in the presence of multipath was required for an SFN to work at all. To have complete coverage, some of the signals from the various transmitters in an SFN will have to overlap, thereby causing multipath interferences. The early DTV receivers did not handle this situation very well, and this was a major reason for the early call to switch over to the COFDM — because it did a better job at handling multipath problems. But with today’s current-generation 8-VSB receivers, this is no longer the case; they can handle a much larger range. The real problem is in how close in time of arrival the two signals are and if the stronger one is advanced or delayed compared to the weaker one. Typically, early 8-VSB receivers incorporate an equalizer that could compensate for multipath disturbances within a certain range (e.g. –2.3uS to 20uS). Outside of this range, the receiver would fail to lock to either signal, resulting in a loss of picture. As long as the timing of the two identical transport streams can be controlled, the receiver can pick one out and decode it. (See Figure 2.)

Today’s DTV receivers can handle two identical signals that are of equal level with a timing displacement of anywhere from-12 to 50uS. With a level difference of 8dB, the timing can be –45 to 55uS.


Richard Schwartz of Axcera contributed to thistutorial.