SFN TV Broadcasting in The United States?

Many readers might remember the enthusiastic support of some for the adoption of coded orthogonal frequency division multiplexing, also known as COFDM, as the DTV transmission mechanism in the United States.

One of the touted advantages of COFDM over the 8-VSB transmission system that has been implemented here was COFDM’s ability to deal with multipath conditions, where the receiver must sort out a number of signals that emanate from the same transmitter, but, due to reflections and circuitous paths, arrive at the receiver at different times.

We are well aware of the kinds of problems multipath reception, caused by terrain and tall buildings, create in analog TV—ghosts and multiple images in the picture and interference to the FM audio signal.

In the DTV world, such multiple signal arrivals at the receiver can confuse the demodulator, and early ATSC receivers really did not exhibit good performance with respect to such multipath conditions.

Time marches on, technology advances, and today’s ATSC receivers deal with such conditions quite capably. So capably, in fact, that we are contemplating the possibility of implementing single frequency network (SFN) television broadcasting in the United States.

(click thumbnail)Richland Towers prepares to install an SFN antenna in Times Square in New York City for a recent DTx test.In TV broadcasting as we know it today, a single, high-power transmitter, its antenna as far above the ground as possible, serves a given metropolitan area—a high-power transmitter driving an antenna atop the Empire State Building in New York City, for example.

The SFN model employs several low-power transmitters driving antennas that are possibly at lower heights than the single-transmitter scheme. In this way, each low-power transmitter serves a specific segment of the metropolitan area; a favorite analogy is to the way cell phone networks are structured, although like many analogies this one is only partially accurate.


Apologies for burying the lead in the middle of the story, but there was an announcement in late August that tower operator Richland Towers, TV station group Ion Media, and other broadcasters have completed a test of an SFN—which is also known as a distributed transmission system—in New York. This sort of transmission system relies on the receiver to sort out multiple, identical signals arriving at different times. Let’s take a quick look at how this works.

In an SFN, two or more transmitters radiate the same signal on the same frequency, in the same geographical area. The receiver is thus likely to simultaneously receive signals from many or all of these transmitters, along with signals that are true reflections or echoes from surrounding terrain or buildings, and it must extract the DTV data from this multiplicity of signals.

The two principal requirements for such a distributed transmission system to work are: first, all the transmitters in the system must transmit identical symbols for the same data inputs; second, all the transmitters must be operating on exactly the same frequency.

In order for the receiver to treat the signals generated by the transmitters of an SFN as echoes of one another, all the transmitters must emit the same output symbols for the same data inputs; in other words, the signals must be identical.

There are two ways to ensure this. Some of the transmitters may repeat the off-air signals of a “lead” transmitter, thereby guaranteeing that all transmitters are emitting identical symbols, albeit with some additional delays. Alternatively, all transmitters in the system may be fed in parallel, with mechanisms implemented to insure that all transmitters are emitting identical symbols.

The second approach is standardized in ATSC A/110A, the Synchronization Standard for Distributed Transmitters (Rev A). It is important for all the transmitters to operate on the same frequency, as any frequency differences between signals will cause their echoes to appear to the receivers to have Doppler shifts, as if they were emanating from moving sources. These Doppler effects stress the receiver’s adaptive equalizer.

Another important factor in distributed transmission is delay spread—the total time elapsed between the earliest-arriving and latest-arriving signals.

When the desired to undesired signal ratio between transmitters falls below some particular value, the delay spread significantly affects the ability of the receiver to recover the desired data. The delay over the service area may be minimized by controlling the timing of the emission of the signals from the system’s various transmitters.

To get back to the real world of TV broadcasting, the entire distributed transmission system for New York City, as conceived by the testers, consists of a “high-powered hub site” in West Orange, N.J., about 11.5 air miles west of Times Square in mid-Manhattan, and several low-powered distributed transmitters around the metropolitan area.

The recent test involved only the New Jersey site plus a transmitter located atop a tall building in Times Square. The testers reported that, on the basis of comparison of the distributed system signals with existing single-transmitter broadcast signals emanating from the Empire State Building, the results for the distributed system were, “at least equal” to those using the single-transmitter approach.

The claimed advantages of the distributed transmission approach include lower costs for towers and transmitters, and the possibility of better coverage of segments of the metropolitan area. This is indeed news for broadcasters as it could drive a change in the fundamental way in which broadcasters emit their signals in the metropolitan areas of the United States. It bears watching.

Randy Hoffner