The ATSC Distributed Transmission System and Applications to Translator Service

Distributed transmission is single frequency network technology applied to the ATSC system for digital television. Rather than using a single transmitter to service a coverage area, multiple transmitters are used. The transmitters are synchronized in frequency and symbol emission. Timing adjustments allow optimization of the system to produce minimum timing skew in areas where the multiple signals overlap. This paper provides a detailed introduction to the technology and hardware developed for distributed transmission. The first station to implement distributed transmission is WPSX-DT in State College, PA. For WPSX, distributed transmission was the only feasible way to provide UHF coverage, mainly because of terrain shielding. This paper presents implementation experiences and data from this real-world installation. Distributed transmission can also be applied to translator systems, by creating distributed translator networks. In a distributed translator network, the translators may all operate on the same channel. This results in greater spectral efficiency. Multiple hop distributed translator systems can be accommodated with a minor change to the ATSC CS/110 candidate standard. The changes necessary to implement distributed translator systems are also described. What It Is Distributed Transmission is the use of multiple transmitters to service your coverage area rather than the traditional single transmitter system. What It Is Not Distributed Transmission is not a booster (repeater) system. Distributed Transmission systems cannot normally receive a signal off-air and retransmit it, because this would violate causality. Distributed transmission systems radiate the same symbol sequence at very nearly the same time from each transmitter. The relatively long time delays associated with demodulation, deinterleaving, and error correction, followed by forward error correction, interleaving, and remodulation, would mean that a distributed transmission slave would need the input data long before it could receive it over the air. On channel repeaters are another beast entirely. On channel repeaters receive an off-air signal and retransmit it within about a microsecond. Most of the delay in an on channel booster is from the IF bandpass filter (usually a SAW filter). Because the delay of an on channel booster is intended to be short, it is not possible to regenerate the data and remove errors. Received signal distortions (including the short echo produced by the booster itself) are cumulative and are simply retransmitted. A distributed transmission system transmits a pristine signal from all of its transmitters. And, each transmitter can be delayed or advanced in time with respect to the other transmitters in the network. On channel repeaters can only be delayed. Why It Is Needed Some DTV stations are “terrain-challenged.” Most DTV allocations are on UHF channels, where terrain shielding is more of a problem than it is on VHF. In some cases, it is impossible to replicate analog service using a single transmitter. For these cases, distributed transmission is needed. In other cases, such as the inability to put up a single tall tower, distributed transmission may be desirable. Putting up a distributed transmission system is more complicated than just feeding the same SMPTE 310 signal to multiple transmitters. In fact, two exciters from the same manufacturer, fed with the same SMPTE 310 input signal, will almost always generate different symbol sequences. In other words, they will become mutual jammers. DTV modulators are generally nondeterministic. A DTV modulator makes an arbitrary decision on where to insert frame sync. Frame sync occurs once every 624 MPEG packets. So the odds of getting two modulators to insert frame sync in the same place is 1 in 624. DTV modulators also include arbitrary initial states in 24 trellis coder bits and 12 precoder bits. (To simplify the terminology in this paper, we will collectively refer to these as “trellis coder” states, data, or bits, even though precoders are also included.) So, this makes a total of 36 arbitrary bits. So the odds of having two exciters fed with the same bit stream producing the same symbols by chance is 1 in 624*2 = 1 in 42,880,953,483,264, or about one in 43 trillion. Starting up an ATSC modulator once a second would result in chance synchronization about once every 1.4 million years. The odds get far worse when we consider having three or four modulators randomly synchronize. Another way to say this is to point out that given a particular SMPTE 310 input stream, there are 42,880,953,483,264 different symbol sequences that can represent the same signal. So to create a distributed transmission system, there must be a means of synchronizing the otherwise arbitrary initial conditions in each modulator, and maintaining synchronism over time. Pilot frequencies also need to be locked, and timing must be adjustable and stable.