Digital TV: Charles W. Rhodes

IEEE Symposium Highlights

By all accounts, the IEEE Fall Broadcast Technology Symposium was a success. Papers relating to DTV were presented, including an important submission by Merrill Weiss on distributed transmission of DTV signals (DTx). I couldn't attend his PowerPoint presentation, so he kindly provided me with a copy of it. Thanks, Merrill.

Dr. Yiyan Wu of the Canadian Research Centre (CRC) opened a session with another paper on DTx. He detailed the three possibilities and their pros and cons in different situations.

DTx OPTIONS

The first mode is a Single Frequency Network of transmitters fed via an STL link. All transmitters are radiating precisely the same data pattern in the same channel synchronously; i.e., their pilot carriers are frequency-locked. There is no main transmitter.

There are constraints on the relative arrival times of these data signals at points where more than one transmitter's signal arrives with significant field strength. This is because those signals arriving outside the receiver's adaptive channel equalizer are seen as noise, not echoes, which the equalizer can filter out. Such noise reduces the signal-to-noise ratio and the service area of the transmitter.

The second mode involves the use of two channels. Transmitters that would cause interference with each other if broadcasting on the same channel would use different channels. The FCC would have to authorize not only the use of multiple transmitting sites, but also multiple channels.

The third mode described is the Digital On-Channel Repeater (DOCR). This mode uses the signal radiated by the main transmitter to feed the on-channel repeaters, which serve terrain-shielded areas. This was first demonstrated in 1998 by the Advanced Television Technical Center Inc. and described in papers presented at the 1998 and 1999 NAB Engineering Conferences.

Undesired feedback of the repeater's re-transmitted signal into the repeater's receiving antenna can lead to howling, just as with feedback from loudspeakers into the microphone of a PA system.

The forward gain of the repeater must be less than the feedback-path loss, or else it will howl. This may limit the power output and thus the coverage area of some repeaters. Each such repeater must be custom-engineered to minimize this feedback.

One problem for all forms of DTx is adjacent channel interference (ACI). At sites where a weak desired signal is being received, the re-radiated undesired adjacent channel signal may be strong enough to cause ACI.

There will be situations in which DTx will serve to improve coverage, but this problem of ACI must be considered in the early stages of planning a DTx system.

A/V CODING

Bill Miller of ABC-TV presented a progress report on the work of the ATSC Specialist Group on audio and video coding.

The session was not limited to ATSC. Europe has DVB-S ("S" for satellite-to-the-home), its near cousin DVB-C ("C" for cable-to-the-home) and DVB-T ("T" for terrestrial broadcasting-to-the-home). DVB-S was the parent system; DVB-C and DVB-T, both minor variants.

In Japan, an entirely different system, ISDB-T, was developed after ATSC and DVB-S. ISDB-T offers many novel features, including better immunity to man-made (impulse) noise than earlier systems. Impulse noise is an issue for low-VHF broadcasters, concerned about the suitability of low VHF for DTV broadcasting.

Recent tests of the ISDB-T system in Brazil were reported at the IEEE Symposium, and from the impressive results, it would not surprise me if Brazil were to adopt it.

Another paper from the CRC proposed ways to provide a "digital-return channel" for iTV and Internet access, especially from the remote areas of Canada.

Dr. Oded Bendov's paper, "Interference to DTTV Reception by First Adjacent Channels," provided a thorough analysis of noise in the desired channel from every possible source including:

• Noise within the channel radiated by the desired transmitter.
• Noise radiated by one or more undesired transmitters, on first adjacent channels (sideband splatter) into the desired channel.
• Sky noise, a consideration in the low-VHF channels.
• Receiver-generated noise due to third-order nonlinearity in the tuner, or tuner overloading.

Dr. Bendov showed calculations where receiver-generated noise due to tuner overload may be enough to block reception, even without any adjacent-channel signals. This is possible when extremely strong signals overload the tuner, generating cross-modulation (X-M). This is well-known in mobile radio and other wireless applications.

I also presented my paper, "Interference between Television Signals Due to Intermodulation in Receiver Front-Ends." This paper covers much of the same ground as Dr. Bendov's, as well as some of the material first presented in this column. Possibly the most significant point is interference generated by pairs of signals on certain UHF-taboo channels.

This was something never explored by the ATTC in 1995. The worst case may be the n-2, n-1 pair and the n+1, n+2 pair. Intermodulation (IM) products generated by strong signals on these channel pairs add to the receiver-generated IM from the two undesired signals and sideband splatter from the adjacent-channel undesired signal. I posed technical questions concerning:

• Insufficient data on planning factors for repacking the remaining broadcast spectrum;
• Obtaining such vital information;
• Locating a suitable laboratory;
• Obtaining funding, and;
• Supervision of laboratory testing.

We have recently heard much about the performance of DTV receivers. An excellent presentation by LG Electronics of Korea, Zenith, (a division of LG) and Sinclair Broadcast Group identified where these improvements were made and how field-testing has shown them to work. Improvements focused on the performance of the signal decoder with (strong) multipath interference.

The time window in which the adaptive channel equalizer works has been greatly increased for both leading and trailing echoes. Not only has the time been extended forward and backward, the ability of the receiver to synchronize to the ATSC signal under adverse conditions has also been improved.

However, I did not hear of any improvement in tuner performance. The need for a wider linear dynamic range in which the desired signal is not distorted by third-order IM and/or X-M may be what defines the next generation of DTV receivers, but I worry about this.

Why would better, more expensive tuners be developed for a minority of potential buyers not served by DBS or CATV? Is this not a case where the FCC must step in before it shuts down free over-the-air television? How can analog transmission be terminated when a significant portion of the population cannot reliably receive all local DTV channels? The other impediments to reception are now behind us (as I predicted in this column). I believe that existing technology can extend the linear-dynamic range of DTV-signal handling in tuners both downward (lower-noise figures) and upward (higher third-order intercept power) by commercially available means, but not quite for free.

Dr. Charles Einolf, and Victor Tawil of the Association for Maximum Service Television (MSTV) co-authored a paper on the future potential of low-VHF for terrestrial DTV. The impulse-noise levels in this band were documented and the problem won't go away. Impulse noise increases with increased electrical power consumption. While analog transmissions continue, it does not appear likely that higher ERP for low-band DTV transmissions will be permitted. It does seem clear that much higher ERP is necessary for successful DTV service in this band, but this may increase interference unacceptably between low-VHF signals.

Gary Sgrignoli presented the technical parameters of DTV translators as recently established by the FCC. Now it will be possible to engineer DTV translators to extend terrestrial DTV coverage where translators for analog transmission are serving millions of rural viewers. Gary had a lot to do with these technical choices and is enthusiastic they will serve as a basis for rational translator development in the next few years.

IEEE members who are also members of the Broadcast Television Society within IEEE will be receiving these papers over the next year. Those who attended already have at least the gist of what these papers provide, and have an advantage over their competitors who didn't attend. Perhaps with DTV looming large in your professional future, you cannot afford to miss the 55th IEEE Fall Broadcast Symposium next October in Washington, D.C. I was gratified to see that many attending the recent symposium are readers of this column and I hope to see more of my readers next October.

Charlie Rhodes is a consultant in the field of television broadcast technologies and planning. He can be reached via e-mail at charleswrhodes@worldnet.att.net