Transmission Technology: Predicting HDTV Coverage
[ MULTIPLE DTV/NTSC ANTENNA SYSTEMS DESIGN ]     [ NTSC COVERAGE AND DTV SERVICE ANALYSES ]     [ INTERFERENCE ANALYSES ]     [ TRANSMITTER-TO-TRANSMITTER ISOLATION ]     [ NTSC GHOSTS FROM TALL TOWERS AND BUILDINGS ]     [ CHANNEL COMBINING AND FILTERING SYSTEM DESIGN ]     [ RF RADIATION HAZARD ANALYSIS ]     [ TECHNICAL PAPERS ]     [ PATENTS ]     [ ABOUT ]     [Back]

O. Bendov
Dielectric Communications, Inc.
Cherry Hill, NJ 08033

This article looks at the differences that exist in predicting coverage patterns for NTSC vs. HDTV systems. HDTV is used rather than ATV to signify the Grand Alliance's transmission format. For NTSC, the conventional predictor for coverage is the field strength in decibels above 1 microvolt/meter (dBu). The prediction of the 64dBu contour (UHF Grade B) is based on the computation of the available effective radiated power (peak of sync ERP) emitted by the antenna.
For NTSC, field strength is a measurable predictor, and the computation of the coverage contour based on the available ERP is valid, because multipath and other signal distortions over the 6MHz channel translates into degraded picture and not into loss of coverage.
In contrast, multipath co-channel interference and other distortions of the digital HDTV signal that can be observed on a spectrum analyzer translate into loss of coverage, not picture degradation. The loss of coverage due to signal distortion must, therefore, be accounted for as a penalty against the available power if a realistic coverage prediction is desired. Furthermore, the field strength of HDTV is not measurable and cannot serve as a service predictor. (See sidebar "Can field strength serve as a predictor of HDTV coverage?" at the end of this paper.)
HDTV signal distortion
The spectra of undistorted NTSC and HDTV channels are shown in Figure 1. The undistorted spectra are a good representation of the NTSC and HDTV signals delivered to the transmitting antenna. What may come out of the transmitting antenna and is finally delivered to the receiver is shown in Figure 2. It is apparent from these figures that NTSC is essentially a narrowband 3carrier transmission in which only the integrity of the area around the carriers is important for picture and sound quality.

Figure 1. The spectra of undistorted NTSC and HDTV channels.

Figure 2. The NTSC and HDTV spectra as they might appear at the receiver.

In contrast, HDTV is a wideband transmission in which the integrity of every portion of the channel is equally critical for coverage. In fact, the distorted HDTV signal, with a deep notch just below the color subcarrier of the overlaid NTSC spectrum, will provide zero coverage regardless of the power level at the receive antenna. The same distortion applied to the NTSC signal would hardly affect the picture and will have no effect on the sound.
Several sources may cause multipath distortion at the transmission facility. For example, it occurs when the transmitting antenna is near the support tower (side-mounted) or is part of an antenna farm. Figure 3 shows what happens to a perfectly omnidirectional Channel 38 antenna once it is installed next to (8-foot separation) a typical tower with a 10-foot face. The antenna pattern is distorted, and in each azimuthal angle, the distortion varies with frequency.
Coverage penalty due to distortion
Any distortion of the HDTV signal results in a loss of coverage. The loss of coverage is best expressed as a power penalty, in decibels, against the available power at the input to the transmitting antenna. The concept of the power penalty is unique to HDTV. It is central to the observation that the net useful power, rather than available power (used in NTSC), is the key to realistic coverage prediction. The computation of the penalty due to linear distortions, such as multipath, is detailed elsewhere. (See "A- New Approach to the Analysis of Adjacent Structure Effects on HDTV Antenna Performance," 1995 NAB Engineering Proceedings.)
The power penalty is a sum of two components: one component represents the actual loss of total power in 6MHz; the second component represents the effective loss of power due to the channel equalization process at the receiver. The equalizer attempts to reshape the distorted spectra shown in Figure 2 to that of the undistorted spectra shown in Figure 1 by increasing the gain at selected frequencies. As a result, the noise level increases and the carrier-to-noise ratio (CNR) decreases, which is effectively a loss of carrier power. It should be mentioned that, in HDTV, the carrier power is the average of total signal power over 6MHz, whereas in NTSC, the carrier power is the rms power at the carrier frequency during sync pulses.

Figure 3. The radiation pattern of an omnidirectional antenna (Channel 38) after it is installed next to (8-foot separation) a tower with a 10-foot face.

Figure 4. Power penalty that results from the pattern shown in Figure 3.

The total penalty due to the antenna patterns of Figure 3 is shown in Figure 4. In two directions, southeast and northwest, the multipath was substantial and the penalty was high. In those directions, coverage would be totally lost for HDTV, while the penalty for NTSC would be some loss of coverage and minor picture degradation.
HDTV contours
HDTV coverage contours cannot be defined by the equivalent NTSC contours (City, A and B grades) because HDTV does not degrade gracefully -- picture and sound are either available and perfect or absent altogether. This is known as the cliff effect, and it takes less than 1 dB loss of CNR to lose service at the fringe.
HDTV coverage contours can be defined and measured in terms of power or CNR with sufficient margin for the desired percentage of time availability. As an example, consider the antenna of Figure 3 with the penalty against the available power of Figure 4. The 15dB CNR (threshold HDTV) contours over a flat terrain are shown in Figure 5.

Figure 5. Antenna countours that would result when the installation
of Figure 3 is combined with the power penalty of Figure 4.

The two contours are shown for two different percentages of time availability. A comparison of figures 3 and 5 illustrates the significance of the penalty. For NTSC, coverage would be predicted everywhere, whereas for HDTV, coverage would not be available everywhere. Just what is the "right" percentage of time availability for HDTV service is still an open question. Suffice it to say that, because of the cliff effect, the percentage required for reliable HDTV service will be higher than the 90% used for UHF-NTSC. While HDTV field strength cannot be measured (and even if it could be measured, it could not serve as a reliable coverage predictor), the concept of "equivalent field strength" could be useful for contour planning. (See "The Effect of Channel Assignment on Transmitter and Receiver Requirements for Equivalent H DTV/NTSC Coverage," 1994 NAB Engineering Conference Proceedings.) The assumption underlying the concept of "equivalent field strength" is that t here are no linear or non-linear distortions by the transmitter/antenna system; then the penalties to spectrum distortion do not apply, and contour planning at a single frequency is valid. In most cases, this assumption is optimistic. Even the transmitter's in-band intermodulation products may contribute a penalty of 0.5dB against the available power. The concept of "equivalent field strength" is also useful because most of the available software is based on field strength.

Following are the equivalent field strengths that, together with the FCC (50, 50) curves, would yield the HDTV 15dB CNR contours with enough margin for the stated percentage of time availability, assuming a receiver noise figure of 10dB and an outdoor antenna with a gain of l0dB:

Channel 15
Channel 39
Channel 61
50-foot Downlead attenuation
4 dB
4.5 dB
5 dB
90% Availability
51 dBu
53 dBu
55 dBu
99% Availability
58 dBu
60 dBu
62 dBu
Conclusions
  • Conventional field-strength meters cannot be tuned to the carrier frequency of an HDTV signal.
  • Neither the field strength of the pilot nor the average power of the HDTV signal can serve as reliable coverage predictors because signal distortion, such as a notch in the spectrum, must be factored as a penalty against the available power.
  • Transmission equipment of two different vendors that would provide identical NTSC coverage contours may provide different HDTV coverage contours because the spectral distortion of HDTV is translated into loss of coverage. it should be remembered that, in HDTV, a ldB penalty is typically translated into a loss of one mile at the fringe.
  • Unless the FCC sets standards for HDTV receivers, the penalty due to the equalizer may vary from one receiver brand to another, thereby affecting actual coverage.

Sidebar
Can field strength serve as a predictor of HDTV coverage?

The Final Technical Report submitted to the FCC by ACATS states that "An objective measurement that should permit reliable prediction of satisfactory HDTV service at UHF is field strength." In fact:

  • HDTV field strength is not measurable.
  • The field strength of the NTSC signal was measured during the Grand Alliance field test in Charlotte using the Potomac Instruments' FIM-72 field-strength meter. This measurement failed when applied to the HDTV signal.
  • The conversion process, from the measured total signal power in 6MHz to incident field strength at the receive antenna as used in the field test report (not documented in the report), is not applicable to HDTV.

Unlike NTSC, the 6MHz spectrum of HDTV shows no distinct carrier, only a signal acquisition pilot. RF field strength is measurable by a narrowband tunable receiver at a single frequency provided the field stays constant during measurement. The field strength of the HDTV signal is random and fast-changing because it depends on the picture content, and the data rate is high.

How then, did the field test report come up with the field-strength data even though it could not be measured? It began with the averaged measured total signal power in 6MHz. That power is represented by the area under the power spectrum curve as seen on the spectrum analyzer. Next, the conventional NTSC formula that relates the measurable received power to the incident field strength was invoked:

To determine the field strength for HDTV using the conventional formula, a wrong assumption had to be made -- that all of the HDTV power is concentrated at a single frequency (carrier was chosen) rather than being spread, generally not uniformly, across the channel. In any case, the value obtained is a derived quantity, not a measurable field strength of the HDTV carrier. The derived field strength using the conventional formula can play a role in service planning as explained in the HDTV contours section on p. 290.

Measurement of the HDTV pilot carrier's field strength might be possible, but the pilot's field strength alone cannot be used as a coverage predictor because the shape of the spectrum must also be accounted for in the prediction of realistic coverage.

Acknowledgment: Figures l and 2 of the NTSC and HDTV spectra courtesy of C. Scarpa, Hitachi America.

Back