In my work, I'm often asked to do TV coverage studies based on Longley-Rice. With the wide availability of software for Longley-Rice coverage studies, including the free software I described last month, this would seem like a simple task. In reality, it isn't because the FCC database used for FCC Bulletin OET-69 interference studies does not include TV stations' real antenna patterns. Use of the default OET-69 elevation pattern can lead to large errors, especially for LPTV stations. The FCC has recognized the limitations in Longley-Rice and is addressing it as it related to LPTV stations in a recent Notice of Proposed Rule Making. This month I'll describe the limitations of the OET-69 implementation of Longley-Rice and offer two ideas on how antenna elevation pattern data could be depicted in the FCC database, making it possible to do more accurate Longley-Rice coverage and interference studies.
How does this affect interference analysis? FCC OET Bulletin 69 uses the desired-to-undesired signal ratio to determine whether the population in a cell is considered to receive service or not. If the signal levels at greater distances from the transmitter site are calculated to be higher than they actually will be when extra beam tilt is used, other stations will be allowed to put more signal into these cells. With the wrong ratio, an OET-69 study could show no interference when interference actually would result. Looking at it from the other side, studies will show the station with additional beam tilt would cause more interference at this greater distance than it actually will.
If you have looked at UHF transmitting antenna patterns, you will see stations at high elevations use additional null fill to provide stronger signals close to the transmitter site. The differences in null fill are not accounted for in OET-69. DTV stations starting out with low power or with limits on power due to their allocation may opt for smaller, lower gain antennas. These antennas may have a down tilt of 0.75 degrees, but will have a much wider elevation pattern, resulting in stronger signal than those predicted on either side of the main elevation beam.
How bad is the problem for full power DTV stations? Based on the Aug. 29, 2003 FCC CDBS, out of 352 licensed UHF DTV stations, 114 had electrical beam tilt of more than 0.75 degrees, but only 26 used beam tilt of more than 1 degree. Using a standard elevation pattern greatly simplifies interference studies. Many interference studies, especially in congested areas, can involve calculating field strength for fifty or more stations. If a custom elevation pattern was used for each of these, both data entry time and, depending on resolution, calculation time could be excessive. Given the small percentage of stations with large amounts of electrical beam tilt, use of a standard pattern like Table 8 may be acceptable.
LPTV stations, however, are likely to use a much wider range of antenna elevation patterns. Antennas can range from Yagis and 4 dipole panels to 32 element slot antennas. Applying the elevation pattern from Table 8 to these stations is likely to lead to significant errors when calculating interference. NEW LONGLEY-RICE INTERFERENCE MODEL FOR LPTV
The standard method for calculating interference to and from analog LPTV stations uses FCC contours, although interference waivers can be requested based on a Longley-Rice showing. In the DTV LPTV NPRM, the FCC considers adopting a contour approach for DTV LPTV, but states, "The DTV methods provide more comprehensive, accurate and realistic analyses than the contour protection method currently used for the LPTV service. Given these advantages and the DTV model's wide-spread use, we are inclined to prefer the DTV methodology over the contour protection method as the basis for accepting digital LPTV and TV translator applications."
This, of course, raises the question of what antenna pattern to use. The FCC NPRM summarizes it this way: Typically, LPTV and TV translator stations use transmitting antennas with less gain and more beam tilt because such antennas are less expensive, smaller and lighter, and transmit a larger proportion of the stations' limited power downward toward the close-in locations these stations want to serve. These antennas generally have broader vertical radiation attenuation characteristics than the values given in Table 8 (i.e., numerically larger relative field strengths for the corresponding vertical angles, particularly for UHF antennas). Further, TV translator stations are typically sited at high elevations (hills or mountain slopes) and commonly employ electrical antenna beam tilt or combinations of mechanical and electrical tilt to maximize their signal down into the served communities."
The NPRM said one way to account for this variability would be to incorporate antenna beam tilt into an LPTV implementation of OET Bulletin 69. This would solve the beam-tilt problem, but would not catch the differences in elevation pattern beamwidth. A SIMPLE, ACCURATE DESCRIPTION OF ELEVATION PATTERNS
It seems to me the best approach would be to incorporate more accurate antenna elevation pattern data into the FCC database. I see two ways to accomplish this-one easy and one hard.
Antenna elevation gain and elevation pattern beamwidth are directly related. With two numbers -- antenna elevation gain and beam tilt -- the elevation pattern of the transmitting antenna could be characterized much more accurately than with the OET-69 Table 8 values. This approach does not describe side lobe response, but neither does Table 8. Null fill is not included directly, but since null fill reduces elevation gain it would fatten the pattern. With this data, it would be easy to include the effect of mechanical beam tilt after the untilted elevation pattern was calculated. Elevation gain is readily available for most antennas and should be easy to add to the FCC antenna database. In some cases, it can be assumed from the antenna model number.
My second alternative is a bit more complex and would be more difficult to implement, but it would also be more accurate since it would catch the side lobes and null fill. I suggest a somewhat larger database for each antenna that would include the amplitude and elevation angle for each maxima and minima between 0 and 10 degrees (or 0.75 and 10 degrees to match Table 8) and the points half way between each maxima and minima. For some low-gain antennas, there would only be three points, although the number of data points would increase substantially with higher gain antennas and may become unwieldy at some point.CALCULATING FIELD STRENGTH FROM RECEIVED SIGNAL LEVEL
I'm often asked how field strength in microvolts per meter is related to received signal power or voltage from an antenna. I described how to calculate this in this column several years (available on my FTP site at ftp://ftp.transmitter.com/pub/RF_Columns/rf-9309.txt
I've taken the formulas and put them into a spreadsheet you can download from my Web site. I didn't have room to add this to my list of tools and software last month. Fig. 1 shows the layout of the spreadsheet. Here's a description of how to use it.
When converting field strength in microvolts per meters to a received power or voltage level, the frequency of the signal and the gain and impedance of the antenna have to be included in the calculation. Enter this data in cells B4 through B7. Note that I've also included space to enter line loss in B6.
To convert from field strength to power and voltage, enter the field strength in dB over 1 microvolt per meter in cell B11 and read the result in cell B17. If you know the received power in dB above 1 mw, such as you would read from a spectrum analyzer, enter that number in cell B22 and read the calculated field strength in B28. The last table can be used to convert dB above one millivolt, as often found on field strength meters, to field strength in dB above one microvolt per meter. I hope you find it useful! Note that if you don't change the antenna or frequency, the dB difference between the field strength and received power will remain constant. If you have a lot of data to convert for one channel and one antenna, work out the conversion factor once and simply add it to all the readings. Your comments are always welcome. Drop me a note at firstname.lastname@example.org