Friday, February 22, 2013

Review of Revised OET Bulletin No. 69, "Longley-Rice Methodology for Evaluating TV Coverage and Interference" - The Repacking Formula

There are eight issues targeted in the Commission’s review of the OET Bulletin No. 69 (Longley-Rice) software.  They are as follows:

1.      Population (census) data.  The DTV allotment channels assigned by the Commission (circa 1998) were based on 1990 Census data.  A number of years after that, they began to incorporate the 2000 Census data in coverage and interference studies.  They now propose to migrate to the 2010 Census data.  I think this is a good thing, since the 2010 Census represents reality more so than use of older census data.

2.       Terrain data.  Currently, most Longley-Rice software (including the Commission’s) uses 3-arcsecond terrain data.  The Commission now wants to use 1-arcsecond terrain data in order to provide more accurate results of coverage and interference.  This is also a good idea, in my opinion, because it will yield a result which closer resembles actual field strength in a given area.

3.       Correction of inaccurate engineering data in records of some stations.  This absolutely needs to be fixed.  I suggest that the Commission request a review by every station of its engineering record and submit any and all corrections.  I can’t tell you how many times I have seen obviously miss-rotated antenna data in CP records (and then carried through into licenses) and use of ERP in dBk instead of kW in license applications (flip-flopping the values in the 302 Form).

4.       Use of actual station elevation patterns in calculations.  Right now, the software assumes a standard elevation pattern for all stations’ antennas.  Since elevation patterns can be quite diverse, use of actual elevation patterns should yield a better depiction of actual signal level, at least on a theoretical basis.  I support the effort, but with reservation.  Based on results of field studies I have seen and participated in, calculated signal level bears little resemblance to actual signal level at particular locations.  This is because RF doesn’t always behave as predicted.  When an antenna is side-mounted on a tower, significant scalloping of the pattern can occur, resulting in peaks and valleys in the amount of radiated signal in a particular direction.  In addition, attenuation effects from manmade structures and vegetation is not factored into the equation.  There are dozens of reasons why we tend not to find a good correlation between predicted signal strength and actual signal strength, except in very general ways.  I also want to mention that nulls and maxima in elevation patterns affect signal levels relatively close to the transmitter site, where receivers tend to have line-of-sight and relatively high signal strengths (even in the elevation pattern nulls).  Finally, many stations use mechanical beam tilt in addition to electrical beam tilt, which significantly changes the power radiated in certain directions and out to significant distances.  I think that if the FCC truly wants a more accurate picture of what is going on in the real world, they need to develop a way to incorporate the use of mechanical beam tilt into the calculations.

5.       Depression angle calculation.  The Commission accidently accepted a mistake in the original software whereby calculation of signal at a given receive site used a depression angle to the transmitting antenna based on the latter’s height above ground rather than the difference between the transmitter site and receive site in terms of height above mean sea level.  This can cause a serious anomaly when the calculations are derived for mountainous areas.  This should be fixed, in my opinion.

6.       Precision of geographic coordinates.  The software presently rounds transmitter and receiver locations to the nearest second.  They now propose to round to the nearest 0.001 second.  Since most antenna sites have not been surveyed to that precision, I don’t really think this is going to make a difference in the real world.  But, hey, go for it if you think it will make things better.

7.       Establishment of a calculation grid.  I don’t really understand what they are talking about here and I am concerned that this might be the way they can get away with repacking stations into tighter spectrum confines.  So, unless somebody can explain to me why this is a good thing for broadcasters, then I am opposed to it.

8.       Treatment of error codes.  Right now, when the software goes bonk along a certain azimuth under calculation (usually along azimuths where the terrain is going through rapid fluctuations), it just goes ahead and assumes that there is coverage by the station in all of the cells where this error code comes up.  The FCC is asking if they should continue to allow the program to do this.  In a perfect world, I would ask the FCC to fix the problem with the software, but in the absence of that, I think it is a conservative (and potentially beneficial) approach to keep things the same (i.e. assume coverage in the error cells).  I say keep it the same.

The NAB came out with its initial comments on these potential changes to the software.  While they are not opposed to the concept of updating the software, they are unhappy that it is being done now, just before the auction.  Many stations have begun conducting coverage and interference analyses on their stations in order to determine whether or not to participate in the auction or to make sure they are not getting screwed by the repack (in terms of interference-free coverage they now have versus what they given on their new channel).  The results of all those studies would be useless if the changes to OET 69 was implemented now.  The NAB is lobbying for the Commission to make the changes after the dust settles with the auction and the spectrum repack.

The concern I have is that there will not be enough time between now and the auction to adequately evaluate the differences that the new and old software have on station coverage and incoming interference.  And, I would be afraid that the changes the Commission makes to the software would disadvantage the broadcaster in some way that would only be discovered after it is too late.  If I could cherry-pick the items that I could support now in the changed software, they would be: use of 2010 Census data, use of 1-second terrain data, correction of station data in the FCC engineering database, correction of the depression angle problem, and use of actual antenna elevation patterns in the calculations.  Numbers 6, 7, and 8 above are not worth consideration at this time.  And I am very leery of the calculation grid thing.

If you have any questions regarding my comments, please feel free to share.