[opendtv] Re: Spectrum Analysis
- From: "Manfredi, Albert E" <albert.e.manfredi@xxxxxxxxxx>
- To: "opendtv@xxxxxxxxxxxxx" <opendtv@xxxxxxxxxxxxx>
- Date: Mon, 20 Jun 2011 20:33:04 -0500
Ron Economos posted:
> Options for broadcasting.
>
http://download.broadband.gov/plan/fcc-omnibus-broadband-initiative-%28obi%29-technical-paper-spectrum-analysis-options-for-broadband-spectrum.pdf
Page 6:
-------------------------
The propagation characteristics of the TV bands, especially
in UHF ranges between 470 MHz and 698 MHz, are
well-suited for wireless broadband applications. Unlike higher
frequency ranges, which comprise the majority of spectrum
licensed for mobile broadband use, the UHF frequency bands
provide excellent coverage over wider areas, as well as better
penetration into buildings and houses. These propagation
characteristics reduce the capital required for network buildout,
especially in less dense areas where cell sizes are largely
limited by propagation rather than by clutter, terrain, or capacity
needs. For example, a simple propagation analysis shows
that approximately one third as many cell sites are required to
cover the same rural area at 650 MHz as are required at 1900
MHz, assuming the same wireless technology is deployed at
both frequencies.
-------------------------
Simplistic notions sound more obtuse every time they are repeated.
Where there is most need for TV spectrum is also where there is most need for
efficient, high density two-way wireless broadband. High population densities.
In these environments, you DO NOT want the propagation qualities that (quoting)
"reduce the capital required for network buildout, ..." Translated, they mean
reduce the number of cell towers per unit area.
It is exactly the other way around.
In these dense population centers, the only way to achieve the required scaling
characteristics for two-way wireless is to LIMIT the range of each tower, and
to reduce the cell size. That's how you get a good multiplier effect (frequency
reuse). For a two-way system, if you try to save money by making cells larger,
you are simply wasting spectrum, and gaining very little extra capacity by
comparison. So in fact, you DO NOT want the propagation characteristics of low
UHF for two-way wireless, in urban and many suburban settings.
Conversely, in rural areas where the propagation characteristics of low UHF
really do help, because you really want large cells, there are a lot more TV
band frequencies available (white spaces).
So why do they insist on this single "propagation characteristics" goal, even
though it does not apply across the board? I think that a report like this one
can't start out with an incorrect premise. The rest of the report depends on
this. So as far as I'm concerned, it taints the rest of the report. For
instance, why not consider the 60 MHz bands for two-way wireless from, say,
light poles, in urban areas? That has more potential for capacity improvement
than the lower UHF bands. Easily.
And too, their trend lines for TV usage stop in 2008. Just enough to show cable
flattening, not enough to show any drop.
Page 15:
-------------------------
For one set of scenarios, we assumed six stations could combine
on one channel. Six-to-one channel sharing would allow each
broadcaster to maintain a primary OTA signal in SD. Although
the bandwidth required to broadcast an SD signal ranges from
1.5 to 6 Mbps, stations could use statistical multiplexing to
combine up to six signals on a given channel by taking advantage
of different "peaks" and "valleys" in bandwidth needs for
each signal.4
-------------------------
Does anyone at the FCC watch OTA TV at all? One of the great enhancements of
digital was multicasting. Any excuse for not even hinting at this at the
outset? Instead, they go on and on about 6:1 reclamation benefit. The only
concession they make is to mention HD channels, for heaven's sake. What about
those stations that are already multicasting five and more SD streams? Or HD
plus several SD? So, this scenario should not even have been introduced.
They do eventually start talking about multicast channels, but only to make
points about 720p vs 1080i efficiency. Not to undo their previous comments
about channel sharing. Parenthetically, Univision, which they claimed had no HD
programming, now has all HD programming. So these things are still changing.
Very interesting charts, p. 20-22. I have to admit, they did pair the correct
stations in most cases (except for the ABC-ION). One of the pair doesn't
multicast. But I don't see even here that they have taken the SD multicasts
into account. For example, they pair up NBC and Univision. But they showed
above that NBC uses a solid 18 Mb/s for its HD + 2 SD streams and mobile. How
do you add the new HD Univision to this?? Even without multicasts or mobile for
Univision? Did they subtract the NBC multicasts?
Fancy charts aside, if NBC used the most up to date MPEG-2 encoding possible,
according to their numbers, the NBC HD channel might often fit in 9 MHz avg.
("HD streams that currently average 11 Mbps would consume 9.4 Mbps with a 15%
improvement in efficiency enabled by more advanced encoders," they claim. And
similar improvements for lower bit rates demands.) Adding the two SD channels
to this means at least 4 more Mb/s. Mobile adds (according to their numbers) 2
Mb/s. That's a total of 15 Mb/s. There's no credible way to add another HD
stream to that, even if the other stream is H.264. So this pairing means
freezing in their tracks the nice innovations we have been seeing of late (e.g.
ION adding HD, Univision adding HD, CW50 adding a second movie channel, NBC4
converting SD postage stamp to SD anamorphic squeeze).
More of the so-called "cellular architecture" ideas, i.e. SFNs, which I very
much insist do nothing for the purpose of saving spectrum in a single-market
structure like we have here.
Bert
----------------------------------------------------------------------
You can UNSUBSCRIBE from the OpenDTV list in two ways:
- Using the UNSUBSCRIBE command in your user configuration settings at
FreeLists.org
- By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word
unsubscribe in the subject line.
Other related posts:
