[opendtv] Re: Spectrum is too valuable

From: Craig Birkmaier <brewmastercraig@xxxxxxxxxx>
To: opendtv@xxxxxxxxxxxxx
Date: Fri, 13 Nov 2015 07:13:58 -0500

On Nov 12, 2015, at 9:46 PM, Manfredi, Albert E <albert.e.manfredi@xxxxxxxxxx>
wrote:

Craig Birkmaier wrote:

As I pointed out, these numbers are on the low side,

No, Craig. You posted that link and those numbers to support your view, but
unfortunately you had not read the numbers you posted. That's why I said,
focus on the numbers that matter. In any of these lossy compression
algorithms, you can use more bit rate and get even better quality. But what I
stated is what your "proof" also states.

I read the numbers, said that they were on the low side, and pointed out that
they did not mention frame rates.

That's why you have to repurpose the cable broadcast spectrum.

All in good time. As you saw in the paper than Ron provided the link for, there
are many potential migration strategies. They will repurpose spectrum as needed
to meet demand for a range of services.

Then you can have it both ways. That's what I've been trying to get across.
Did you miss that point all along? Look at the subject line: "Spectrum is too
valuable." That's the whole point.

This started talking about the broadcast spectrum Bert. The cable guys have
many options that are not available to broadcasters.

Did you read it, Craig?

Yes, I found it and posted it, although I went back and could not find the
stuff you pulled out.

Ron posted some viewgraphs. Look toward the end. If you do not repurpose the
cable broadcast spectrum, then you will not achieve that 10 Gb/s capacity.
You'll get something less.

Of course Bert. That's why I did the numbers for you. I ran numbers for a 1 GHz
HFC plant assuming ideal conditions...your mileage may vary.

Here is the quote from the standard, Section 5.1:

"DOCSIS 3.1 ultimate service goal of multi-gigabit per second in the
downstream direction and gigabit per second in the upstream direction
resulted in significant changes in the PHY layer approach compared to earlier
DOCSIS versions in addition to changes on the cable network assumptions.
DOCSIS 3.1 focuses on the eventual use of the entire spectrum resources
available in cable environment by the CMTS and CM and on scalable cost
effective techniques to achieve full spectrum use."

So?

The reality, as the slides Ron linked to is that DOCSIS 3.1 will be phased in,
just as the other DOCSIS Standards have been. And I pointed out that there is a
big difference between theoretical and reality.

DOCSIS 3.1 offers many options to tune the profiles used to the ACTUAL
conditions in the neighborhood being served. Did you look at the slide
"Downstream Profiles?" It addresses the fact that SNR is not uniform across the
coax links fed by the PON. It suggests that multiple profiles can be used in a
PON to better tune the system and optimize throughput for the real world
conditions.

The slides also talk about extensions to 1.218 GHz and eventually 1.788 GHz.
But they do not say much about the implications of doing this. The 1.218
upgrade may be possible with existing cabling; the 1.788 will likely require
new cabling, not unlike what the industry went through upgrading from 400 MHz
(or less) to 900 MHz and 1 GHz. Again this is based in the reality of the
physical conditions, and the need to limit RF radiation from the system.

So bottom line, your notion of reclaiming spectrum now used to provide money
making services, just so a cable system can switch immediately to 100%
broadband is FAR from reality.

Read the standard. Why do you think you can wing it and get away with it,
Craig? You never have in the past.

I was asking for the specific part of the document from which you found this.

But don't bother, as the information Ron provided was far more useful.

But you didn't consider DOCSIS 3.1, which doubles the spectrum.

I gave you the numbers for the plants that exist today. Physical cabling
upgrades are very expensive. I really don't give a rat's ass about the
theoretical. What matters is how a standard can be implemented in the real
world, and the migration strategies that the industry will use.

You could just overbuild with fiber to the curb, which some systems are doing,
and get even better performance. But there are economic realities that must be
considered.

In case you didn't get this earlier, the spectral efficiency of DCOSIS 3.1 is
NOT significantly greater than that of DOCSIS 3.0. It is marginally greater,
mostly by saving on the 6 MHz channel guardbands. The real difference comes
from the fact that DOCSIS 3.1 can use a whole lot more spectrum.

That's crap. The numbers I provided show a huge gain in existing spectrum.
A theoretical limit of 5.697 Gbps for DOCSIS 3.0 vs 7.379 Gbps for DOCSIS 3.1.

You may claim that that is not significantly greater, but it's more than enough
to cause the industry to start deployment. The reality of moving to a 1.788 GHz
infrastructure is that it will be significantly more expensive.

So the answer is, with DOCSIS 3.1, if all the broadcast spectrum goes to
DOCSIS, a single PON can support 2000 homes with 5 Mb/s service, or 667 homes
with 15 Mb/s service (enough for three simultaneous HD streams).

How does that compare with your figures of between 25 and 2000 homes per PON
today, with an average of 500? Sounds to me like you can have it both ways.
But only if you repurpose that "too valuable" spectrum.

I already ran the numbers. The marketplace is moving to 25 Mbps as the base; I
already get 50 Mbps, and Google is providing gigabit. So stop trying to
reinvent reality.

The industry will migrate at the pace the market dictates.

Regards
Craig

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.

----------------------------------------------------------------------
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.

References:
- [opendtv] Re: Spectrum is too valuable
  - From: Manfredi, Albert E

[opendtv] Re: Spectrum is too valuable

Other related posts: