[opendtv] Re: Spectrum is too valuable

  • From: Craig Birkmaier <brewmastercraig@xxxxxxxxxx>
  • To: opendtv@xxxxxxxxxxxxx
  • Date: Thu, 19 Nov 2015 07:23:16 -0500

Sorry, hit send before finishing...

Regards
Craig

On Nov 19, 2015, at 6:59 AM, Craig Birkmaier <brewmastercraig@xxxxxxxxxx>
wrote:

On Nov 18, 2015, at 8:00 PM, Manfredi, Albert E
<albert.e.manfredi@xxxxxxxxxx> wrote:


Whoa, Craig, you're *so seriously* uninformed. You're talking nonsense. Why
do you think they are called passive optical networks, Craig? Answer: they
are passively split, shared among multiple households. Read up what a PON is:

No need. I know what a PON is. I know how to read. And I am seeing
commercials from Cox right now promoting Gigablast.

They ARE deploying 1 Gbps to homes now via fiber, and have announced plans to
offer this speed via HFC with DOCSIS 3.1.

Far fewer? Do the math, Craig. We already determined that with DOCSIS 3.1,
in the old 900 MHz plant, you can offer about 5 Mb/s aggregate downstream
capacity in the neighborhood (don't forget that some MHz is taken up by
upstream traffic), but ONLY if you eliminate all broadcast traffic. So tell
us, Craig. **If you repurpose all of the broadcast spectrum**, how many
homes can be passed with a single DOCSIS 3.1 PON, in a 900 MHz cable plant,
if each home wants 1 Gb/s? It's a really simple calculation.

Do you remember a post from 11/7 with the subject Gigablast?

I asked how blasting files at 1 Gbps impacts utilization stats within a PON.

Nobody answered.

You are looking at this problem from the wrong perspective. You are trying to
design the system assuming sustained bit rates for every home. That is, the
bandwidth of the PON divided by the number of homes. But that's not how we
use broadband.

If you have a max bit rate to a home of 5 Mbps, it takes 133 minutes to
download a 5 GB file, as you might need for a HD movie. If you have 1 Gbps to
the home it takes 40 seconds.

This is not idle conjecture. I used to download movies from iTunes way back
in 2005. It took several hours on my DSL broadband. The broadband was not
fast enough to support real time streaming of high quality video.

So the dynamics of how TV programming is delivered with gigabit service to
the home are very different than for "slow" broadband. It is not the
instantaneous max bit rate that a PON can support, but the throughput of the
PON that matters. Things may get tight during peak hours, but most programs
will be downloaded quickly, making the PON available to other users.

Only in the worst case, where everyone served by the PON is watching the same
live program does the max bit rate matter; and in that case, an IP multicast
can support everyone watching the program with one stream.

So when I said that there would be fewer homes per PON, I did not mean that
there would only be 10 homes per PON. And remember, they are not turning off
the MPEG streams, which is what many homes will be watching. The cable
company just needs to have enough bandwidth dedicated to DOCSIS 3.1 to
support the broadband demand, which will become relatively short bursts for
on demand TV programming.

First, almost anyone who wants "Internet-delivered TV" can get it today
from the cable/FIOS systems. 80% of U.S. homes have access to 25/3
broadband service today.

A refreshing change of tune, from someone who was claiming we are decades
away from being able to do this. So tell us, Craig, if you have a 2.4 Gb/s
GPON feeding the neighborhood, how many homes can be served, if each one
wants 25 Mb/s downstream>

No change Bert. Just reality.

The claim was that it may be a decade or more before everything is delivered
OTT. The live linear streams are still doing the heavy lifting, with a growing
percentage of TV viewing slowly moving to OTT, mostly for on demand, not live
programming. That is still true.

The reality is that this is a long term transition, not the flash switchover
you are advocating. The cabled providers have the ability to reallocate
spectrum resources as needed, and to build new PONs as needed.

Simply reclaiming the analog spectrum will make a huge difference, making room
for two 196 MHz DOCSIS 3.1 channels, perhaps more depending on the upper band
pas for the analog service. And if the number of live linear channels is
reduced, as I have been claiming will happen, those channels can be repurposed
for DOCSIS 3.1 as well. Assuming 4-5 programs per 6 MHz channel, you could
recover more than 100 MHz by eliminating 100 channels.

So yes, it is amazing, if you look at the big picture, rather than getting hung
up on theoretical number games.

Regards
Craig




Amazing.

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.

Other related posts:

  1. Home
  2. opendtv
  3. Archive
  4. 11-2015