nanog mailing list archives

Re: FCC Chair Rosenworcel Proposes to Investigate Impact of Data Caps


From: sronan () ronan-online com
Date: Tue, 20 Jun 2023 10:09:35 -0400

Or the investment to upgrade doesn’t make financial sense.

On Jun 20, 2023, at 9:54 AM, Mark Tinka <mark@tinka.africa> wrote:

 

On 6/20/23 15:20, Mike Hammett wrote:

Sometimes yes, sometimes no.

When you go down in density, your fixed cost per customer really escalates and you simply can't afford to provision 
as much as you'd like to. When you leave glass as a transport mechanism, scaling isn't easy. When you don't have a 
wireline to the customer prem, scaling isn't easy.

You might have a licensed backhaul going 10 - 20 miles to feed a remote cluster of customers (be it wireless, 
copper, coax, or glass as the last mile). Those are more or less limited to about 1.5 gb/s. Spectrum availability 
can reduce that. You can sometimes stack them, but again, spectrum availability would be king in that decision.
You might have fixed wireless as the last mile. We're starting to see platforms capable of multi-hundred megabit per 
customer with a sector capacity of low gigabits, but again, spectrum availability comes into play here. Those 
solutions require line of sight (or close to it) and only go a few miles. The systems that can penetrate foliage 
really cut your per-sector capacity to around 100 megabit, shared amongst all customers. Those are simply 
limitations of physics.


When you don't have the benefits of scale, the only viable path forward in a managed setting is usage-based billing, 
with some amount of included data.

We are saying the same thing re: mobile (when I say mobile I mean wireless) providers. Because spectrum is a 
limitation, capped services make sense. 

When I say "fixed line", I mean end-to-end, i.e., from CPE to nearest ISP PoP, all on wire. In such a case, if an 
operator is still offering a capped service, it is because they have no incentive (competition) to do otherwise.

Mark.

Current thread: