Comments
Newest First | Oldest First | Threaded View
glennfish
glennfish
1/8/2020 6:23:20 PM
User Rank
Narrowbander
Re: question on under reporting
Regarding your question about under reporting...

The FCC database has over 2,000 ISPs with data.  We think the actual number is closer to 3,000 based on crowd sourced ASNS in rural areas.


 We did a preliminary look at the level of effort required to determine under reporting by sampling some rural ISPs and determined that under reporting is primarily a phenomenon of fixed wireless providers and small regional ISPs (Cable & Fiber).

We believe, but do not know, that about 5% of the wireless ISPs never report to the FCC.  These tend to be ma/pa type enterprises with a few hundred customers each.

We believe, but do not know, that about 50% of the remaining wireless ISPs under report simply because they don't know their coverage areas until a field tech approves an install, and it never works back to the FCC as a served location.

For small non-wireless ISPs, we believe, but do not know, that for FTTP ISPs that have less than 10,000 customers, less than half of their coverage areas are reported to the FCC.  These tend to be small shops who are more involved in deployment than reporting.

It's a good research topic for sure.  :)

50%
50%
Alison D
Alison D
1/8/2020 5:50:31 PM
User Rank
Narrowbander
Re: Reliance on ISP sourcing & the challenge process
Thanks for this insightful response, Glenn. When you walked me through your organization's broadband map last year, it was fascinating to see the detailed display of infrastructure -- fiber, DSL, satellite etc. -- come to life (or not) around the country. When compared with the census-block-generated data, it was even more interesting, to put it politely. With states and counties (and perhaps municipalities) now apparently requiring more complete data about their existing and non-existent infrastructure before they hand out grants or loans, I imagine there'll be more of a call for services from the companies, organizations, universities and agencies already able to provide a more complete picture. 

Interesting point, out of many you made, about under-reporting... wonder how widespread that is? And whether it's out of disinterest or understaffing, or whether there's a real or perceived business reason?

50%
50%
glennfish
glennfish
1/8/2020 5:04:33 PM
User Rank
Narrowbander
Reliance on ISP sourcing & the challenge process
While any measures to improve on the FCC's broadband map are to be applauded, the single most egregious error is to rely on ISP generated data to indicate where broadband is, and is not.

Several examples, irrespective of the more recent efforts in Virginia, Georgia, and Kansas (not mentioned in the article)...

1.  ISPs routinely mis-report.  Even with a challenge process, the fact that a challenge is required, means the process is broken.  Most citizens would be unable or unwilling to challenge ISP provided data.  By example, I live in Le Seuer County, Minnesota.  There are 1,995 block groups in the county.  One particular ISP has reported to the FCC that they have 1 GIG FTTP in 1,478 of these block groups.  This particular ISP is a fixed wireless provider and actually covers a small number of block groups (< 100) with a fixed wireless solution of less than 6 meg down.  The protest process would require residents of 1,478 block groups to challenge this particular ISP's reported coverage.  This is one county of 3,007 in the U.S.  To rely on ISP self reporting and a challenge process is simply not going to work.

2.  ISPs routinely do not report meaningful data.  In the public 2018 release of the FCC 477 data, over 10% of the coverage reports showed ISPs reporting download speeds and upload speeds of ZERO for both residential and businesses.  After that data was scrubbed, nearly 1/4 of the remaining records reported residential download speeds of ZERO.

3.  Not every ISP reports to the FCC.  We have numerous cases where we can show that ISPs are present in an area and have never bothered to inform the FCC through the 477 process.  While over-reporting is routinely claimed, under reporting also happens.

4.  Reporting of coverage by an ISP by a type of technology does not indicate the actual consumer throughput at any point.  For fixed wireless, long range connections may not have anything close to the channel capacity claimed by the ISP.  For fiber connections, insufficient backhaul, aged equipment, may dramatically reduce the throughput well below the "rated" capacity.

With these examples, I suggest that ISP self reporting is a proven means of having unreliable data and guarantees that public money is being withheld from where it is needed, and allocated to where it is not.

The States which have undertaken their own initiatives are to be praised for their efforts, and cautioned by the lessons learned.

In my opinion the whole process should be reversed.  Citizens should be the original source of data, and the ISPs may challenge citizen data.  Crowd sourced data initiatives already exist and albeit limited in scope to date, work quite well.

 

100%
0%


Latest Articles
A budget of $61 billion to $118 billion would enable the US to bring 1-Gig to all locations that get less than 100/20 Mbit/s service that aren't tied to RDOF phase I, ACA Connects/Cartesian study finds.
Satellite-powered broadband service provider is in talks with 'several' airlines about providing connectivity for in-flight W-Fi services.
As operators mull plans for future DOCSIS 4.0 deployments, many are considering expanding the upstream to 396MHz, a move that would deliver upstream speeds of about 3 Gbit/s, CommScope exec says.
Set for launch in early 2022, ViaSat-3 will cover the Americas and represent the first in a set of new satellites that will eventually deliver global coverage.
The UK is bustling with companies eager to build the next generation of broadband networks in what looks like a potentially nasty development for BT.
Information Resources
All resources
Webinars
ARCHIVED
Thursday, December 17, 2020
12:00 p.m. New York / 5:00 p.m. London

Today’s access network architecture is under mounting pressure due to a continued surge in the number of connected devices, a proliferation of bandwidth-intensive customer applications and dramatic shifts in usage patterns related to the pandemic, such as work-from-home and e-learning.

Learn why now is the right time for cable operators to build greenfield networks or expand their existing networks with 10G PON, arming customers with high-speed symmetrical broadband. Gain a clear understanding of the drivers impacting the access network and the various approaches being considered to deliver higher speed services. Plus, find out the best practices that operators are employing as they leverage the latest in passive optical technology to future-proof their networks.

Topics to be covered include:

  • Node + 0 (Fiber Deep)
  • DOCSIS 3.1, DOCSIS 4.0 (FDX/ESD)
  • FTTP and 10G PON
  • XGS-PON, NG-PON2
  • Provisioning 10G PON within a DOCSIS B/OSS environment

Broadband World News
About Us     Advertise With Us     Contact Us     Help     Register     Twitter     Facebook     RSS
Copyright © 2021 Light Reading, part of Informa Tech,
a division of Informa PLC. Privacy Policy | Cookie Policy | Terms of Use
in partnership with