A Look Under The Hood Of The FCC National Broadband Map

By | July 27, 2023
FCC Broadband Map - BEAD Funding

The FCC recently released version 2 of the National Broadband Map. It’s an ambitious project that aims to provide information on the available connectivity services at every location in the US. The NTIA plans to use the map to identify unserved and underserved communities to provide funding under the Broadband Equity, Access, and Deployment (BEAD) program. I was very intrigued by the project and spent some time to explore what the underlying data says. With millions of data points, I thought I could do some statistical analysis to answer questions on the broadband market in general. Here are some observations.

The Who and What

The first thing I did was to understand what exactly are the different parameters in the database are. Then I progressively un-layered the data to distill some information. I looked at licensed fixed wireless access first. This is how the service providers explain the data they provided the FCC to develop the map:

  • AT&T: “Each AT&T living unit address is geocoded using Pitney Bowes advanced licensing dataset.  Address level data is then extracted from AT&T’s Enterprise Consolidated Data Warehouse (ECDW) with the associated network qualification information for each address, capturing all technologies and speeds available to each location.  Addresses are then matched to the Fabric addresses that have a BSL Flag of TRUE, using match fields of “address”, “state” and “zip”.  Like technologies are then aggregated to the more general tech code categories specified, and max speeds are identified for each major grouping (10, 50, 71) by address.  Finally speeds available to the address are adjusted as prescribed in the BDC instructions. Uploaded addresses are those matched to Fabric locations pursuant to FCC instructions. Unmatched addresses will be investigated and uploaded at the next opportunity.
  • T-Mobile: “T-Mobile makes excess network capacity available for fixed wireless broadband services within its national wireless footprint.  T-Mobile selected three speed tiers of expected speeds for BDC reporting purposes. It determined expected fixed broadband speeds for Fabric locations by applying link budget constraints to its national wireless footprint and by taking into consideration network capabilities as of 12/31/2022.  T-Mobile then performed a geolocation match of its address database, provided third party data vendor and used in the ordinary course of business, to the FCC Fabric, finding numerous inconsistencies and inaccuracies in the Fabric street addresses.  T-Mobile has submitted availability at expected tiers where it could find a geolocation match to the Fabric.
  • Verizon: “Verizon used its internal “open for sale” address databases, developed and used for commercial purposes, to derive this service availability data. Development included steps  to ensure any address listed as open for sale met internally set network reliability metrics.

Essentially, service providers determine whether they have coverage at a geolocation, and determine the capacity and speed of the service. For fixed wireless access, service providers use link budgets and propagation models to identify coverage and capacity as T-Mobile explicitly stated. The method is technically sound, but that does not mean the final result will match reality. For instance, Verizon said that they have an “open for sale” address database which indicates a filter that excludes geolocations which Verizon does not want, or cannot serve for different reasons. On the other hand, T-Mobile does not seem to apply such a filter which results in a much larger footprint and service availability numbers than T-Mobile.

MNO Fixed Services

The data shows a huge difference in FWA service availability between the three mobile network operators.  T-Mobile shows huge lead in service availability; while AT&T is virtually non-existent. This, I believe, is not a result of the actual coverage footprint, but of the logic on deciding what service level to assign for a geolocation. Another interesting point is that T-Mobile shows greater availability for business locations than residential ones.

Number of units of broadband serviceable residential locations for licensed fixed wireless access.
Number of units of broadband serviceable residential locations for licensed fixed wireless access.
Number of units of broadband serviceable residential locations for licensed fixed wireless access.
Number of units of broadband serviceable business locations for licensed fixed wireless access.

Yet, despite these shortcomings, we can gleam some interesting information. For instance, Verizon could provide minimum 250 Mbps service in some locations. We could attribute this to millimeter-wave FWA technology which Verizon has deployed. Verizon also has 160 MHz of C-band spectrum in many markets which could provide such level of service once fully deployed.

Fixed broadband service penetration for major mobile network operators.
Fixed broadband service penetration for major mobile network operators (residential).

Technology Performance

We also see penetration of different technologies. GEO provides ubiquitous coverage, while LEO ever so slightly trails, presumably for lack of coverage in the extreme north. GEO speeds lag that of LEO, while both trail licensed FWA. We also see unlicensed FWA trails in speed, presumably because of the channel bandwidth used in most deployments, but is capable to provide gigabit speeds where only fiber and cable dominate.

Technology penetration for residential units.

Concluding Thoughts

The FCC broadband map and underlying database present some very interesting information on the state of broadband. However, one should be careful on how to interpret the data. Many had voiced concerns about this project because the data is supplied by the service providers. They questioned whether service providers would have vested interest to inflate service availability to pre-empt potential competition in their service areas. While this is possible, there is a need for independent validation given some of the descrepencies. The challenge process seeks to address some of that; and time will tell on how successful it is. In the meantime, some states and other jurisdictions may opt to use their own maps for assessment.

A final note: We, at Xona Partners, are involved in providing advisory services to BEAD funding applicants. This includes strategic business and technology planning (wireline, wireless). In this process, we have partnered with a premier company that provides wireline and wireless techno-financial planning tools, to which we add our own proprietary methods and models. We can literally reduce the length of some planning/design cycle processes from days to hours. If you are planning to tap the BEAD funding and need assistance in strategic planning, contact me to learn more on our services and to start the conversation.