Defining broadband down

Federal Communications Commission chairwoman Jessica Rosenworcel wants to study changing the definition of broadband from 25MB/s to 100MB/s for downloads, and from 3MB/s to 20MB/s for uploads.

Making this change would significantly impact the $42.45-billion federal Broadband Equity, Access & Deployment Program, or BEAD, grants, for broadband infrastructure to the states, passed by Congress in 2020.

But would changing the definition of broadband have any real-life impact?

Originally, broadband didn't mean a particular bandwidth (or higher), but described a particular set of communication technologies.  As is often the case with technology, however, marketing departments seized on the label broadband to denote higher bandwidth than you can get with what our competitors are offering, and bandwidth became the primary measure of an internet connection's usefulness for consumers.  From there, broadband became a sort of shorthand for high bandwidth access.

Is rating a particular access link based on its speed useful, even if we accept the shorthand equivalence between broadband and high bandwidth?

No, it isn't.

While raw bandwidth might be helpful in marketing (although most technical folks consider this a gimmick), we need to develop a new standard for acceptable consumer-grade (last-mile) internet access performance.  We need to consider factors beyond bandwidth — and be realistic about the meaning of bandwidth.

Starting with bandwidth — the bandwidth available from any given consumer-grade connection technology is highly dependent on the length of the electrical connection.  A house a mile from a cable company's facility will receive lower bandwidth than a house half a mile away just because of the realities of electrical transmission over copper wires or a wireless connection.

The physical path matters as well.  For copper, whether the cable is hanging or buried and what other cables it is near impact the amount of data the cable can carry.  For wireless links, precipitation, ambient temperature, relative humidity, and even local spectrum crowding can significantly impact available bandwidth.  (Think of a user living next to an airport versus one living on a farm with no other houses within several hundred feet.)

Many last-mile technologies share the available bandwidth across all connected users.  Two cable runs connected to the same point in the provider's network, but with two different user densities, will provide different amounts of bandwidth — even at different times of the day.

Moving beyond bandwidth, however, many other factors can impact each user's experience.  For most modern applications, the amount of jitter matters as much as the available bandwidth.  Jitter is the difference in delay between any pair of packets transmitted across the link.

Resilience and reliability also significantly impact the value of a last-mile connection, as noted by Docket FCC 22-50.  Providers with more connectivity to their peers and upstream providers will generally provide more stable service.  Emphasizing last-mile and internal network resilience also increases last-mile service value.

There is almost no way to know how an individual link will perform in the real world without testing — and ultimately using — it.

Suppose the FCC is going to create a new definition of broadband.  In that case, it should include a realistic meaning and consider the many other factors impacting consumer-grade internet services.

Russ White is a well known voice in computer networking, where he advocates for simplicity, privacy, and the decentralized internet.  He co-hosts the Hedge podcast, serves on the Internet Architecture Board, and serves in a leadership role in the FR Routing open-source community.  Russ has co-authored more than forty-five software patents, fifteen books, and many hours of video training.  He holds a Ph.D. in philosophy from Southeast Baptist Theological Seminary, an MACM from Shepherds Theological Seminary, an MSIT from Capella University, and several expert and architectural information technology certifications.  Russ recently published Unintended Dystopia and Unfriending Dystopia, two books about social media's impact, and is working with the Discovery Institute.  His LinkedIn page is here.

Image: Pixabay, Pixabay License.

If you experience technical problems, please write to