Happy Birthday to Ethernet

The idea for Ethernet was born fifty years ago in May 1973 when Robert Metcalf coined the word Ethernet. He had been studying ALOHAnet, which was developed at the University of Hawaii in 1971 and was the first public demonstration of a wireless packet data network. Metcalf used the work Ethernet as a reference to luminiferous aether, a concept that been postulated in the 17th century to explain how light could be transmitted through a vacuum.

Xerox filed a patent for Ethernet in 1975 that listed Metcalf, David Boggs, Chuck Thacker, and Butler Lampson as the inventors. Metcalf left Xerox in 1979 and formed 3Com, which eventually was purchased by Hewlett-Packard. If you’ve been in the industry for a while, you’ll remember the first 3Com routers and switches. As part of the transition to 3Com, Xerox gave up the Ethernet tradename and Ethernet became a worldwide standard in 1980 with the first standard labeled as “The Ethernet, A Local Area Network. Data Link Layer and Physical Layer Specifications”.

Ethernet is a family of computer networking technologies that is the heart of modern computing. The Ethernet standard divides a stream of data being transmitted into shorter pieces called frames or packets. Each frame contains the needed details of the data transmission, including the source of the transmission and the destination. Each frame includes error-checking data that can be used at the receiving end of a transmission to make sure that all intended data packets are received.

The use of frames is the basis for how the Internet functions. If data is sent to your home broadband router, the router checks to see if each incoming frame is complete. Your home router will send a request to retransmit any frame that showed up incomplete. Most people think that data is simply sent and received, but the reality is that there is a complex system of error-checking and resending missing frames that happens behind the scenes.

The system of resending incomplete packets explains why connecting to a server on another continent is slower than connecting to a server much closer to where you live. The farther data has to travel, and the greater the number of intermediate data switching points, the higher the chance that incomplete frames make it to the destination – and the longer it takes to replace incomplete or missing frames.

This is also part of the reason why a connection like a Zoom call is often of lower quality on a low bandwidth connection. When the broadband speeds aren’t fast enough, there is a delay in getting enough frames delivered to create a coherent picture in a real-time Zoom connection. When frames are missing or show up late, the Zoom call pixelates or has voice quality issues because not all needed data frames were delivered in time.

Ethernet was first used inside wired local networks, such as between the buildings on a campus, or inside of a business to create WAN and LAN connections. But it quickly became the standard for public data transmission and became the heart of the Internet. Ethernet is used in all common broadband technologies from DSL, fixed wireless, and fiber data transmission. Ethernet is used in the WiFi connection between your home broadband router and your home devices.

The protocol had an original data rate of 2.94 Mbps. The current Ethernet standard supports speeds up to 400 Gbps, but data rates up to 1.6 Tbps are under development. Since the 1980s, the Ethernet standard has continuously been updated and improved and now allows for the complex routing protocols in place to operate the Internet.

Designing for Growth

I live in Asheville, North Carolina, where a lot of the neighborhoods like mine have been here for a long time. In fact, most of the houses in my neighborhood were built one hundred years ago – and there are plenty of neighborhoods that are older.

I also live in a city that has been booming over the last decade or two. There have been a lot of new houses built at the fringe of the City and a lot of infill construction where somebody has built a house on almost every vacant lot in the City. Folks are even starting to build houses in what used to be their backyard. I also know of a half dozen multi-dwelling buildings and complexes being built.

This kind of growth means that there is a lot of stress on the existing utilities. Both the electric grid and water system were built for a smaller city, yet each has been expected to somehow accommodate the outward and infill expansion. As you might expect, the over-taxed utilities are showing the strain. There seem to be electrical outages almost every time we have serious rain. We have a few major problems every year with the water system.

Our telecom networks are not immune from growth problems. The Charter cable network clearly has occasional problems. The company never explains problems, but it’s not hard to imagine that there are neighborhoods where the cable network is overloaded due to growth. I also imagine that the Charter network is a hodge-podge of different neighborhoods built over time – it is not one big ubiquitous network but rather a patchwork of smaller networks built in different decades that are all grafted together. That means different generations of coaxial cable, different sizes of cable nodes, and different schemes and designs of taps and amplifiers.

The growth story in Asheville is mild compared to some of the places I’ve lived in my life. In the early 80s, I lived a block outside of the Dallas city limits in Richardson, Texas. At that time, there wasn’t a whole lot in Richardson, which was the outer northern fringe of the suburbs. North of Richardson were places like Plano, which I remember as having a few bars and a rodeo, but which was mostly open country. Today, the Dallas suburbs spread far north of where I used to live forty years ago.

My latest electrical outage made me think about designing fiber networks for growth. There will surely be some fiber networks built with grant funding in rural areas that, over the next fifty years, will be overrun by urban expansion. How easy will it be to expand fiber networks in areas with fast growth?

Passive PON networks are the best at handling growth. The good analogy for thinking of a PON network is to envision one of the many board games that are based upon hexagons. PON technology can handle growth by adding new hexagons as long as the ISP can get enough bandwidth to feed the new neighborhood. A PON network can theoretically handle nearly unlimited growth. If an existing hexagon adds a lot of new homes, it’s just as easy to add new PON core OLTs.

This is not to say that PON networks can’t get into trouble with fast growth. If there had been a PON network in Richardson in the 1980s, there would be a good chance that expansion happened so quickly that there would not have been enough extra fibers in the network to fully handle growth. It’s almost certain that if PON had been in Richardson in the 1980s, there would have been several upgrades to the backbone fiber feeding the neighborhoods to get enough broadband to satisfy customer demand.

Active Ethernet networks have a harder time handling runaway growth. Since there is a last-mile fiber for every customer, it’s not hard to envision adding more homes than existing fiber can handle. There is no real issue expanding active Ethernet into new greenfield neighborhoods, but fast infill growth might mean building new neighborhood fiber at some point.

Not all fiber networks are alike. I know some ISPs today that are building what I consider to be slim fiber networks, meaning a network with a minimum of extra fibers. The chances are that in most rural places this will be okay – but it means a lot of future investment in areas where growth shows up unexpectedly. And that means that it’s going to be important to be served by an ISP that is ready and willing to invest in keeping up with growth. We’ve seen copper and coaxial networks deteriorate when the network owner wouldn’t spend the needed capital.

I am positive that the city planner in Richardson in 1980 did not foresee that Richardson would become a densely populated inner suburb today – or if they did, nobody believed them. Fifty years is a long time, and I’m positive that some of the folks building rural fiber networks will be equally surprised by the growth. But they can make growth easier with some early planning during the original design.

Are You Ready for WiFi 7?

It wasn’t that long ago that we saw a major update to WiFi standards with the release of WiFi 6 in 2019 and WiFi 6E in 2020. But we’re on the verge of the next generation of WiFi with the official launch of the new WiFi 7 standard in November 2022. There has already been a soft release of WiFi 7 routers in China, and we’ll start seeing the new routers in the market here sometime this year.

The performance of WiFi 7 is much improved over its predecessor:

  • A WiFi 7 router will be able to connect to more simultaneous devices, which will be useful in hotels and other public settings.
  • WiFi 7 has vastly improved theoretical speeds. WiFi 6 could support a speed as fast as 9.6 Gbps. WiFi 7 will allow for a connection as fast as 36 Gbps.
  • The new standard will be better able to maintain a low-latency connection to devices.
  • WiFi 7 will allow for WPA4 security, but this part of the standard has not been fully ratified.

The new standard achieves improvement in several ways. First, the new standard can spread a single broadband connection across frequency bands and effectively create wider bandwidth channels. Where WiFi 6 had a maximum channel size of 160 MHz, the new WiFi 7 can create a channel as wide as 320 MHz.

The new WiFi standard will use the three existing WiFi frequencies of 2.4 GHz, 5 GHz, and 6 GHz. The big improvement over WiFi 6 is that WiFi 7 can reach across all three frequencies to create a connection, while WiFi 6 was limited to only use 6 GHz spectrum.

WiFi 7 also introduces a new feature called Multi-Link Operation (MLO) which will allow for the simultaneous sending and receiving of data within one channel – something that’s never been available with WiFi before.

WiFi 7 also has an interesting feature called Multi Resource Unit Puncturing that allows for a router to grab any unused spectrum within channels that are in use. This means that all of the bandwidth in the frequency band is available for use. In current WiFi technology, when a channel is connected to a device, the entire channel is dedicated to that one use, even if the bandwidth use is small.

Interestingly, these features look a lot like the features that are supposed to be used by 5G. It seems WiFi manufacturers have beaten the cellular companies to the market with some of these features that more efficiently use bandwidth.

An immediate use could be in offices where WiFi 7 can allow the transfer of data between devices at 10 Gbps without wires. An office only has to upgrade to WiFi 7 devices when they become available. While this also could mean a big improvement in places like a stadium or a convention center, it will take years until enough users have WiFi 7 capable devices. The ultimate benefit of the larger channel sizes and faster speeds is to enable technologies like ultra-high definition video or augmented reality.

Just like with WiFi 6, we won’t see much WiFi 7 in use until there are both routers and devices enabled to use the routers. Most of the devices in our homes are still not WiFi 6 compatible, and there are still not a huge number of homes with a WiFi 6 router. But over time, new WiFi 7 devices will include the standard and will creep into our homes and businesses.

The Latest FCC Maps

As promised, the FCC released a new set of maps on May 30. These are supposed to be the maps that will be used to allocate the $42.5 billion in BEAD grant funding to states. Broadband analyst Mike Conlow quickly published a blog on Substack about the new mapping data that includes a summary of the new map in easy-to-understand tables. Mike’s summary shows that there are more than 114.5 million broadband passings in the country – locations that could be broadband subscribers). That’s an increase of over 1 million locations since the last version of the FCC maps.

More importantly, the new maps can be used to count the number of households that can buy broadband at various speeds. The $42.5 billion in BEAD grant funding will be allocated to states according to the number of unserved locations – places that can’t buy broadband at a speed of at least 25/3 Mbps. Locations are underserved if there is an ISP that offers broadband between 25/3 Mbps and 100/20 Mbps. According to Mike’s quick math, there are 8.67 million unserved locations and 3.55 million underserved locations. Mike subsequently corrected the number of unserved locations to 8.3 million.

Anybody who is intimately familiar with the FCC maps knows that there is a lot of fiction buried in the reporting. There is one huge flaw in the FCC mapping system that has carried over from the previous FCC mapping regime – ISPs self-report the speeds they can deliver. Per the FCC mapping rules, ISPs can claim broadband marketing speeds rather than some approximation of actual speeds. In every county where I’ve delved deep into the local situation, I’ve found multiple ISPs that are overclaiming broadband speeds.

ISPs vary widely in how they report broadband speeds to the FCC. I see some ISPs who meticulously categorize customers into a dozen or more speed tiers. It’s fairly obvious that these ISPs are trying to accurately show the speeds that are available. But there are also ISPs that claim the same speed over a large geographic area. In today’s world, I’m always instantly suspicious of any ISP that claims exactly 100/20 Mbps broadband since that conveniently classifies those locations as served. An ISP making that claim is telling the FCC that everybody in their service footprint already has adequate broadband and that there is no need to give grant money to anybody to compete with them.

But such a claim is ludicrous if the ISP is deploying a technology like DSL, cellular wireless, or fixed wireless where it is impossible that every customer over a wide geographic area to get the ISP’s top claimed speed. Such claims are easy to debunk when you look closely. For example, customers only a few miles from a DSLAM or a tower can’t get the fastest speeds. There are multiple reasons why a given customer’s speed might be slower. Such claims are even more quickly debunked when looking at detailed Ookla speed tests.

A second flaw in the FCC maps is the coverage areas claimed by ISPs. The FCC is counting on public broadband challenges or challenges by State Broadband Offices to somehow fix this problem – but that’s an unrealistic hope. Most people don’t know about the FCC maps and the challenge process – and even people who know about it are not motivated to file a challenge about an ISP that claims service at their home that’s not really available. This issue can apply to any technology, but it’s particularly a problem for WISPs and cellular broadband. It’s not easy for a knowledgeable engineer to accurately judge the coverage area of a wireless network from a given tower – I have to think it’s beyond the capability of the folks at a State Broadband Office to understand it enough to challenge coverage. But it doesn’t take any expertise to know that a WISP or a cellular company claiming ubiquitous 100/20 Mbps coverage across large areas is exaggerating both speed and coverage capabilities.

It’s going to be interesting to see how States react to these final counts. There have been rumors about states ready to sue the FCC and the NTIA if they feel these maps will cheat them out of funding. There has been legislation introduced in the Senate that would force the NTIA to wait longer for better maps before allocating most of the funding. It’s going to be surprising if nobody pops up to challenge the allocation of the $42.5 million dollars. A challenge could pluge the BEAD grants into huge uncertainty.

An even bigger issue is if the FCC maps will be used to determine the locations that are grant eligible – because that would be a travesty. That would mean that every ISP that claims a bogus 100/20 Mbps broadband coverage will be rewarded by keeping out competition from grant funding. Regardless of how the funding is allocated to States, Broadband Offices need to be the ones to determine which locations in their State don’t have good broadband.

Broadband Customers 1Q 2023

Leichtman Research Group recently released broadband customer statistics for the end of the first quarter of 2023 for the largest cable and telephone companies. Leichtman compiles most of these numbers from the statistics provided to stockholders other than for Cox and Mediacom, which are estimated. Leichtman says this group of companies represents 96% of all US landline broadband customers.

The first quarter of the year shows a continuation of the trend where all of the growth in broadband is coming from T-Mobile and Verizon FWA fixed cellular wireless. Those two companies added 916,000 out of 962,000 total customer growth for the quarter. T-Mobile passed Lumen and Frontier in the quarter in terms of the total number of broadband customers.

% 1Q
1Q 2023 4Q 2022 1Q Change Change
Comcast 32,324,000 32,319,000 5,000 0.0%
Charter 30,509,000 30,433,000 76,000 0.2%
AT&T 15,345,000 15,386,000 (41,000) -0.3%
Verizon 7,528,000 7,484,000 44,000 0.6%
Cox 5,565,000 5,560,000 5,000 0.1%
Altice 4,612,700 4,632,000 (19,300) -0.4%
T-Mobile FWA 3,169,000 2,646,000 523,000 19.8%
Lumen 2,981,000 3,037,000 (56,000) -1.8%
Frontier 2,863,000 2,839,000 24,000 0.8%
Verizon FWA 1,866,000 1,473,000 393,000 26.7%
Mediacom 1,470,000 1,468,000 2,000 0.1%
Windstream 1,175,000 1,175,000 0 0.0%
Cable ONE 1,063,000 1,060,400 2,600 0.2%
Breezeline 689,903 693,731 (3,828) -0.6%
TDS 515,400 510,000 5,400 1.1%
Consolidated 369,862 367,458 2,404 0.7%
Total 112,045,865 111,083,589 962,276 0.9%
Cable 76,233,603 76,166,131 67,472 0.1%
Telco 30,777,262 30,798,458 (21,196) -0.1%
FWA 5,035,000 4,119,000 916,000 22.2%

The telcos collectively lost 21,000 customers for the quarter, but unlaying that number is the success story where telcos are collectively now adding as many customers on fiber as are being lost on DSL. After so many years of seeing Frontier bleeding broadband customers, it’s interesting to see the company now growing faster than the big cable companies. It’s hard to think that the telcos overall won be in a positive growth mode soon.

The only cable company with any significant growth is Charter – and that growth likely comes from the company now constructing fiber in some of the markets where it won the RDOF subsidies.

The only company on the list with a significant loss is Lumen, which lost 1.8% of its customers in the quarter. Lumen is also the telco with the least aggressive fiber growth plans.

There are several companies conspicuously missing from the list. It’s hard to think that Brightspeed and Google Fiber are not larger than the companies at the bottom of the list.

2023 Broadband Legislation

I guess it’s inevitable that a $42.5 billion grant program would attract a lot of legislative action trying to set some parameters on how to spend the money. This was witnessed at the recent hearings at the House Commerce Committee’s Subcommittee on Communications and Technology which looked at issues that would expedite the deployment of broadband infrastructure. At the hearing, it was mentioned that the subcommittee was reviewing 32 House bills related to broadband infrastructure, many of the bills still in draft form. I’ve followed broadband legislation for decades, and I can’t ever recall seeing this much attention on broadband issues by federal legislators. Following are just a few bills that show the breadth of House legislation being considered:

Broadband Expansion and Deployment Fee Equity and Efficiency or “BEAD FEES” Act. This draft bill from Rep. Rick Allen of Georgia would require states that accept federal BEAD money to ensure that application fees charged by state and local governments are transparent, competitively neutral, and cost-based.

Reducing Barriers for Broadband on Federal Lands Act. Rep. Russ Fulcher of Idaho is drafting a bill that would eliminate the need for environmental or historic preservation reviews for deploying broadband projects on previously disturbed federal lands.

Timely Replacement Under Secure and Trusted for Early and Dependable Broadband Networks or “TRUSTED Broadband Networks” Act. Rep. Brett Guthrie of Kentucky would eliminate the requirement for an environmental or historic preservation review before removing and replacing network equipment that affects national security.

Facilitating DIGITAL Applications Act. Rep. Mariannette Miller-Meeks of Iowa wants the NTIA to update Congress on whether the Departments of Interior and Agriculture have created an online portal for tracking applications to deploy communications infrastructure on federal property.

The Broadband Incentives for Communities Act (H.R. 1241). Rep. Lizzie Fletcher of Texas would require the NTIA to establish a grant program to assist local governments and Indian tribes with the efficient review of zoning or permitting applications for broadband infrastructure.

The Senate has also been busy. I wrote recently about a bipartisan bill from Sen. John Thune (R-S.D.) and Jacky Rosen (D-Nev.), the Accurate Map for Broadband Investment Act that would require the FCC to fix its broadband maps before dispersing the majority of the BEAD funding to states.

Four senators recently introduced the bipartisan Rural Internet Improvement Act that would streamline the grant funding process for the USDA’s ReConnect grants.

There is also legislative activity in some states. The bill that might get the most attention comes from the Washington legislature, which has included language in the annual House budget bill that would require any ISP that accepts BEAD grant funding to provide open-access. That means ISPs taking grant funding would have to sell wholesale connections to any other ISP on grant-funded networks. Washington has a long history of open-access, and for years it was the only allowable business model for the many countywide municipal electric companies. This law is facing obvious major opposition from cable companies and big telcos which have always resisted all laws that would require them to share their networks. One, article I read said that cable companies are threatening to stop making infrastructure investments in cities if this law passes.

It’s hard to know what this flurry of legislative activity means. It’s likely that most of the proposed bills won’t make it through the process to become laws. But some likely will. It’s hard not to think that some of this legislation will change the BEAD grant rules between now and the time that ISPs file for and receive BEAD grants. I guess we have to add legislative uncertainty to the list of issues that are complicating the BEAD grant process.

FCC Touts 6G

The FCC has seemingly joined forces with the marketing arm of the cellular industry in declaring that the spectrum between 7–16 GHz is now considered to be 6G. Chairman Jessica Rosenworcel recently announced that the agency would soon begin looking at the uses for this spectrum for mobile broadband. Specifically, the agency will be looking at 550 MHz of spectrum between 12.7-13.25 GHz for what Rosenworcel characterized as airwaves for the 6G era.

This 7-16 GHz spectrum is already used for a wide range of purposes, including fixed point-to-point microwave links, radio astronomy, communications with airplanes, and various military uses. Probably the biggest current use of the spectrum is for communicating with satellites. Rosenworcel said the agency would consider ways to share some of the spectrum between satellite and terrestrial uses.

The use of the 6G description for this spectrum is a big departure from the recent past. It was just in 2019 when Verizon defined 5G to include the millimeter-wave spectrum as high as 28-39 GHz as part of 5G. I’m sure most of you remember the never-ending TV commercials showing cellphones receiving 1-gigabit speeds. Verizon and a few other cellular carriers had deployed millimeter-wave spectrum in downtown areas of a few major cities as a gimmick to show how fast 5G could be. Verizon labeled this as Ultra Wideband to distinguish it from the 4G LTE spectrum that Verizon and others were starting to label as 5G.

It has to be confusing to be a cellular customer because I try to follow this stuff, and I can’t keep up with the cellular marketers. When Verizon used millimeter-wave spectrum and labeled it as Ultra Wideband, the company flashed a 5G UW icon to users to denote having access to the superfast speeds. But I’m hearing that people are now getting the 5G UW icon when connecting to Verizon’s C-Band spectrum, which is mid-range spectrum between 3.7-4.2 MHz.

The funny thing about everything that cellular marketers are doing is that 5G has nothing to do with any specific frequency range. 5G is a set of specifications to define how cell towers work, and the specification can be used with any spectrum. The 5G spectrum can work in the mid-range spectrum, in the band that the FCC just labeled as 6G, and in the higher millimeter wave spectrum.

I’m mystified that the FCC would suddenly label the spectrum between 7-16 GHz as 6G. There will be no 6G specification – anything we do in this spectrum will still either use the 4G LTE or 5G specifications.  Wireless scientists around the world have started experimenting with what they are calling 6G using terabit spectrum that ranges between 100 GHz and 1 THz. These high frequencies sit right below light and have the capability of being harnessed to transmit huge amounts of data for short distances, such inside superfast computer chips. Scientists expect within the next decade to develop the new 6G specifications.

Scientists understood that the 5G specifications would cover all spectrum up to 100 GHz. But apparently, we’re going to now carve up spectrum into tiny slices and label each tiny slice as a new generation of G. I’ve always joked that we’re going to be to 10G before we know it – and it turns out that was no joke at all and extremely conservative.

Behind all of the confusion behind mislabeling things as 5G and 6G is the fact that we will eventually need new cellular spectrum. Cellular networks seem robust today, but the demand for mobile data keeps growing. There are already a lot of complaints that the new spectrum labeled as 5G is overcrowded. The FCC knows it takes many years after declaring a new cellular spectrum until it shows up in the market. This is the time to look at new spectrum bands to put into use a decade from now. This is not going to be easy because satellite companies will be screaming loudly that cellular companies are trying to steal their spectrum. They aren’t completely wrong about this, and I don’t envy the FCC the job of refereeing between the competing uses of spectrum. Just recently, the FCC made it easier for satellite providers to share in existing spectrum bands. But when the FCC labeled this spectrum as 6G, I think we already know it ultimately favors the cellular companies.

Is Charter the Largest Rural ISP?

Once in a while, I see something in the industry press that gives me a pause. Telecompetitor reported that Charter CEO Chris Winfrey said on the company’s first quarter earning call that Charter is the “largest rural provider today.” As much as I work in and track the industry, I would never have connected the dots enough to think that.

I can see how Charter is on the way to being a big rural player. The company was the largest winner of the RDOF reverse auction in terms of passings and is slated to bring broadband to pass over 1 million rural homes and businesses. The company says it is ahead of schedule and has already built 40% of those passings. But does passing 400,000 homes make Charter the biggest rural provider in the country?

In the last few years, there has been an explosion of FWA fixed cellular wireless from T-Mobile and Verizon. At the end of 2022, T-Mobile had over 2.6 million FWA customers and added 524,000 in just the fourth quarter of 2022. Verizon had almost 1.5 million customers and added 389,000 in the fourth quarter. While not all of those customers are rural, it seems likely that both companies have a lot more rural customers than Charter.

It’s hard to get specific statistics from the big telcos, but it’s hard to imagine that CenturyLink and Frontier don’t still have more rural customers than Charter. In all fairness, the rural telco DSL customers are the prime target for Charter and everybody else who is building rural networks – but it’s unlikely that Charter has yet eclipsed them in customer counts.

Jonathan Chambers of Conexon wrote a recent blog that notes that electric cooperatives collectively have more rural customers than Charter.

Nobody knows who the eventual biggest rural winner will be. There are somewhere north of 10 million rural passings that will be tackled by the upcoming BEAD grants. Meanwhile, huge amounts of funding have been provided in rural America through CARES and ARPA funding administered through states or awarded by local governments. I think we’re going to have to wait for the BEAD grants to play out to see who will ultimately be the largest rural ISP.

And even at the end of those grants we might not know. I’ve been predicting that there will be a major roll-up of last-mile fiber networks, and there is no reason that won’t include rural properties. We might have to wait a decade to see who the biggest rural players will be.

I have to think that Winfrey knew his statement wasn’t factual, and I think that he was making the point that Charter is now a major player in rural America. He caught the industry’s attention through the statement which was aimed at Charter’s stockholders. We’re seeing big cable company customer counts level off after a decade of spectacular growth, and I think his message was that Charter is still a growing company.

One thing that Charter didn’t say is that whoever builds fiber in rural areas today is creating monopoly markets. It’s going to be hard for anybody to compete against rural fiber over the long run, and Charter and other companies pursuing grants are counting on being the monopoly provider across large swaths of rural areas. I see a lot of speculation asking why companies are pursuing rural broadband – and I think the appeal of having markets where a company will eventually have an 80%+ market penetration is something that pencils out well.

Unwinding the PSTN

This blog is aimed mostly at telephone companies and various CLECs who have been operating on the legacy Public Switched Telephone Network (PSTN). This network has been used for interconnection to the local incumbent offices and tandem switches, for connecting to 911 centers, for connecting to operator services, for connecting to cellular carriers, or for connecting to other neighboring carriers.

At CCG, we are finally starting to see that network being shut down, route by route and piece by piece. But like everything related to operating in the regulated legacy world, it’s not easy to disconnect the PSTN connections called trunks. The big incumbent telcos like AT&T, Verizon, CenturyLink, and others will continue to bill for these connections long after they stop being functional.

I don’t use this blog to make many pitches for my consulting practice, but I think we’re one of the few consultants left in the industry that can help to unwind and stop the billing of the old PSTN network arrangements. We spent many years helping ILECs and CLECs originally order these connections. The ordering process for the PSTN has always been complicated and cryptic. Carriers need to go through those same systems to cut a circuit dead. You often can’t stop the billing by calling or writing to the incumbents – network arrangements need to be unwound in the reverse manner they were built in the first place.

It’s not surprising that this is hard to do. The ordering system was made difficult on purpose after the big telcos decided they didn’t like the requirements of the Telecommunications Act of 1996 that required them to share their networks with other carriers. After that FCC order, big telcos purposefully made it hard to initiate a connection with them – and now it’s just as hard to disconnect. The big telcos will be glad to continue to bill for circuits for years after they no longer work.

I have no idea how long it’s going to take the PSTN to die, but it’s finally starting to be disassembled, piece by piece. In some ways, it’s a shame to see this network die because it was the first nationwide communication network. It was built right, and it was reliable. Outages came from the same issues that still plague networks, and a fiber cut has always been able to isolate a town or a region from the PSTN.

Sadly, the big telcos never spent the money to create route redundancy. Folks like me have shouted for decades that there was no way to justify multi-day rural network outages when we know how to solve the problem. These outages are still happening today – and the fibers that carry the PSTN are often the same fiber routes that act as the only broadband backbone route into a rural area.

I remember twenty years ago when I had a few small telephone company clients who were willing to solve the redundancy problem by building a new fiber route. We were shocked when Verizon and AT&T refused to connect the new routes into the PSTN. Apparently, the big telcos were more worried about being bypassed than they were about having a more reliable network.

Over time, and as a result of some orders from State regulators, the big telcos allowed route redundancy when somebody else paid for it. Today, large carriers like Level 3, Zayo, and many others cross the country with alternate transport routes, but unfortunately, there are still a lot of rural places where the only available fiber comes from the incumbents.

If you are having problems disconnecting or rearranging connections with other carriers, give us a shout. This could be connections with a large telco, with cellular towers, or to other local carriers. You can contact Derrel Duplechin at CCG at dduplechin@ccgcomm.com. We hate to see the PSTN starting to go. But even more, we hate to see folks who can’t figure out how to get a divorce from the big telcos.

Starlink’s New Business Broadband

Starlink has quietly updated its business broadband offerings. The original plan for businesses was $500 per month with a two-terabyte data cap. If a customer exceeded the data cap, the speed reduced to 1 Mbps for the remainder of the month unless a customer bought additional broadband at $1 per gigabyte. Starlink business comes with a premium antenna from HP at a one-time cost of $2,500.

The new plans are:

  • 1 TB Data Cap. $250/month plus $2,500 equipment costs.
  • 2 TB Data Cap. $500/month plus $2,500 equipment costs.
  • 6 TB Data Cap. $1,500/month plus $2,500 equipment costs.

Extra data now costs $0.50 per additional gigabyte.

Starlink promises faster speeds for businesses with the HP business antenna. This antenna has a 35% better field of view, is less sensitive to hot weather, handles rain better, and melts snow faster. The company now claims the following speeds on its website:

Download Upload
Residential 20 – 100 Mbps 5-15 Mbps
Business 40 – 220 Mbps 8-25 Mbps
RV 5-50 Mbps 2-10 Mbps

Interestingly, the speed claims above from the Starlink website are a lot slower than what was promised as recently as September 2022. For example, residential customers were told in 2022 that download speeds would be between 50 – 200 Mbps with upload speeds of 10 – 20 Mbps. Customers have been saying online that speeds are getting slower – something that has been validated by Ookla speed tests.

In the most recent FCC maps, Starlink claims speeds up to 350/40 Mbps. That matches the maximum speeds that Starlink promised to business customers in September 2022. We’ll have to see if the company drops the speeds claimed to the FCC now that it has dropped its maximum claimed speed down to 220/25 Mbps.

On May 9, Starlink notified customers that it would no longer deprioritize traffic after a customer hits the monthly data cap. Customers were being slowed to speeds as slow as 1 Mbps. Now, customers can sign-up to automatically be billed for excess data usage.

To some degree, the business offering is going to be a concern for some residential customers since business customers will get bandwidth priority. That might make a difference in neighborhoods with multiple business customers.

It will be interesting to see how Starlink performs over the long run. The company still has plans to add many thousands of satellites. But the company still has a waiting list of customers – and the company admits that it can get easily get oversubscribed in a neighborhood.

 

In 2021, Elon Musk said that he foresaw a future where Starlink could provide backhaul bandwidth to rural cell towers. That may still be coming in the future, but not with the current constellations. The speeds above are not nearly what a cell tower owner wants to buy. Even the most rural small cell site is going to want a steady 500 Mbps of bandwidth, with a more typical requirement of 1 Gbps. I would think that residential subscribers have to hope the company never sells to cell towers, or the coverage at peak times could suffer.