Why Starlink Will Win Rural Broadband

Update 2021-05-25 - I now consider this article Starlink, Part 2 of a series. For the entire series, click on the Starlink tag.

Update 2021-05-13 - I published a followup article to this one - More on Starlink.

...

I'm fortunate to be part of a certain online community (which I cannot name). That community has many sharp intellects and a wealth of experience and backgrounds. You know who you are :-) Thanks for all the help in fleshing out my discussion points in this article.

For purposes of this article, I posit that getting Broadband Internet to rural areas has always been problematic and the solutions to date...

• Geosynchronous Earth Orbit (GEO),
• Digital Subscriber Line (DSL) over copper telephone lines,
• Cellular data,
• Wireless Internet Service Provider - WISP,
• Cable television,
• Fiber

all suffer from quality or availability or affordability issues. Those issues have been amplified in 2020 and beyond due to the COVID-19 pandemic necessitating a shift to online work and education.

In 2020, SpaceX Starlink became a new technology for Broadband Internet Access in rural areas (beta testing phase). Starlink is a Low Earth Orbit (LEO) constellation of satellites orbiting ~210 miles above earth. Starlink doesn't intend to provide Broadband Internet Access service in urban areas, but rather underserved markets of rural areas in the US and International. Recently Starlink announced that it intended to provide mobile broadband such as semi trucks and ships. In late 2021, Starlink will commence commercial service.

Starlink differs from all of the above Broadband Internet Access methods in significant ways:

• Low latency - the latency of a 420 miles round-trip is minimal; by contrast GEO satellite round trip is 44,600 miles (twice for each packet - user <-> satellite <-> ground station and ground station <-> satellite <-> user).
• Near universal availability - nearly every spot in North America will have full-time coverage from one or more satellites at any given moment.
• High data rate - Broadband Internet speeds are a complex mix of number of users, the available bandwidth of the signal and other factors. 
• Minimal ground infrastructure needed.

Basically, buy a Starlink terminal, set it up, it orients itself, begins tracking Starlink satellites above, and you have near-continuous broadband Internet Access at acceptable latency. In 2020 and 2021, the price of the Starlink terminal (minus various optional mounting accessories) is $499, and the service fee is $99/month. As I write this in May, 2021 Starlink is still in beta testing, but claims that commercial service will commence by late 2021.

Demonstrably, even in the "beta testing" phase, Starlink just works. There are hundreds of testimonials available online. There are multiple Facebook groups which provide self-support. Rural residents are desperate for Broadband Internet Access that provides acceptable performance for videoconferencing (work, school, medical), streaming (Netflix, etc.) that is reliable.

I believe Starlink is far superior to the other options for Broadband Internet Access available in rural areas. To wit:

• vs Geosynchronous Earth Orbit (GEO) satellite - Starlink has low latency, higher speeds, is user-installable, and (at the moment) does not have transfer caps (amount of data transferred before reducing the data rate). I think that the GEO satellite Internet business, at least for consumers, is going to effectively evaporate the day that Starlink begins commercial service and providing service on demand.
• vs Digital Subscriber Line (DSL) - The copper telephone lines that DSL uses have not been maintained by the large telecommunications companies such as Frontier and CenturyLink. Because of that poor quality infrastructure, DSL is infuriatingly unreliable and data rates are highly variable. Starlink is much faster and more reliable than DSL.
• vs Cellular data - Cellular data in rural areas suffers from fewer towers and thus long distances between customers and towers. Thus the data rate is low, especially for "hotspots" (small hockey puck sized units that use cellular connections as backhaul and Wi-Fi to connect devices. Cellular hotspots are typically "capped" at some amount of data transfer per month, for example 10 GB, at which each additional GB is charged an additional fee. This can add up quickly if you stream a 4K movie or sporting event. Starlink uses the same idea - Internet Access via satellite, and Wi-Fi (or Ethernet) to connect devices. Starlink's data rate is much higher, and does not have data transfer caps.
• vs Wireless Internet Service Providers (WISPs) - WISPs are usually (but not always) handicapped by limited budgets and coverage areas limited to line of sight from their towers. Typically WISPs use the license-exempt 5 GHz band which does not have the ability to penetrate trees. (Some WISPs are beginning to use CBRS systems the 3.5 GHz band, which may provide better coverage and tree penetration.) In addition, because of the limited amount of spectrum WISPs have available, they generally cannot offer speeds above 25 Mbps to most customers. At shorter ranges, WISPs can use other spectrum to provide higher speeds. Essentially, each Starlink satellite is a "super tower" and thus can "see" a much larger geographic footprint than a terrestrial tower. Starlink offers much higher speeds than WISPs can - typically greater than 100 Mbps.
• vs Cable modem or fiber - Cable modem service, nor fiber Internet Access is rarely available in rural areas. There are exceptions, especially by progressive electricity co-ops and some privately owned and co-op telephone companies. Anyone who has fiber or cable modem service is rarely interested in Starlink because, generally, they don't need it.

Starlink Future Competition
I expect that the emergence of Starlink will catalyze all of the above options to "get serious" about deploying new generations of their tired existing services, since their current customers are no longer "captive" - they now (or soon will) have the option of Starlink. For example, in April 2021 T-Mobile debuted their T-Mobile Home Internet service which T-Mobile claims will be an improvement over current "hotspots" because of additional towers in rural areas, as well as new service centers that they will place in rural areas. I think that T-Mobile is trying to get ahead of losing rural customers to Starlink service

Similarly, in emerging from bankruptcy in May, 2021, Frontier "promises... really, this time" to begin aggressively deploying fiber in its service areas (those that it has not sold to other entities, such as Ziply Fiber in the Pacific Northwest). CenturyLink makes similar promises.

While serious competition to Starlink will undoubtedly emerge, especially if various stimulus funds from the Biden Administration actually materialize... such programs will take years to become reality. Even T-Mobile Home Internet will require a significant investment and deployment which won't be fully realized for years.

Update after hitting the Publish! button: Related article of mine - Federal money to SpaceX may hurt public broadband efforts in Washington state.

In contrast, Starlink "merely" has to do three things (that it's proven it can do well) to expand Starlink and provide the capability to service additional customers:

• Build more Starlink satellites
• Launch more Starlink satellites
• Install more Starlink Ground Stations (fiber-to-satellite transition points)

Starlink's Potential to Scale
(For a more hyperbolic take than mine on Starlink's potential to scale, see Starlink is a global ISP built at ZERO COST to SpaceX, enabling NASA’s Artemis launch. You'll enjoy it.)

Update after hitting the Publish! button: One point I meant to make is that I think it's equally as important to look at the potential of Starlink as it is to evaluate what they've accomplished to date. That's one of the primary lessons I learned from the seminal book The Innovator's Dilemma. That is, pay attention the potential to scale the technology / product / company. Are all the elements of the technology / product / company in alignment for being able to scale? As I discuss below, in the case of Starlink, I think the answer is Yes!

Why Starlink will likely become the preferred Broadband Internet system for rural residents is that it has the potential to scale far beyond the current Starlink system. Here are the salient points about that potential to scale:

• Starlink is making satellites at scale. They have a factory with a production line in Redmond, WA.
• Starlink is launching satellites at scale, 60 at a time (with current rockets).
• Starlink can expand its satellite constellation "at will" by building and launching more satellites. There are, of course, constraints such as requiring permission from various government agencies. But with their record to date, and proven demand from customers that really need good quality Broadband Internet now, Starlink has a credible cudgel to use in its expansion requests.
• Starlink satellites in low earth orbit have a limited lifespan. They will require replacement eventually. New generations of Starlink satellites will have better onboard systems enabling more users. "Radios" are now almost entirely in the computing domain rather than the hardware domain. Not only will software allow upgrades in performance in the existing satellites, but better compute capability in coming years will allow newer satellites to have even better performance. 
• Starlink will soon incorporate the ability for inter-satellite links. This will reduce the dependencies on ground stations being within the same geographic footprint as a user (bent pipe architecture).
• Starlink will continue to improve their ground terminals to be able to serve more satellites and thus more users.
• Starlink will continue to deploy additional, denser, ground terminals to be able to serve more satellites and thus more users.

Beyond the technical improvements that Starlink can, will, and is making, the economic issues about Starlink scalability are even more compelling:

• Simple economies of scale for satellites. Starlink is making more satellites than any other entity. My math is horribly inexact, but my perception is that Starlink will soon have manufactured, launched, and orbited more satellites than all the satellites that have been manufactured, launched, and orbited to date. The more you manufacture, the more efficient you get, and the less expensive each unit becomes.
• Simple economies of scale for launch. Starlink is making highly efficient use of SpaceX launch capability and refurbished rockets that are largely already paid for by previous launch customers.
• Inter-satellite links will allow economical service in very sparsely populated areas where very low populations don't currently justify placement of a ground station.
• Starlink isn't only an Internet Access system for the US; Starlink is simultaneously commencing service in other countries. Each Starlink satellite isn't idle when it's not above North America - as soon as it comes within coverage of a paying customer anywhere on Earth, it's earning revenue. This will be especially true when the inter-satellite links are activated.
• What's most compelling about Starlink's economies of scale is that they have announced that Starlink will also be capable of mobile Internet Access with the constraint that the "vehicle" must be able to accommodate the form factor of "Dishy McFlatface". Imagine having full streaming Internet on every ship, many boats, almost all recreational vehicles (campers), and semi trucks.

Even other "proposed" Low Earth Orbit satellite broadband Internet access systems won't be able to catch up with Starlink's lead. Starlink has the established (revenue) customers (will, by the time a competing system is ready to begin [if ever]), it has the production process down for manufacturing satellites at scale, and it has the cost-effective launch capability. I just can't envision other systems overcoming Starlink's first mover advantage for capturing rural customers of Broadband Internet Access. 

In conclusion, I don't see any other Broadband Internet Access technology for rural customers that can compete with Starlink on the combination of:

• Immediacy - Once commercial service commences, you can have Broadband Internet Access as fast as you can get the Starlink terminal shipped to your house).
• Availability - Can you see the sky? You can get Starlink.
• Performance - At 100 Mbps and low latency, it's good enough for videoconferencing and streaming.
• Easy (inexpensive) user installation. It doesn't get much simpler than putting it on the picnic table in the back yard and plugging it in. Oh, you'll need to read the one page of instructions and install an app on your phone.

As for price... if $100/month is unaffordable, I suspect that there will be subsidy programs that offset part or all of that monthly fee, especially when factoring in remote school and remote medicine.

I suspect that Starlink is the missing piece for many remote workers to relocate to rural areas now that they are all but guaranteed to have Broadband Internet Access that's reliable, capable, and affordable.

As I complete this article, an Ars Technical article stated that Starlink has 500k orders on the books (all with a deposit of $100). Elon Musk tweeted yesterday:
"Only limitation is high density of users in urban areas. Most likely, all of the initial 500k will receive service. More of a challenge when we get into the several million user range."

Lastly, to answer a plaint from one of the members of the "certain online community (which I cannot name)":
500k? That's just not good enough. We need to provide Broadband Internet Access to every rural resident, and there are a lot more than 500k of them.

My answer:
Point taken... but for those 500k, Starlink is a good enough solution for Broadband Internet Access that's available now... not years from now. It's my guess / bet / hope that Starlink is currently under-promising and planning to over-deliver. All the trends I'm seeing point to Starlink ultimately being able to scale to significantly more than 500k customers.

Disclaimer: I am one of those who have placed a $100 deposit for Starlink service, despite the availability of Comcast cable modem service at my house (I am currently a Comcast subscriber). My motivations for wanting Starlink service include predictable Comcast outages during emergencies.

Thanks for reading!
Steve Stroh
Bellingham, Washington, USA
2021-05-05
Copyright © 2021 Steven K. Stroh