Thread by @steve_larrison

Has the partnership between Verizon and Kuiper made SpaceMobile obsolete?

A thread

$ASTS
#SpaceMob

1/n

I didn't have time to dig in at the time, but in late October, a deal was announced between Verizon and Kuiper to develop technology to harness the power of LEO based satellite broadband to provide mobile data services.

cnbc.com/2021/10/26/ver…
2/n

The only thing that could have stoked the FUD machine more is if it was StarLink instead of Kuiper that was teaming up with Verizon.

Face it, since Jeff Bezos stepped down as CEO of Amazon, he lost some of the "it" factor.

But those powers still run strong in Elon Musk.

3/n

The idea behind pairing up a satellite based broadband network that was NOT designed to work with cell phones and a cell tower is a relatively simple concept.

But there are certainly complexities involved in building out a cellular network.

Essentially a cell tower is a
4/n

fixed structure with radio gear that can broadcast a signal that mobile phones are capable of receiving. It also contains antennas capable of receiving signals from cell phones within range.

This structure is connected to a phone line through cable or fiber, and it also
5/n

needs power to operate.

The cost of building a cell tower is highly variable - more on that later. But one variable is proximity to a phone line. (Proximity to power is also a concern)

If you are constructing a cell tower in a remote area, the cost of tapping into a phone
6/n

line can be quite high - or certainly higher than if you were constructing a tower in a developed area.

So the idea behind the marriage of a satellite broadband network and a cell tower would look something like using the cellular radio and antennas to talk to phones and
7/n

using a satellite based internet backhaul to transmit voice using VOIP to transmit conversations to a location where voice signals could be put onto a traditional phone network.

It's a bit clunky, but on paper it sounds like it will work.

So should SpaceMob be worried about
8/n

this partnership?

Or about future partnerships with Elon Musk? (You have to admit that he creates some pretty good memes)

In a word - NO.

But if you have read any of my past threads, you already know that I never give one word answers when there is an opportunity to
9/n

let people understand a topic.

The primary advantage that SpaceMobile has over the competition is that it doesn't require the end user to acquire additional hardware.

The secondary advantage is that it is compatible with existing cell phones.

That means that if your LEO
10/n

constellation can't talk directly to a cell phone, you have to take on additional infrastructure costs. And to date, there is only one other company - Lynk Global who is attempting to do what AST is doing.

So a logical question would be "How much does it cost to build a
11/n

cell tower?"

When I typed "average cost to build a cell tower" into Google, it quickly came back with some extracted text from an article in the search results with a figure of $175,000.

But when I read the article that Google extracted the text from and looked at some of
12/n

the other results on the page, I quickly became dissatisfied with $175,000 as an answer.

The answers were from old articles and there was a wide variance in price points.

Thinking about the problem, I assumed that location would have a lot to do with costs.

Rural areas
13/n

would be likely to have lower permitting costs and land acquisition costs than developed areas.

But rural areas would be likely to have higher infrastructure development costs, shipping costs (equipment) and transportation costs (people building the cell towers)
14/n

Also, another thing that I found dissatisfying with my initial research into the costs of building cell towers was lack of detailed cost breakdowns of the various things that go into the construction of a cell tower.

Then I ran into one of those "Be careful what you wish
15/n

for, because you just might get it" moments.

As part of COVID relief funding in 2020, the US government taked on a bill to fund replacement of all telecom equipment in use in the US that was manufactured by ZTE and Huawei

reuters.com/article/usa-in…

16/n

If you think about it, it makes sense to spend $1.9 billion replacing telecom equipment as part of a COVID bill.

Everyone knows that 5g cell towers cause COVID.

(One downside of tweeting using my real name is that I have to put disclaimers like this in my tweets - that was
17/n

intended as a joke.)

At the cost of $1.9 billion to US tax payers, I got a good source of detailed data on the cost of constructing cell towers!

Here is a link to the cost estimates for replacing the Chinese telecom equipment.
ecfsapi.fcc.gov/file/032507837…
18/n

As soon as you finish reading the 152 page report, you will be ready to continue reading this thread.

Just kidding.

I didn't even read the entire report.

I just read enough to understand some of the cost drivers behind a more current cost estimate of the cost to
19/n

construct a cell tower.

I found the previous link in this article.

urgentcomm.com/2021/04/29/her…

The cost to build a 5g cellular network for 50,000 customers ranges from $250,000 to $1,200,000.

That's still an extremely broad range, but now I have access to data to back up
20/n

those numbers. (Without data, all you have is an opinion.)

So what causes such a broad range of costs?

The factors I mentioned earlier certainly play a role, but that is not all.

The first thing that needs to be examined is the type of cell tower being used.
21/n

There are a few different styles of cell towers. Honestly, which type of tower is best for given situations goes beyond my area of expertise. (But it leads to an area of costs that I hadn't considered)

The most common type of tower is a mono tower - these range in height
22/n

50 to 200 feet.

I am guessing that these work best in areas where there is high density usage - they aren't very tall, and for inter-city usage, their range doesn't have to extend very far.

23/n

Another type of tower is the Lattice tower - this is also known as a self supporting tower.

They typically range in height from 200 to 400 feet

This isn't my area of expertise, but I could imagine this being more useful in rural areas that are somewhat sparsely populated

24/n

It seems like the extra height would be advantageous in helping the cellular radio and antennas operate at longer ranges.

Another type of tower is the guyed tower. These range in height from 300 to 2,000 feet.

Unlike the lattice tower, the scafolding of the guyed tower

25/n

is very light weight. It is held in place by guy wires that are anchored into the ground.

26/n

This brief piece of types of towers makes me aware of a couple of expenses that I ignored.

Engineering services and Site Preparation.

Here is what it will cost someone who understands this stuff to advise you on what you need to do at your site:

27/n

Here is what your network engineers will need to be paid to design a new cell tower.

28/n

This is a look at site preparation costs

29/n

And the costs go on and on.

For example, a 5g cellular radio - which sits at the top of the cell tower, weighs 3,200 lbs.

The only way to get the radio in place is by using a heavy lift helicopter.

Here is the price tag on that service.

30/n

Unfortunately I don't have good data on what it would cost to extend telephone lines to rural locations lacking service - this project was for replacement of existing equipment, not putting new equipment in place.

But when you look at the $250k to $1.2 million price tag
31/n

per cell tower, the network portion of that cost is below:

32/n

put another way, even when phone network extension is a trivial portion of the cost of installing a cell tower, everything else adds quite a bit of cost.

It is unknown how many user terminals that a Kuiper or StarLink solution to connect cellular phones to an satellite
33/n

internet backbone, or what additional costs would be involved in getting the VOIP signal from the Kuiper link to a phone network, but one thing is certain - the costs are much higher than they are for a system that doesn't require a cell tower at all.

And at the end of the
34/n

day, it all comes down to economics.

The economics aren't there to provide broadband to remote locations through terrestrial means because infrastructure and labor is expensive, and by definition, population bases in rural areas are small.
35/n

But the analysis so far has been focused on costs in the United States.

There are a few reasons for that:
1) I live in the United States
2) I found detailed data that I could make inferences from
3) I haven't found detailed data for costs in foreign countries.
36/n

But I have found a way to use *SOME* data as a proxy for costs of cell towers in foreign countries - I have done searches for sales of cell towers in other countries.

It isn't a perfect proxy - a sales price of a cell tower takes into account more than just the cost to
37/n

replace a tower - it also factors in the fact that a cell tower is an asset that generates revenue.

It's imperfect. But I am not being paid for this research, and I believe that even though it is imperfect data, it helps tell the story on a broader scale.
38/n

Here is a link to a company that is going to build 2,000 cell towers in Ethiopia over the next 5 years - the cost - $1.5 billion dollars. That's $750,000 per tower

capitalethiopia.com/capital/ethiop…
39/n

Here is an article about Telefonica selling a company that owns and supports 30,722 cell towers spread throughout Spain, Germany, Brazil, Peru, Chile and Argentina for $9.4 billion - a cost per tower of 305,969.

commsupdate.com/articles/2021/…
40/n

You can find other articles for cell tower sales in other countries through a little googling.

The exact numbers of cost per tower don't matter.

What does matter is that technology that allows mobile phones to work without massive infrastructure costs is the only way to
41/n

connect the unconnected.

The technology that AST SpaceMobile is developing is essential. And if AST doesn't succeed in their mission of connecting the unconnected, someone else needs to.

Its that important.
42/n

@threadreaderapp unroll