LEO Broadband - Crystal Ball

Forums › New Zealand Broadband › LEO Broadband - Crystal Ball

pdh

345 posts

Ultimate Geek

#245477 6-Feb-2019 00:28

Using LEO satellites (1,000 km) for domestic broadband seems to be technically feasible.
It just needs to reduce costs - to become attractive economically.

Firstly commercial traffic - Telesat (Ottawa) is 'targeting' those services for 2022.

They launched a Phase 1 satellite a year ago - offering it for testing by partners.
Other satellites will go up to form a networked constellation.
SpaceX, Google (and I imagine others) have talked of doing something similar.

Advocates laud low-latency, high capacity, redundancy, global coverage... big claims.

Telesat is an outfit with some credibility - and it wouldn't surprise me if they're first out of the gate. They trace their roots back to Bell Labs and Telstar 1 (the world's first Comms satellite) and have a history of solid solutions and 'firsts'. Canada, after all, has a population density that makes NZ look crowded - and couldn't have a deeper history with telecoms ;-)

I noticed that Telesat is suddenly upping their workforce by 50%.

I guess that recent (last 2 years) significant advances by Blue Origin, Rocket Labs and SpaceX - dramatically dropping launch costs - may be a tipping point for the economics of the LEO scheme.

Canadian Governments are now budgeting money for Advanced Rural Broadband services - based on Telesat's LEO system.
As these LEO satellites are not geostationary, they will roam the globe and pass over our heads.

I have no idea how quickly this will enable direct satellite broadband to NZ's household modems or mobile devices - but you can now (several years) use an InReach to send data over the 20+ year-old satphone satellite network. That's still costly - but it's now _very_ old tech...

Anybody see LEO broadband affecting NZ within 5 years ?

Will fibre in the countryside be obsolete in 10 years ?

hio77

12999 posts

Uber Geek

ID Verified

Trusted

Lizard Networks

#2173776 6-Feb-2019 01:26

satellite comms has come down a fair bit in price. @gravityinternet being solid performers in the resi market right now here.

There was the Iridium NEXT network recently going live.

As someone who does purchase services over Iridium, i can certainly say it's even less competitive than the older offers (be it with slightly more bandwidth)

As with anything, fixed installations are a fair bit cheaper monthly. but there are still other factors to consider such as cost to entry.. Who wants to drop a few grand on a satellite when they can moan that fibre should come to their trees (i'll take the fibre plox ;) )

#include <std_disclaimer>

Any comments made are personal opinion and do not reflect directly on the position my current or past employers may have.

pdh

345 posts

Ultimate Geek

#2173779 6-Feb-2019 04:47

As you say, satellites have always been super costly to get into place.
Also costly to build - since you weren't going to be fixing them economically.

So, existing satellite comms are all based on the old costs to launch.

Google tells me that used to be 50-200 M$ (launch only) and that it cost Iridium 5 B$ to launch 95 (52 M$ each).
Note that was 20+ years ago - when a B$ was worth something...

The replacement (last 5-10 years) Iridium Next Satellites cost (to build) was 2.1 B$ for 81 (26 M$ each) and is expected to average less than 10 M$ each to launch. 75 of the new ones are already up - SpaceX got 10 up, on a Falcon 9, just 3 weeks ago.

Note that SpaceX is now talking 2500 $ per pound to orbit.
For a 700 kg satellite, that's down from 52 to 4 M$.

So, to compare with a fresh rural broadband site in NZ, I would expect that to be much cheaper than 26+4 M$ for gear & install - and the throughput probably isn't comparable anyway. But the satellites could (maybe ?) replace many terrestrial sites. Plus some marine cables.

Also, if you were going to scale things up for broadband-service satellites built with 2020 gen tech - you'd be looking to build 100's or 1000's of them and the cost would go down. Cost of failure would also be a lot less. Then you'd do a deal with somebody with a BFR to take them up by the sack full...

I'm anticipating our NZ services to be 'on the back of' international satellite constellations - just as happens with Iridium & GPS. So, the satellites' main income will be while they're passing over North America - anything they earn in the South Pacific would be gravy.

Will the dropping satellite costs of electronics and launch fees cross the rising terrestrial costs of land, concrete, steel & labour anytime soon ? Watch this space, I guess.

quickymart

13983 posts

Uber Geek

ID Verified

#2173785 6-Feb-2019 06:54

pdh:

Will fibre in the countryside be obsolete in 10 years ?

Fibre obsolete in 10 years? I doubt it - but what do you mean by "in the countryside"? Rural fibre?

pdh

345 posts

Ultimate Geek

#2174170 6-Feb-2019 22:41

What I meant by 'in the countryside'...

The economic advantage of any land-based connection is highest for a big apartment building - or in a high-density area like Stonefields in Auckland or Kensington in Orewa - where you've got 700+ residences in very close proximity.

At the other extreme, out in the wops - it's expensive to run fibre/wire from the road up to the house. It may also be expensive just to get the fibre down the road.

The cost of a satellite connection is (I guess) exactly the same in either case - rural or urban - so rural is a lower-hanging fruit.

Why would NZ spend tons of money trying to eradicate the fibre-gap (to the last 10% or 5% of the population) if an effective satellite service is just over the (time) horizon ?

Maybe later there comes a point where cellular service costs more than satellite service and we just go with one big happy service provider for mobile & home - voice & data. Country Telcos would like that to be them - global telcos might want to use the same satellites as they transit all the way round the planet. That would also retire the concept of 'roaming'. As with a sat-phone now, you'd have one contact ID for everything - all countries - anywhere.

Sort of redefining the cloud.

All registered in & billed from the Caymans, I imagine ;-)

nickb800

2723 posts

Uber Geek

Trusted

#2174238 7-Feb-2019 07:13

pdh:

What I meant by 'in the countryside'...

The economic advantage of any land-based connection is highest for a big apartment building - or in a high-density area like Stonefields in Auckland or Kensington in Orewa - where you've got 700+ residences in very close proximity.

At the other extreme, out in the wops - it's expensive to run fibre/wire from the road up to the house. It may also be expensive just to get the fibre down the road.

The cost of a satellite connection is (I guess) exactly the same in either case - rural or urban - so rural is a lower-hanging fruit.

Why would NZ spend tons of money trying to eradicate the fibre-gap (to the last 10% or 5% of the population) if an effective satellite service is just over the (time) horizon ?

Maybe later there comes a point where cellular service costs more than satellite service and we just go with one big happy service provider for mobile & home - voice & data. Country Telcos would like that to be them - global telcos might want to use the same satellites as they transit all the way round the planet. That would also retire the concept of 'roaming'. As with a sat-phone now, you'd have one contact ID for everything - all countries - anywhere.

Sort of redefining the cloud.

All registered in & billed from the Caymans, I imagine ;-)

There's also an opportunity for price discrimination based on location - so satellite phone services to high country station may be reasonably priced to compete with fixed line connections, but there's a higher pricing tier for roaming into other parts of the country or overseas. A bit like we have for mobile data - relatively cheap (per GB) fixed 4g broadband, and relatively expensive 4g mobile data (although 4g is slightly different as we have over and under utilised spectrum in different areas).

All using GPS to figure out where you are and whether you're complying with the location requirements of your chosen package.

wellygary

8345 posts

Uber Geek

#2174274 7-Feb-2019 09:24

pdh:

Why would NZ spend tons of money trying to eradicate the fibre-gap (to the last 10% or 5% of the population) if an effective satellite service is just over the (time) horizon ?

Because unless you have *big* land masses and low pop densities (ie Canada/Australia/Russia/Africa etc,) its cheaper to just keep extending the margins of your existing ground based systems, -Cell based RBI will arguably provide a better and cheaper service for most of the potential "rural" customers,

Granted, there are probably a few thousand outliers in NZ where a sat based service is the most economic, but In a small country like NZ the economics for most users don't really stack up... especially as you really need a ground station or two in your country to avoid packets hopping all around the globe to find a downlink to the "rest of the world internet"

There is a good look at what's likely on offer in the next few years here..

http://www.mit.edu/~portillo/files/Comparison-LEO-IAC-2018-slides.pdf

pdh

345 posts

Ultimate Geek

#2174529 7-Feb-2019 18:08

Many thanks Gary - that's an interesting analysis.

It does look like this will be (initially) designed more for thousand's of ground stations - than for billions.

So not targeted as a direct connection to a hand-held device (or an individual home).

Have you seen any papers discussing the target audience ?