Starlink now open for pre-orders.

social butterfly 2020 · 22-03-2021 05:49PM

corazon wrote: »

SpaceX’s Starlink satellite internet in talks for a place in the UK’s $6.9 billion ‘Project Gigabit’
https://www.cnbc.com/2021/03/22/elon-musks-spacex-starlink-in-talks-with-uks-project-gigabit.html

Why not do the same in Ireland?

Because contracts have already been signed

plodder · 23-03-2021 09:06AM

corazon wrote: »

SpaceX’s Starlink satellite internet in talks for a place in the UK’s $6.9 billion ‘Project Gigabit’
https://www.cnbc.com/2021/03/22/elon-musks-spacex-starlink-in-talks-with-uks-project-gigabit.html

Why not do the same in Ireland?

Fibre broadband is much better than satellites.

social butterfly 2020 · 23-03-2021 09:47AM

plodder wrote: »

Fibre broadband is much better than satellites.

Starlink satellites are at a much lower orbit compared to traditional satellite broadband suppliers.

Latency is lower and speeds of between 250mbps to 400mbps are achievable. Far eclipsing some fibre bb providers.

To compare this to previous satellite bb is void.

Messer1 · 23-03-2021 10:51AM

social butterfly 2020 wrote: »

Because contracts have already been signed

The contracts need changing to take account of disruption caused by Covid-19 and Government's wish to speed up delivery. The Government could avail of this opportunity to encourage market entry by Starlink specifically to target more remote rural areas likely to be waiting years for fibre.

This case is made in a detailed "Plan B for NBP" which I compiled and circulated to ministers etc. last month. Summary at <https://bit.ly/3cyDb1A>.

23-03-2021 11:00AM

social butterfly 2020 wrote: »

Starlink satellites are at a much lower orbit compared to traditional satellite broadband suppliers.

Latency is lower and speeds of between 250mbps to 400mbps are achievable. Far eclipsing some fibre bb providers.

To compare this to previous satellite bb is void.

How is this eclipsing some fibre broadband that is capable up up to 1 gigabit?

Messer1 · 23-03-2021 11:22AM

Deleted User wrote: »

How is this eclipsing some fibre broadband that is capable up up to 1 gigabit?

What type of household needs a Gigabit? For what?

23-03-2021 11:28AM

Messer1 wrote: »

What type of household needs a Gigabit? For what?

They may if they run a business from home but certainly will need it in future if you look at bandwidth usage graphs over the last number of years. Gigabit isn't even the limit on fibre, change the optics either end and you can go to 10G and more, it's in it's infancy

social butterfly 2020 · 23-03-2021 11:39AM

Deleted User wrote: »

How is this eclipsing some fibre broadband that is capable up up to 1 gigabit?

I have vigrin fibre and I'm getting 250mb wireless

23-03-2021 11:42AM

social butterfly 2020 wrote: »

I have vigrin fibre and I'm getting 250mb wireless

Virgin is cable, not fibre! And wireless is your own concern nothing to do with your provider. If you want better wireless buy a better wireless access point and put it in a better location

Messer1 · 23-03-2021 11:44AM

Deleted User wrote: »

They may if they run a business from home but certainly will need it in future if you look at bandwidth usage graphs over the last number of years. Gigabit isn't even the limit on fibre, change the optics either end and you can go to 10G and more, it's in it's infancy

According to Microsoft, a power user of Office needs about 15 Mbps. The FCC recommends 25 Mbps for teleworking or streaming UltraHD 4k vid. The FCC and EU have set 100+ Mbps targets for homes by 2025. Still don't see need for a gigabit for households even running a micro business.

cruizer101 · 23-03-2021 11:45AM

Deleted User wrote: »

They may if they run a business from home but certainly will need it in future if you look at bandwidth usage graphs over the last number of years. Gigabit isn't even the limit on fibre, change the optics either end and you can go to 10G and more, it's in it's infancy

Vast majority of businesses don't need it.
Log on to some cloud based apps, emails, edit documents online, conference calls. No need for Gigabit.

23-03-2021 11:46AM

Messer1 wrote: »

According to Microsoft, a power user of Office needs about 15 Mbps. The FCC recommends 25 Mbps for teleworking or streaming UltraHD 4k vid. The FCC and EU have set 100+ Mbps targets for homes by 2025. Still don't see need for a gigabit for households even running a micro business.

So what! This makes a ftth network even more future proof and a great asset for rural people

BailMeOut · 23-03-2021 11:57AM

is there anyone here that already has this service who can run a jitter test and post results.

The test I'd like to see if a jitter test to a VOIP service and the command to run and let run for 2-3 minutes would be something like this:

C:\>tcping -t -j sipdir.online.lync.com 5061

(tcping can be found here)

Example output from my Eir service is below but would love to see how Starlink compares and if the service is good enough for VOIP.

C:\>tcping -t -j sipdir.online.lync.com 5061

** Pinging continuously. Press control-c to stop **

Probing 52.112.194.11:5061/tcp - Port is open - time=18.058ms
Probing 52.112.194.11:5061/tcp - Port is open - time=15.481ms jitter=-2.577
Probing 52.112.194.11:5061/tcp - Port is open - time=15.180ms jitter=-1.589
Probing 52.112.194.11:5061/tcp - Port is open - time=15.093ms jitter=-1.146
Probing 52.112.194.11:5061/tcp - Port is open - time=14.993ms jitter=-0.960
Probing 52.112.194.11:5061/tcp - Port is open - time=16.150ms jitter=0.389
Probing 52.112.194.11:5061/tcp - Port is open - time=15.399ms jitter=-0.427
Probing 52.112.194.11:5061/tcp - Port is open - time=15.324ms jitter=-0.440
Probing 52.112.194.11:5061/tcp - Port is open - time=15.058ms jitter=-0.652
Probing 52.112.194.11:5061/tcp - Port is open - time=15.822ms jitter=0.184
Probing 52.112.194.11:5061/tcp - Port is open - time=16.138ms jitter=0.482
Probing 52.112.194.11:5061/tcp - Port is open - time=15.245ms jitter=-0.455
Probing 52.112.194.11:5061/tcp - Port is open - time=15.293ms jitter=-0.369
Probing 52.112.194.11:5061/tcp - Port is open - time=15.192ms jitter=-0.442
Probing 52.112.194.11:5061/tcp - Port is open - time=16.275ms jitter=0.673
Probing 52.112.194.11:5061/tcp - Port is open - time=15.138ms jitter=-0.509
Probing 52.112.194.11:5061/tcp - Port is open - time=15.161ms jitter=-0.454
Probing 52.112.194.11:5061/tcp - Port is open - time=20.175ms jitter=4.587
Probing 52.112.194.11:5061/tcp - Port is open - time=15.218ms jitter=-0.625
Probing 52.112.194.11:5061/tcp - Port is open - time=15.163ms jitter=-0.647
Probing 52.112.194.11:5061/tcp - Port is open - time=15.467ms jitter=-0.311
Probing 52.112.194.11:5061/tcp - Port is open - time=15.049ms jitter=-0.713
Probing 52.112.194.11:5061/tcp - Port is open - time=19.113ms jitter=3.383
Probing 52.112.194.11:5061/tcp - Port is open - time=16.962ms jitter=1.085
Probing 52.112.194.11:5061/tcp - Port is open - time=15.061ms jitter=-0.862
Probing 52.112.194.11:5061/tcp - Port is open - time=15.403ms jitter=-0.485
Probing 52.112.194.11:5061/tcp - Port is open - time=15.345ms jitter=-0.525
Probing 52.112.194.11:5061/tcp - Port is open - time=15.258ms jitter=-0.592
Probing 52.112.194.11:5061/tcp - Port is open - time=15.535ms jitter=-0.294

babelfish1990 · 23-03-2021 11:59AM

Messer1 wrote: »

The contracts need changing to take account of disruption caused by Covid-19 and Government's wish to speed up delivery. The Government could avail of this opportunity to encourage market entry by Starlink specifically to target more remote rural areas likely to be waiting years for fibre.

This case is made in a detailed "Plan B for NBP" which I compiled and circulated to ministers etc. last month. Summary at <https://bit.ly/3cyDb1A>.

Messer1 - your Plan B fails to address the capacity needed to meet requirements. Currently, with 1,200 satellites, there is typically 1 Starlink satellite serving Ireland at any time. The capacity of the whole satellite is 20Gbps. If you assume that the average capacity required is 30Mbps per household (1 x 4K TV channel) at peak times, the current satellite could serve around 600 households in Ireland. Even the ships in the seas surrounding Ireland would probably use this capacity alone.

If Starlink increases to 12,000 satellites, they could serve 6,000 households with one TV channel. This represents around 1% of the required households. Even if Starlink goes to 30,000 satellites, and a couple of other LEO constellations come on board, the available capacity is unlikely to exceed 10% of rural households not already served by fibre.

Starlink is absolutely transformational compared to former geo-stationary satellites - and it will be hugely successful. When it becomes popular, the current subsidized tariff plans are likely to be adjusted to discourage heavy users. There will be enough business users willing to pay for small amounts of bandwidth in remote locations to fill up Starlink's capacity. Anyone who wants to use it for HD TV will likely have to pay much higher tariffs. It is likely that tariffs will become very smart - offering cheap rates at off-peak times and over sparsely populated regions such as oceans and deserts. Tariffs over densely populated cities will have to be prohibitive. It is possible that Starlink could have a very small role in a rural broadband scheme - perhaps covering the 1% most remote homes? It is also possible, that it may be a temporary solution, before too many users sign up, particularly if users focus on low bandwidth applications. The Govt rural broadband is very stupid in many ways and way over-priced - particular because they have decided to overlay separate fibre cables where Eir has already laid cables - greatly increasing the cost and delaying the rollout timelines. However, Starlink or other Leos are not a substitute.

Marlow · 23-03-2021 12:06PM

Deleted User wrote: »

Gigabit isn't even the limit on fibre, change the optics either end and you can go to 10G and more, it's in it's infancy

Neither are the current speeds Starlinks limits. The satellites literally use the same technology as fibre uses to talk between them, just that it's faster, as light through a vacuum is faster than light through glas.

Either way, this thread is about Starlink and it's progress. If you want to discuss the NBI, post in the NBI thread.

/M

Marlow · 23-03-2021 12:09PM

babelfish1990 wrote: »

Messer1 - your Plan B fails to address the capacity needed to meet requirements. Currently, with 1,200 satellites, there is typically 1 Starlink satellite serving Ireland at any time. The capacity of the whole satellite is 20Gbps. If you assume that the average capacity required is 30Mbps per household (1 x 4K TV channel) at peak times, the current satellite could serve around 600 households in Ireland. Even the ships in the seas surrounding Ireland would probably use this capacity alone.

The usage per household is vastly smaller. Even today.

Also, for reference: most OpenEIR exchanges currently only have a 10 Gbit/s connection back into the network. That can of course be upgraded, but that gives you a bit of a picture.

/M

babelfish1990 · 23-03-2021 12:29PM

Marlow wrote: »

The usage per household is vastly smaller. Even today.

Also, for reference: most OpenEIR exchanges currently only have a 10 Gbit/s connection back into the network. That can of course be upgraded, but that gives you a bit of a picture.

/M

Perhaps - but we are rapidly moving towards a model where the vast majority of TV viewers will view TV online, instead of off-air, and content is migrating towards the 4K standard. It is reasonable to assume that within 10 years, the average household demand is likely to exceed 30Mbps at peak viewing times - ie one 4K TV channel.

Messer1 · 23-03-2021 12:32PM

babelfish1990 wrote: »

Messer1 - your Plan B fails to address the capacity needed to meet requirements. Currently, with 1,200 satellites, there is typically 1 Starlink satellite serving Ireland at any time. The capacity of the whole satellite is 20Gbps. If you assume that the average capacity required is 30Mbps per household (1 x 4K TV channel) at peak times, the current satellite could serve around 600 households in Ireland. Even the ships in the seas surrounding Ireland would probably use this capacity alone.

If Starlink increases to 12,000 satellites, they could serve 6,000 households with one TV channel. This represents around 1% of the required households. Even if Starlink goes to 30,000 satellites, and a couple of other LEO constellations come on board, the available capacity is unlikely to exceed 10% of rural households not already served by fibre.

As I write there are 3-4 sats over Ireland (out of a thousand deployed). Fortunately for us 53 degrees seems to be a sweet spot based on current launch inclinations. I have seen reports that about 40 sats will be available to serve a particular location once the constellation fills out we don't know what capacity really is now (hence beta), or what it will be as more sats are launched and tech changes (ISLL etc.).

Messer1 · 23-03-2021 12:39PM

Marlow wrote: »

Neither are the current speeds Starlinks limits. The satellites literally use the same technology as fibre uses to talk between them, just that it's faster, as light through a vacuum is faster than light through glas.

Either way, this thread is about Starlink and it's progress. If you want to discuss the NBI, post in the NBI thread.

/M

Yes, once inter-satellite laser links are operational, LEO sats will, according to simulations, prove faster than terrestrial fibre for distances of over c.3k kms. All Starlink sats to be launched after this year will incorporate IISL as will OneWeb and Telesat from the outset (though they may not be offering strictly consumer services). Most likely for Amazon also when it launches.

Messer1 · 23-03-2021 12:44PM

babelfish1990 wrote: »

Perhaps - but we are rapidly moving towards a model where the vast majority of TV viewers will view TV online, instead of off-air, and content is migrating towards the 4K standard. It is reasonable to assume that within 10 years, the average household demand is likely to exceed 30Mbps at peak viewing times - ie one 4K TV channel.

You dont need low-latency LEO sats to stream vids. Very high throughput geo sats could be used where latency isn't such an issue.

KildareP · 23-03-2021 01:13PM

It's disappointing that these threads always descend into a "NBP, but Starlink" discussion. Starlink is still a very early beta system.

It has "more than 10,000" subscribers globally but we aren't given any more information than that. To put that into perspective, just 1% of the premises in the NBP intervention is 5,000 premises alone.

In a few years we will have a better idea of what Starlink is truly capable of in the real world and no doubt can review as the harder-to-reach areas come into play.

However proposing we reopen NBP contracts and carve off areas to Starlink today here and now, based on promising (but as yet unproven*) technology and some impressive speedtests on a system with practically no load, has no basis for argument...

* What is the real-world benefit of FSO when both ends of the connection will be microwave RF?

babelfish1990 · 23-03-2021 01:17PM

Messer1 wrote: »

As I write there are 3-4 sats over Ireland (out of a thousand deployed). Fortunately for us 53 degrees seems to be a sweet spot based on current launch inclinations. I have seen reports that about 40 sats will be available to serve a particular location once the constellation fills out we don't know what capacity really is now (hence beta), or what it will be as more sats are launched and tech changes (ISLL etc.).

Starlink submitted their specs to FCC for US licence - they stated that capacity of each satellite is circa 20Gbps - so we do know the total capacity available per satellite.

The performance for individual users will improve as the number of satellites increase, but I have allowed for this in my calculation. It is unlikely that the total capacity of each satellite can easily be increased, as nobody can swap-out the hardware in the sky! There will be other improvements - such as the FSO laser links, and a ground-station in Ireland, which will support local peering and reduce latency. This won't increase overall capacity per satellite.

Occasionally more than 1 satellite appears over Ireland - but on average there is only one serving us at the moment. Also - remember that typical rural areas not already served by fibre are in mountainous and wooded areas with tall trees providing shelter-belts, blocking some of the satellites. This will improve with more satellites - but the capacity is still tiny by comparison with the requirement.

This technology is really no different to a mobile base-station in the sky, except that the base-stations are moving and the users are stationary. We know that Ireland's three mobile operators already have thousands of base-stations - orders of magnitude more than Starlink plans, - and they cannot reliably provide high-bandwidth services such as TV - hence the need for fibre.

Starlink is amazing technology - but you need to do your sums.

Messer1 · 23-03-2021 01:27PM

KildareP wrote: »

It's disappointing that these threads always descend into a "NBP, but Starlink" discussion. Starlink is still a very early beta system.

It has "more than 10,000" subscribers globally but we aren't given any more information than that. To put that into perspective, just 1% of the premises in the NBP intervention is 5,000 premises alone.

In a few years we will have a better idea of what Starlink is truly capable of in the real world and no doubt can review as the harder-to-reach areas come into play.

However proposing we reopen NBP contracts and carve off areas to Starlink today here and now, based on promising (but as yet unproven*) technology and some impressive speedtests on a system with practically no load, has no basis for argument...

* What is the real-world benefit of FSO when both ends of the connection will be microwave RF?

The discussion should be about NBP+Starlink. It will take a few years for Starlink to grow and bed in just as it will take till 2026 ish to finish NBP. There is no need to carve off areas specifically for Starlink as homes starved of BB at the moment can, if they wish, subscribe to it shortly and when fibre eventually arrives either stick or switch.

Messer1 · 23-03-2021 02:11PM

babelfish1990 wrote: »

Starlink submitted their specs to FCC for US licence - they stated that capacity of each satellite is circa 20Gbps - so we do know the total capacity available per satellite.

The performance for individual users will improve as the number of satellites increase, but I have allowed for this in my calculation. It is unlikely that the total capacity of each satellite can easily be increased, as nobody can swap-out the hardware in the sky! There will be other improvements - such as the FSO laser links, and a ground-station in Ireland, which will support local peering and reduce latency. This won't increase overall capacity per satellite.

Occasionally more than 1 satellite appears over Ireland - but on average there is only one serving us at the moment. Also - remember that typical rural areas not already served by fibre are in mountainous and wooded areas with tall trees providing shelter-belts, blocking some of the satellites. This will improve with more satellites - but the capacity is still tiny by comparison with the requirement.

This technology is really no different to a mobile base-station in the sky, except that the base-stations are moving and the users are stationary. We know that Ireland's three mobile operators already have thousands of base-stations - orders of magnitude more than Starlink plans, - and they cannot reliably provide high-bandwidth services such as TV - hence the need for fibre.

Starlink is amazing technology - but you need to do your sums.

First ground station? https://www.irishexaminer.com/news/arid-40232972.html
Others if needed could be located at well-connected data centres around Dublin. As regards satellites, 90% have still to be launched and they will most likely will be very superior to V1s which have a life of about 5 years.

<https://satellitemap.space/> is a good place to watch sat movements. Co-incidentally a train of 60 have just now passed over Ireland on their way to operational orbits.

47akak · 23-03-2021 05:43PM

babelfish1990 wrote: »

Starlink submitted their specs to FCC for US licence - they stated that capacity of each satellite is circa 20Gbps - so we do know the total capacity available per satellite.

...as nobody can swap-out the hardware in the sky! This won't increase overall capacity per satellite.

Starlink is amazing technology - but you need to do your sums.

There's more than 1 satellite. The vast majority apparently have "4x" the capacity of the first generation's 20gbps. So up to 80 Gbps.

joe123 · 23-03-2021 07:14PM

Messer1 you will get nowhere if you continue to ignore and/or disregard opposing opinions or arguments.

I'm keeping an eye on Starlink progress but a quick scan through comments on reddit suggest this is far from being feasible for some so far. And again, Musk himself said this is not for your typical consumer.

So to continue to push for contract changes and a "plan B" for a technology that hasn't been properly tested yet is madness imo.

What happens if you got your way? It would result in more delays, more fluting about the place and potentially it could transpire that Starlink would not be able to supply 500k houses with sufficient speeds.

babelfish1990 · 23-03-2021 08:08PM

Messer1 wrote: »

You dont need low-latency LEO sats to stream vids. Very high throughput geo sats could be used where latency isn't such an issue.

The suggestion that geo-stationary satellites could be used for on-line TV distribution is nonsense. Geo satellites can only be used to beam large numbers of TV channels when vast numbers of viewers are watching the same content at the same time. That model no longer meets the requirements of most consumers, who want to watch content in their own time, hence requiring separate bandwidth for each consumer. The latency issue is irrelevant for TV.

The train of 60 Starlink satellites occasionally over Ireland are no use for providing service - when they reach their final operating orbits they will be well dispersed.

I have nothing against Starlink. I don't accept the argument that it is an unproven technology - the concept of LEO has been around for a long time although it lacked implementation on scale until now. The engineering principles are well understood, and the Beta to date has given excellent results. I think it is wonderful technology, and may play a tiny role in extreme rural broadband provision - probably around 1% of households not already served by fibre (maybe 5,000 households). However, there is a real danger that people who don't understand the details will be deluded into thinking this is a substitute for scaled delivery of video streaming on fibre, or that it is a cheaper solution - which it is not. Starlink don't need to go after the mass market. It is much more profitable for them to go after niche markets that are desperate for small amounts of bandwidth in remote locations. Businesses, aircraft, ships, oil-rigs, weather buoys, telemetry from utility infrastructure etc will all be prepared to pay dearly for ubiquitous service, and will generally have low requirements for bandwidth compared to consumer users.

Messer1 · 23-03-2021 08:35PM

joe123 wrote: »

Messer1 you will get nowhere if you continue to ignore and/or disregard opposing opinions or arguments.

I'm keeping an eye on Starlink progress but a quick scan through comments on reddit suggest this is far from being feasible for some so far. And again, Musk himself said this is not for your typical consumer.

So to continue to push for contract changes and a "plan B" for a technology that hasn't been properly tested yet is madness imo.

What happens if you got your way? It would result in more delays, more fluting about the place and potentially it could transpire that Starlink would not be able to supply 500k houses with sufficient speeds.

My Plan B NEVER advocated that the NBP be delayed in any way or suggested that Starlink should supply 500k houses.

47akak · 24-03-2021 12:33AM

Obviously it's a more complicated subject than simple multiplication would suggest but I think people would be surprised at how little the aggregate bandwidth consumption in Ireland even at peak ads up to vs what you would expect it to taking for example the Netflix HD rate and guessing how many people might be watching at any one time. The total bandwidth available for a full Starlink constellation (even the % over Ireland at any one time) would compare favourably with that and might make it seem like it's feasible to use it for a significant percentage of NBP premises but for a few different reasons even very best case scenario it's unlikely Starlink could ever serve more than 100k premises in Ireland (it's probably half that or less), peaking at that number in 7 years, around the same time NBP will complete 600k premises. Those 100k will be paying double anyone selling NBI wholesale will charge too.

Starlink is not designed for, nor intended to be, a competitor to fixed-line broadband operators. For a start those operators don't need to replace their whole network every 4 years like Starlink. Starlink is an ingenious system and exciting for those who need and afford it but even a cursory run through the numbers makes it clear it's not feasible for a large scale rollout. I can't find the number of houses which won't be connected by NBI but it's only those houses which should be looked at for LEO internet access, a subsidy perhaps.

Marlow · 24-03-2021 02:31AM

47akak wrote: »

Starlink is not designed for, nor intended to be, a competitor to fixed-line broadband operators.

Elon made it very clear, that Urban areas should rely on FTTH or 5G. It's mostly a solution for rural or ad-hoc connectivity. A very powerful at that.

And it's a serious backup for everything else.

And yes ... stating peak usage is nonsense. Aggregate usage is a whole different kettle of fish, but most consumers won't know what that actually entails.

/M

Starlink now open for pre-orders.

Comments