SpaceX is one of the RDOF Auction Winners

[volker2020]

According to telecompetitor.com they won $885.5 million for 642,000 locations, making them one of the major receivers.

In the USA, RDOF = Rural Digital Opportunity Fund

edit/zubenelgenubi

[gongora]

https://www.fcc.gov/document/auction-904-winning-bidders

[gongora]

SpaceX amounts (by state).  Reminder this is over a 10 year period.

[rockets4life97]

Can someone explain the significance of this? Thanks in advance.

[gongora]

There were 180 winning bidders in the auction, with the 10-year support amount totaling $9.23 billion and covering
5,220,833 locations in 49 states and one territory. Of the 5,295,771 locations in the 61,766 eligible census block groups, approximately 99% of the locations are covered by winning bids. While winning bids are for a range of performance tiers, winning bids for downstream speeds of at least 100 megabits per second (Mbps) cover 99.7% of these locations, with over 85% of locations covered by winning bids for Gigabit speed service.

So obviously they counted Starlink as at least 100Mbps service.  I doubt they counted it as gigabit, so Starlink (with 12% of locations) would be one of the few applicants to win with less than gigabit service.
SpaceX totals: $885,509,638.40 for 642,925 locations

[gongora]

Can someone explain the significance of this? Thanks in advance.

It's a program to subsidize broadband service to underserved areas of the U.S. over the next 10 years.  Just a few months ago it was unclear if SpaceX would even qualify to participate.  Now they have the opportunity to get payments from the government to subsidize service to the listed locations.  It looks like companies offering geostationary satellite service basically got nothing from the auction.

[gongora]

Results Map: https://www.fcc.gov/reports-research/maps/rdof-phase-i-dec-2020/

SpaceX is one of the few companies in the "Above Baseline" (100Mbps) / low latency tier, so it's easy to see from the colors where they won.

[volker2020]

Can someone explain the significance of this? Thanks in advance.

Basically they got federal grants to further develop there starlink constellation. While there already was a huge demand from investors, it will make the start of the starlink constellation much smoother.

It means that they have convinced the deciders, that they could offer a stable 100 MBit connection, which is no small accomplishment. Starlink is the money cow, that could make colonizing mars possible, and this gives them a solid lead to other space companies playing with the idea of setting up there own mega constellations. Since at least with the second generation, each satellite could be used world wide, this gives them further advantages.

[gongora]

Lots of money left over for the second round in a few years, which will be welcomed by any other LEO constellations in or near service at the time.

Competitive bidding brought the auction in significantly under budget, allocating $9.2 billion in support out of the $16 billion set aside for the Phase I auction. The $6.8 billion in potential Phase I support that was not allocated will be rolled over into the future Phase II auction, which now can draw upon a budget of up to $11.2 billion in targeting partially-served areas and the few unserved areas that did not receive funding through Phase I.

All of the bids can be found here: https://auctiondata.fcc.gov/public/projects/auction904

(For some reason Excel is having issues opening an 890MB .csv file on my computer)

[oldAtlas_Eguy]

The basic implications is that the FCC does not consider there to be much question as to whether technically Starlink can deliver the specified service. But that they can deliver it at a competitive price in these rural areas.

Although there is a question.
What is the payout profile by year such as up front loading to cover "infrastructure emplacement"?

I imagine it is heavily loaded on the front end and tapering off to nearly nothing by year 10.

[wannamoonbase]

Seems like a good source of revenue that can be driven into customer antenna cost reduction.

[niwax]

The subsidies appear to be anywhere from $500 to $2000 per customer. Does anyone know why they would have different amounts for different places? Or are they simply bidding higher where there's the least competition?

Anyway, it would at least cover the initial charge for the antenna. That makes it quite the attractive package even at current pricing.

[gongora]

The subsidies appear to be anywhere from $500 to $2000 per customer. Does anyone know why they would have different amounts for different places? Or are they simply bidding higher where there's the least competition?

There were multiple rounds of bidding.  The winning bids would depend on the competition in the area.

[abaddon]

Lots of money left over for the second round in a few years, which will be welcomed by any other LEO constellations in or near service at the time.

Are first-round winners allowed to bid next time around?

[Robotbeat]

Lots of money left over for the second round in a few years, which will be welcomed by any other LEO constellations in or near service at the time.

Competitive bidding brought the auction in significantly under budget, allocating $9.2 billion in support out of the $16 billion set aside for the Phase I auction. The $6.8 billion in potential Phase I support that was not allocated will be rolled over into the future Phase II auction, which now can draw upon a budget of up to $11.2 billion in targeting partially-served areas and the few unserved areas that did not receive funding through Phase I.

All of the bids can be found here: https://auctiondata.fcc.gov/public/projects/auction904

(For some reason Excel is having issues opening an 890MB .csv file on my computer)

Good to hear there’s still plenty for other LEO megaconstllations. Competition is good, and this should also help the development of RLVs generally.

[abaddon]

The subsidies appear to be anywhere from $500 to $2000 per customer. Does anyone know why they would have different amounts for different places? Or are they simply bidding higher where there's the least competition?

There were multiple rounds of bidding.  The winning bids would depend on the competition in the area.

Right.  And remember, the other option here is fiber, and unlike Starlink [or other satellite technology] those costs vary substantially based on location.

[oldAtlas_Eguy]

The amount of subsidy for an area probably has several factors. Number of potential subscribers in an area. Remoteness of the area (distance from any broadband services as in cable). The amount for infrastructure and the amount for monthly service cost reduction. Each area could have very significantly different subsidies payout profiles and total amounts.

SpaceX infrastructure costs directly related to an area would be Gateway installations and the "pizza on a stick" terminals.

[DreamyPickle]

Subsidy seems to be 1377$ per location. If dishy is indeed 2600$ then the subsidy covers more than half with SpaceX and the customers splitting the rest.

[Nomadd]

Subsidy seems to be 1377$ per location. If dishy is indeed 2600$ then the subsidy covers more than half with SpaceX and the customers splitting the rest.

And if the dish is 35 cents, they'll make a bundle. But it won't be.

[ncb1397]

Can someone explain the significance of this? Thanks in advance.

Depends on how you look at it. The glass half full perspective from the SpaceX point of view, they are one of the largest recipients (4th overall). On the other hand, a handful of their competitors got significantly more to expand coverage. For instance, LTD broadband which currently offers fixed wireless coverage in parts of Minnesota, South Dakota, Iowa, Nebraska and Wisconsin got ~$1.3 billion dollars to improve their service and expand coverage into Texas, Oklahoma, Ohio, North Dakota, Missouri, Kansas, Indiana, Illinois, Colorado and California. So, this one company went from a competitor in 5 states to competing in 15 states (and they offer plans significantly below $100 per month). So, it is a mixed bag, which was probably why SpaceX was pretty lukewarm and/or hostile to these types of programs prior to this.

“SpaceX believes that it is more effective to leverage advanced technology and smart private sector infrastructure investment to reach America’s unserved and underserved population, rather than seek Government subsidization for this effort,” SpaceX’s Vice President of Satellite Government Affairs, Patricia Cooper, wrote in a May 8 letter to FCC Chairman Ajit Pai.

Cooper thanked the FCC for revising the Connect America auction rules, but said systems like Starlink won’t need government funding to connect rural and other remote areas.

https://spacenews.com/spacex-wont-seek-u-s-rural-broadband-subsidies-for-starlink-constellation/

[oldAtlas_Eguy]

That subsidy amount of $885M at $1337 per customer. Results in a subscriber base for those areas of 661,000. That is at $99 per month, a yearly revenue from all of these areas of $786M.

[Asteroza]

That subsidy amount of $885M at $1337 per customer. Results in a subscriber base for those areas of 661,000. That is at $99 per month, a yearly revenue from all of these areas of $786M.

If this is Elon's doing, that $1337 price is not a coincidence...

[gongora]

It's $1377, divided over 10 years

[gongora]

Although there is a question.
What is the payout profile by year such as up front loading to cover "infrastructure emplacement"?

I imagine it is heavily loaded on the front end and tapering off to nearly nothing by year 10.

It's not front loaded but it starts immediately when the paperwork is done (which could still be a few months), so the companies can use some of the money before service starts.

[oldAtlas_Eguy]

Although there is a question.
What is the payout profile by year such as up front loading to cover "infrastructure emplacement"?

I imagine it is heavily loaded on the front end and tapering off to nearly nothing by year 10.

It's not front loaded but it starts immediately when the paperwork is done (which could still be a few months), so the companies can use some of the money before service starts.

Thanks.

So 2021 will have a revenue income for Starlink just from the subsidies of $88.5M. At some point highly likely in mid/late 2021 once the number of sats reach that majical number somewhere around 1400 then initial operational services would start. Which also starts increasing the revenue coming in for Starlink.

[freddo411]

Although there is a question.
What is the payout profile by year such as up front loading to cover "infrastructure emplacement"?

I imagine it is heavily loaded on the front end and tapering off to nearly nothing by year 10.

It's not front loaded but it starts immediately when the paperwork is done (which could still be a few months), so the companies can use some of the money before service starts.

Thanks.

So 2021 will have a revenue income for Starlink just from the subsidies of $88.5M. At some point highly likely in mid/late 2021 once the number of sats reach that majical number somewhere around 1400 then initial operational services would start. Which also starts increasing the revenue coming in for Starlink.

Slight correction. I’m writing checks to SpaceX right now for using star link

[oldAtlas_Eguy]

Although there is a question.
What is the payout profile by year such as up front loading to cover "infrastructure emplacement"?

I imagine it is heavily loaded on the front end and tapering off to nearly nothing by year 10.

It's not front loaded but it starts immediately when the paperwork is done (which could still be a few months), so the companies can use some of the money before service starts.

Thanks.

So 2021 will have a revenue income for Starlink just from the subsidies of $88.5M. At some point highly likely in mid/late 2021 once the number of sats reach that majical number somewhere around 1400 then initial operational services would start. Which also starts increasing the revenue coming in for Starlink.

Slight correction. I’m writing checks to SpaceX right now for using star link

How many are in the Beta and how much $ per month during Beta?

[freddo411]

Although there is a question.
What is the payout profile by year such as up front loading to cover "infrastructure emplacement"?

I imagine it is heavily loaded on the front end and tapering off to nearly nothing by year 10.

It's not front loaded but it starts immediately when the paperwork is done (which could still be a few months), so the companies can use some of the money before service starts.

Thanks.

So 2021 will have a revenue income for Starlink just from the subsidies of $88.5M. At some point highly likely in mid/late 2021 once the number of sats reach that majical number somewhere around 1400 then initial operational services would start. Which also starts increasing the revenue coming in for Starlink.

Slight correction. I’m writing checks to SpaceX right now for using star link

How many are in the Beta and how much $ per month during Beta?

Don’t know

99 a month

[oldAtlas_Eguy]

Although there is a question.
What is the payout profile by year such as up front loading to cover "infrastructure emplacement"?

I imagine it is heavily loaded on the front end and tapering off to nearly nothing by year 10.

It's not front loaded but it starts immediately when the paperwork is done (which could still be a few months), so the companies can use some of the money before service starts.

Thanks.

So 2021 will have a revenue income for Starlink just from the subsidies of $88.5M. At some point highly likely in mid/late 2021 once the number of sats reach that majical number somewhere around 1400 then initial operational services would start. Which also starts increasing the revenue coming in for Starlink.

Slight correction. I’m writing checks to SpaceX right now for using star link

How many are in the Beta and how much $ per month during Beta?

Don’t know

99 a month

~$12M a year(12 months) for each 10,000 subscriptions in the Beta or full operations.

[su27k]

This may be just sour grapes, but this article argues there's unusually large amount of fixed wireless bidding on gigabit tier with very low cost in this auction, which they may not be able to achieve in reality. This may explain why Starlink didn't get a bigger piece of the pie.

RDOF Auction Ends but Confusion and Corruption May Just Be Beginning

[gongora]

This may be just sour grapes, but this article argues there's unusually large amount of fixed wireless bidding on gigabit tier with very low cost in this auction, which they may not be able to achieve in reality. This may explain why Starlink didn't get a bigger piece of the pie.

That "article" is a blog post by a lobbying group.  It also shows that a lot of the top winners bid fiber, not wireless.

[su27k]

Some additional grumbling about RDOF results, I guess they're hoping the incoming FCC chief can change it somehow:

Incoming FCC chief could inherit RDOF boondoggle

FCC’s Rural Digital Opportunity Fund Auction Was Supposed to Significantly Reduce America’s Rural Broadband Gap

[RedLineTrain]

Some additional grumbling about RDOF results, I guess they're hoping the incoming FCC chief can change it somehow:

Incoming FCC chief could inherit RDOF boondoggle

FCC’s Rural Digital Opportunity Fund Auction Was Supposed to Significantly Reduce America’s Rural Broadband Gap

These are pretty bad arguments.  But they should be careful what they wish for.  Any redo of the auction probably would benefit SpaceX greatly.

[su27k]

Tim Farrar is also harping on his new (conspiracy) theory that Starlink couldn't support the 675k locations they won from RDOF (and hoping Starlink will fail in the long form stage), here's just a simple back of envelope math to show he's wrong:

1. Data cap: minimal of 2TB per month is required for RDOF, which is 8*2,000,000 Mb/month = 16,000,000 Mb/(30*24*3600s) = 6.2 Mbps, so as long as Starlink uses an oversubscription ratio less than 100/6.2 = 16, they don't need to worry about data cap.

2. Bandwidth: 100Mbps with oversubscription ratio of 16 for 675k locations, from Robotbeat's post here "total capacity in the US is about 1/25th the total capacity of the network.", assuming 20Gbps per satellite, so how many satellites do we need:

x * 0.04 * 20,000 Mbps * 16 / 675,000 = 100 Mbps

x = 5,273, very doable in 10 years.

They'll probably want to use an oversubscription ratio larger than 16x, which means the RDOF data cap requirement is actually a bigger constraint than bandwidth.

[M.E.T.]

Tim Farrar is also harping on his new (conspiracy) theory that Starlink couldn't support the 675k locations they won from RDOF (and hoping Starlink will fail in the long form stage), here's just a simple back of envelope math to show he's wrong:

1. Data cap: minimal of 2TB per month is required for RDOF, which is 8*2,000,000 Mb/month = 16,000,000 Mb/(30*24*3600s) = 6.2 Mbps, so as long as Starlink uses an oversubscription ratio less than 100/6.2 = 16, they don't need to worry about data cap.

2. Bandwidth: 100Mbps with oversubscription ratio of 16 for 675k locations, from Robotbeat's post here "total capacity in the US is about 1/25th the total capacity of the network.", assuming 20Gbps per satellite, so how many satellites do we need:

x * 0.04 * 20,000 Mbps * 16 / 675,000 = 100 Mbps

x = 5,273, very doable in 10 years.

They'll probably want to use an oversubscription ratio larger than 16x, which means the RDOF data cap requirement is actually a bigger constraint than bandwidth.

Just build more powerful satellites as V1.1.

[oldAtlas_Eguy]

Tim Farrar is also harping on his new (conspiracy) theory that Starlink couldn't support the 675k locations they won from RDOF (and hoping Starlink will fail in the long form stage), here's just a simple back of envelope math to show he's wrong:

1. Data cap: minimal of 2TB per month is required for RDOF, which is 8*2,000,000 Mb/month = 16,000,000 Mb/(30*24*3600s) = 6.2 Mbps, so as long as Starlink uses an oversubscription ratio less than 100/6.2 = 16, they don't need to worry about data cap.

2. Bandwidth: 100Mbps with oversubscription ratio of 16 for 675k locations, from Robotbeat's post here "total capacity in the US is about 1/25th the total capacity of the network.", assuming 20Gbps per satellite, so how many satellites do we need:

x * 0.04 * 20,000 Mbps * 16 / 675,000 = 100 Mbps

x = 5,273, very doable in 10 years.

They'll probably want to use an oversubscription ratio larger than 16x, which means the RDOF data cap requirement is actually a bigger constraint than bandwidth.

Just build more powerful satellites as V1.1.

NOTE: That once Starship starts launching Starlink sats the V2.0's will be deployed. The second generation sat is likely to be heavier but not that much. Produce possibly 4X the power and at least 4X the throughput per sat by more spot beams and 4X frequency reuse. Increased onboard computing and Internet routing. Possibly even some level of onboard data caching or increased packet buffer retransmit capability lowering the overhead for handling failed packet transmissions to clients. Basically a very significant capability jump. Probable introduction of around 2023. So any arguments that say that Starlink will not have the capacity are definite bunk.

The only thing that can hold back capacity expansion is lack of funds.

[Faerwald]

Well that article is crap... In the same breath it say:
A. Starlink shouldn't get money as it will be built anyway, and ,
B. Starlink shouldn't get any money because it has no paying subscribers?

Talking about wanting it both ways..

Snip

FCC’s Rural Digital Opportunity Fund Auction Was Supposed to Significantly Reduce America’s Rural Broadband Gap

These are pretty bad arguments.  But they should be careful what they wish for.  Any redo of the auction probably would benefit SpaceX greatly.

[Hauerg]

Well that article is crap... In the same breath it say:
A. Starlink shouldn't get money as it will be built anyway, and ,
B. Starlink shouldn't get any money because it has no paying subscribers?

Talking about wanting it both ways..

Snip

FCC’s Rural Digital Opportunity Fund Auction Was Supposed to Significantly Reduce America’s Rural Broadband Gap

These are pretty bad arguments.  But they should be careful what they wish for.  Any redo of the auction probably would benefit SpaceX greatly.

Basically the article is complaining that SpaceX has an anfair advantage because its technology is better fit to solve the problem.

[Robotbeat]

The whole point of RDOF is to help enable providers to have the resources to expand coverage. It’s absurd to say at this point that SpaceX should be denied RDOF because they need it, even tho the basic tech is now proven.

[vsatman]

NOTE: That once Starship starts launching Starlink sats the V2.0's will be deployed. The second generation sat is likely to be heavier but not that much. Produce possibly 4X the power and at least 4X the throughput per sat by more spot beams and 4X frequency reuse.

Frequency reuse cannot help here
You have only 4000 MHz in Ka band between Sat and GW . To increase this you have to change in V band  with 10000 MHz 37,5..42,5 GHZ or work with 2 or more GateWAys in Ka...

[Robotbeat]

NOTE: That once Starship starts launching Starlink sats the V2.0's will be deployed. The second generation sat is likely to be heavier but not that much. Produce possibly 4X the power and at least 4X the throughput per sat by more spot beams and 4X frequency reuse.

Frequency reuse cannot help here
You have only 4000 MHz in Ka band between Sat and GW . To increase this you have to change in V band  with 10000 MHz 37,5..42,5 GHZ or work with 2 or more GateWAys in Ka...

wrong! Yes it does! You can use smaller and more numerous spot beams with larger and more capable phased array antenna.

[su27k]

Yeah they really really want to claw it back somehow: Elon Musk’s SpaceX Riles Its Rivals for Broadband Subsidies - Competitors say government should think twice about backing SpaceX’s satellite-based system with nearly $1 billion

quote from reddit:

Rivals of SpaceX for subsidy dollars are calling on the FCC and its new leadership under the Biden administration to give those plans a closer look, and they are drumming up support for their cause on Capitol Hill.

More than 150 members of Congress wrote the FCC on Jan. 19 urging it “to thoroughly vet the winning bidders to ensure that they are capable” and to “consider opportunities for public input on the applications.”

The letter, which didn’t mention SpaceX or other companies by name, was subsequently promoted online by two trade groups that have competed for the federal subsidies: the National Rural Electric Cooperative Association and the Rural Broadband Association.

“We are in effect funding an experiment here,” said Jim Matheson, chief executive of the National Rural Electric Cooperative Association, which represents electricity providers also in line for subsidies to build out fiber-optic broadband networks. “We don’t know if it works or doesn’t work,” he said in an interview, referring to the SpaceX system.

[Lar]

Yeah they really really want to claw it back somehow: Elon Musk’s SpaceX Riles Its Rivals for Broadband Subsidies - Competitors say government should think twice about backing SpaceX’s satellite-based system with nearly $1 billion

quote from reddit:

Rivals of SpaceX for subsidy dollars are calling on the FCC and its new leadership under the Biden administration to give those plans a closer look, and they are drumming up support for their cause on Capitol Hill.

More than 150 members of Congress wrote the FCC on Jan. 19 urging it “to thoroughly vet the winning bidders to ensure that they are capable” and to “consider opportunities for public input on the applications.”

The letter, which didn’t mention SpaceX or other companies by name, was subsequently promoted online by two trade groups that have competed for the federal subsidies: the National Rural Electric Cooperative Association and the Rural Broadband Association.

“We are in effect funding an experiment here,” said Jim Matheson, chief executive of the National Rural Electric Cooperative Association, which represents electricity providers also in line for subsidies to build out fiber-optic broadband networks. “We don’t know if it works or doesn’t work,” he said in an interview, referring to the SpaceX system.

Betting against SpaceX is foolhardy. Unless you aren't competing on merits and think you have more clout than they do.

[docmordrid]

From the reddit quote,

“We are in effect funding an experiment here,” said Jim Matheson, chief executive of the National Rural Electric Cooperative Association,...“We don’t know if it works or doesn’t work,” he said in an interview, referring to the SpaceX system.

Says no one who has seen the YouTube videos or news reports, or talked to WA state officials, native people's who now have it, beta testers, etc.

Reeks of desperation.

[aero]

Seems to me that SpaceX is going to put Starlink up, with or without the RDOF money. They will have it up and operational well before the competition gets their systems in place which will, in the end, will pretty much eliminate the market for the competition.

[su27k]

Viasat wants FCC to review Starlink’s government funding

Satellite operator Viasat is stepping up efforts to stop Starlink’s growing constellation, taking aim at the nearly $900 million of rural broadband subsidies that SpaceX won in December.

The operator is asking the Federal Communications Commission to review decisions made around the Rural Digital Opportunity Fund (RDOF), claiming differential treatment and a lack of transparency.

In an Application for Review (AFR) filed June 1, Viasat calls on the regulator to probe a series of decisions related to bidding to provide low-latency internet service under RDOF’s Phase 1, also known as Auction 904.

[docmordrid]

ISTM Viasat thinks they're in a cage death match (they aren't wrong).

[M.E.T.]

How long can Viasat survive once Starlink is operational?

Just trying to adjust my expectations as to how long we’ll still have to endure this cynical, litigation-based rearguard action before Viasat goes quietly into the night.

[electricdawn]

You know, Viasat could've just took the punch in the gut, roll with it and then get on with developing something better.

Instead, they decided to go the way of the lawyer. IMHO, not a good move in this case.

Compete on your merit as a technical company, not on how expensive your lawyer is.

[jpo234]

How long can Viasat survive once Starlink is operational?

Just trying to adjust my expectations as to how long we’ll still have to endure this cynical, litigation-based rearguard action before Viasat goes quietly into the night.

Imagine Nokia petitioning the FCC in 2007 to block the iPhone because it would disrupt their business. Companies should compete on the merit of their offerings.

[RedLineTrain]

To the merits of Viasat's arguments, they aren't very strong.  Their LEO constellation was only proposed.  They didn't have the regulatory approval for the constellation.  Also, they said the only reason for the LEO constellation was the subsidies.

Contrast to SpaceX, which was in the middle of a constellation launch campaign, was able to provide testing showing the low latency to a skeptical FCC, and had regulatory approval for a full constellation.

In my opinion, the disparate treatment, such as it existed, was warranted.  But in the end, all of these arguments by Viasat and others such as the rural electricity cooperatives hinge on whether SpaceX can deliver this constellation.  So far, SpaceX has shown it is likely to do so in short order.  And so those arguments will fall away as time goes on.

[gongora]

How long can Viasat survive once Starlink is operational?

Just trying to adjust my expectations as to how long we’ll still have to endure this cynical, litigation-based rearguard action before Viasat goes quietly into the night.

Viasat has multiple business lines and Starlink isn't going to have the capacity to completely replace all of their competitors.

[oldAtlas_Eguy]

This current attack  (strategy) by Viasat will very unlikely do anything to SpaceX's Starlink launches. Since the Starlink launches are decoupled from RDOF since they are not paid for by RDOF. Only the purchases of the UT's by the Rural customers is subsidized making the service affordable to the Rural communities. Such that a suit about RDOF would only affect that UT subsidy program which whether it exists or not will not affect service offerings to rural US customers. RDOF for Starlink is an accelerator not a lifeline.

This suit can only delay the funds by RDOF. Such that if Starlink's funds are suspended likely the complete program's funds would get suspended to all the providers that won bids. Either the bid process was faulty or it was not. Such that all bids and payouts are halted or non are.

The suit is a nuisance to the FCC RDOF program. Not to Starlink. Starlink does not need RDOF to thrive.

[Jcc]

How long can Viasat survive once Starlink is operational?

Just trying to adjust my expectations as to how long we’ll still have to endure this cynical, litigation-based rearguard action before Viasat goes quietly into the night.

Viasat has multiple business lines and Starlink isn't going to have the capacity to completely replace all of their competitors.

Imagine you live in a forested area and like it that way, all you need to find is a narrow line of site to a Geo satellite to mount your antenna, and it stays there. With a Leo constellation you ideally need a clear view all the way around to get the max bandwidth, because the sats will be all over the sky. Some people may prefer to stick with Geo for that reason.

[Robotbeat]

How long can Viasat survive once Starlink is operational?

Just trying to adjust my expectations as to how long we’ll still have to endure this cynical, litigation-based rearguard action before Viasat goes quietly into the night.

Viasat has multiple business lines and Starlink isn't going to have the capacity to completely replace all of their competitors.

Imagine you live in a forested area and like it that way, all you need to find is a narrow line of site to a Geo satellite to mount your antenna, and it stays there. With a Leo constellation you ideally need a clear view all the way around to get the max bandwidth, because the sats will be all over the sky. Some people may prefer to stick with Geo for that reason.

On the contrary, I think as Starlink fills out, you can use a small area of the sky for access. In the early days with a thin constellation, you still need a really good view of the sky.

Also, if you live at like 45 degrees latitude or higher (i.e. northern Europe, like the UK and Germany, etc) it's probably more convenient to use LEO which allows you to access the sky upwards than GSO which requires a view  toward the horizon.

[Robotbeat]

How long can Viasat survive once Starlink is operational?

Just trying to adjust my expectations as to how long we’ll still have to endure this cynical, litigation-based rearguard action before Viasat goes quietly into the night.

Viasat has multiple business lines and Starlink isn't going to have the capacity to completely replace all of their competitors.

I am not sure that's true. In a few years, Starlink may indeed have much more capacity than everyone else combined.

Still, there will be some residual business for Viasat as they cut their prices. Some businesses don't care about latency.

[zubenelgenubi]

Moderator: The fan-boy / fan-boy opposite argument is off-topic. It appears to have come to a halt. Please keep it that way.
Posts deleted.

[vsatman]

NOTE: That once Starship starts launching Starlink sats the V2.0's will be deployed. The second generation sat is likely to be heavier but not that much. Produce possibly 4X the power and at least 4X the throughput per sat by more spot beams and 4X frequency reuse.

Frequency reuse cannot help here
You have only 4000 MHz in Ka band between Sat and GW . To increase this you have to change in V band  with 10000 MHz 37,5..42,5 GHZ or work with 2 or more GateWAys in Ka...

wrong! Yes it does! You can use smaller and more numerous spot beams with larger and more capable phased array antenna. 

Better just explain how  antenna on the satellite will understand between two signals on the same frequency of the Ku band (500 MHz in which 8 beams of 60 MHz are currently operating) coming from two different terminals. Now in satellite communication it is radio interference (jamming)

[oldAtlas_Eguy]

NOTE: That once Starship starts launching Starlink sats the V2.0's will be deployed. The second generation sat is likely to be heavier but not that much. Produce possibly 4X the power and at least 4X the throughput per sat by more spot beams and 4X frequency reuse.

Frequency reuse cannot help here
You have only 4000 MHz in Ka band between Sat and GW . To increase this you have to change in V band  with 10000 MHz 37,5..42,5 GHZ or work with 2 or more GateWAys in Ka...

wrong! Yes it does! You can use smaller and more numerous spot beams with larger and more capable phased array antenna. 

Better just explain how  antenna on the satellite will understand between two signals on the same frequency of the Ku band (500 MHz in which 8 beams of 60 MHz are currently operating) coming from two different terminals. Now in satellite communication it is radio interference (jamming)

Something for those interested to read.
Simulation and experimental measurement of digital multi-beamforming phased antenna array in the frequency range C
https://www.sciencedirect.com/science/article/pii/S2211379718320990

It is not that difficult from a mathematical standpoint but becomes much more difficult from a practical engineering standpoint due to the speeds needed by the digital signal processors involved. But even then for what we are talking about is still doable as long as the beams are non moving with fixed directions. In order to have say 4 separate spot beams in different directions, you would probably need a phased array about 2X the diameter of that of a single beam array. To increase the gains and to make enough physical separation of the formed beams so there is margins between the beams to lessen the interference between beams on the same frequency. Note is that for just 4 beams there is not much of an incentive to do multiple beams in the same array. But for 9,16 or more beams will be more space efficient (as in the surface area that the phased arrays take up) using a single array is less than using separate ones.

[Robotbeat]

NOTE: That once Starship starts launching Starlink sats the V2.0's will be deployed. The second generation sat is likely to be heavier but not that much. Produce possibly 4X the power and at least 4X the throughput per sat by more spot beams and 4X frequency reuse.

Frequency reuse cannot help here
You have only 4000 MHz in Ka band between Sat and GW . To increase this you have to change in V band  with 10000 MHz 37,5..42,5 GHZ or work with 2 or more GateWAys in Ka...

wrong! Yes it does! You can use smaller and more numerous spot beams with larger and more capable phased array antenna. 

Better just explain how  antenna on the satellite will understand between two signals on the same frequency of the Ku band (500 MHz in which 8 beams of 60 MHz are currently operating) coming from two different terminals. Now in satellite communication it is radio interference (jamming)

Because a phased array will only have all the antenna elements in phase if the signal is coming from only a certain area in the sky. All other areas will be out of phase and thus not highly amplified.

[oldAtlas_Eguy]

Even though I think we are getting a little to far afield for this thread.

Frequency reuse spot pattern:
- Each spot of a given frequency is not adjacent to another spot of the same frequency.
- It requires at least 4 separate frequencies to be able to reuse frequencies.
- Use of QAM or QPSK allows for multiple bits per Hz of frequency bandwidth used. Such that 256QAM using 125MHz of bandwidth per channel generates a bit rate per channel of 1Gbps. Thus you need 500MHz of frequency bandwidth to support 4 separate frequency channels each of 1Gbps data rates. Note is that there is a 500 MHz down and 500MHz up frquency allocation at the Ku band for Starlink.
Spot Patterns
F1 F2 F3 F4 F1 F2 F3 F4
F3 F4 F1 F2 F3 F4 F1 F2
F1 F2 F3 F4 F1 F2 F3 F4
F3 F4 F1 F2 F3 F4 F1 F2
F1 F2 F3 F4 F1 F2 F3 F4
F3 F4 F1 F2 F3 F4 F1 F2
F1 F2 F3 F4 F1 F2 F3 F4
F3 F4 F1 F2 F3 F4 F1 F2

4 frequencies but 64 spot beams. An 8X frequency reuse factor in this example.

[vsatman]

Better just explain how antenna on the satellite will understand between two signals on the same frequency of the Ku band (500 MHz in which 8 beams of 60 MHz are currently operating) coming from two different terminals. Now in satellite communication it is radio interference (jamming)

Because a phased array will only have all the antenna elements in phase if the signal is coming from only a certain area in the sky. All other areas will be out of phase and thus not highly amplified.

You persistently tell me about the receive at the user terminal, but above I spoke exclusively about the antenna on the satellite (it is works in the 500 MHz band). And according to your words,  Antenna will receive N signals at the same time at the same frequency from UT in N beams.

[vsatman]

Such that 256QAM using 125MHz of bandwidth per channel generates a bit rate per channel of 1Gbps.

You forgot to mention just one little thing - what signal-to-noise ratio is needed for 256QAM. (it seems to be around 36 dB) let me remind you that now the StarLink user terminal has 9 dB, since 3 dB is 2 times .
Ttat is, the size of the antenna on the user terminal needs to be increased by only 2x2x2x2x2x2x2x2х2 = 512 times . And we will get the size 2 times larger than a football field.....

[Robotbeat]

Better just explain how antenna on the satellite will understand between two signals on the same frequency of the Ku band (500 MHz in which 8 beams of 60 MHz are currently operating) coming from two different terminals. Now in satellite communication it is radio interference (jamming)

Because a phased array will only have all the antenna elements in phase if the signal is coming from only a certain area in the sky. All other areas will be out of phase and thus not highly amplified.

You persistently tell me about the receive at the user terminal, but above I spoke exclusively about the antenna on the satellite (it is works in the 500 MHz band). And according to your words,  Antenna will receive N signals at the same time at the same frequency from UT in N beams.

It works basically the same on both ends as both ends have phased arrays and transmit and receive work the same way for a filled phased array.

[oldAtlas_Eguy]

Such that 256QAM using 125MHz of bandwidth per channel generates a bit rate per channel of 1Gbps.

You forgot to mention just one little thing - what signal-to-noise ratio is needed for 256QAM. (it seems to be around 36 dB) let me remind you that now the StarLink user terminal has 9 dB, since 3 dB is 2 times .
Ttat is, the size of the antenna on the user terminal needs to be increased by only 2x2x2x2x2x2x2x2х2 = 512 times . And we will get the size 2 times larger than a football field.....

This is getting way to far down into detail designs. But a 256QAM is an 8 bits encoded into a single HZ. Such that you need 3DB every time you double the number of bits. 2X2X2 = 8 bits. Or 9DB. The other item is that Starlink already is using the multiple bits per Hz data stuffing technique with its current system. So what is being described other than reusing the frequencies by physically pointing separation of the beams is identical to the current system.

Also nearly 20 years ago cell phones were doing direct RF conversion at 2.4GHz Analog to Digital signal conversion. No IF stage. Just an RF preamp then the signal is digitized and then all further signal detection and manipulation was done all in digital. I reiterate this was being done 20 years ago in cell phones.

Each element of the phased array has its own RF amp located next to the antenna element. Then for each beam direction to be resolved a separate signal line then wanders around to a central point where for that beam direction the signals from all the elements meet up and are added together by an multi input RF Op Amp. There is one of these for each beam. The signal lines are precisely measured for their lengths to provide the phase angle differences to create a synthetic aperture at the desired vector direction from the phased array surface. The only limit to how many different beam directions that can be setup is primarily limited by "real estate": the space for the signal lines and the microelectronics + handful of discrete components.

This is a description of a multi beam phased array with fixed non changing directional beams. There are other methods to make it possible to be able to electronically stear each of the beams separately. But this takes more hardware and more space which may not be available in the size of the array. At this time Starlink would not really need to perform the dynamic steering of the beams so the simple solution will work.

For further detailed design discussion of digital domain controlled microwave dynamic steared multi simultaneous beam forming phased arrays I suggest at least a different thread.

[Robotbeat]

Pretty sure Starlink uses some sort of intermediate frequency.

[vsatman]

But a 256QAM is an 8 bits encoded into a single HZ. Such that you need 3DB every time you double the number of bits. 2X2X2 = 8 bits. Or 9DB. The other item is that Starlink already is using the multiple bits per Hz data stuffing technique with its current system.

8 bit/Hz  is only in theory . Real  is about 6,3
https://www.cablefax.com/archives/spectral-efficiency

and again what  Signal /Noise ratio you need for   256QAM??
Now StarLink   terminal has 9 dB - this is only  8PSK... or 3 bit/Hz..

[su27k]

Parking lots and airports don’t count for rural broadband funding, FCC tells SpaceX

The Federal Communications Commission told SpaceX and other companies on Monday that the billions in rural broadband subsidies it doled out last year can’t be used in already connected areas like “parking lots and well-served urban areas,” citing complaints. The commission, in an effort to “clean up” its subsidy auction program, offered the companies a chance to rescind their funding requests from areas that already have service.

The companies that got the subsidies must do the work to determine they qualify for the money, wrote Michael Janson, director of the FCC’s Rural Broadband Task Force, in a letter addressed to SpaceX’s finance director David Finlay. Similar letters, first reported by Bloomberg, were sent to other recipients of the commission’s Rural Digital Opportunity Fund, a $9.2 billion auction to expand broadband into rural areas that lack or have no service.

[thirtyone]

Parking lots and airports don’t count for rural broadband funding, FCC tells SpaceX

The Federal Communications Commission told SpaceX and other companies on Monday that the billions in rural broadband subsidies it doled out last year can’t be used in already connected areas like “parking lots and well-served urban areas,” citing complaints. The commission, in an effort to “clean up” its subsidy auction program, offered the companies a chance to rescind their funding requests from areas that already have service.

The companies that got the subsidies must do the work to determine they qualify for the money, wrote Michael Janson, director of the FCC’s Rural Broadband Task Force, in a letter addressed to SpaceX’s finance director David Finlay. Similar letters, first reported by Bloomberg, were sent to other recipients of the commission’s Rural Digital Opportunity Fund, a $9.2 billion auction to expand broadband into rural areas that lack or have no service.

I'm actually a little concerned about this one. They better look at these areas very closely. Many people who are underserved are often in strange places no one thinks anyone's living in or seem like they're already covered. I know someone in one of the biggest cities in the US, which you'd think means everyone has Internet. But they happen to be living in a low income public housing area near a highway onramp, with a LOT of characteristics that sound like what this group is complaining about:

$111 million of SpaceX’s $886 million share, the report found, was going to well-served urban areas and random patches of land with no infrastructure, from thin highway medians and empty patches of grass to New York City parking lots and big-box stores.

You wouldn't know it existed unless you knew to physically come down and ask people. Due to infinite amounts of bureaucracy and really because everyone forgets the poor, there is virtually no good Internet of any sort there, even at $100/month (remember - that is still much, much cheaper than moving to non-low-income housing here...). On a map it's close enough to populated areas you probably couldn't imagine there wasn't Internet there.

They could not get Internet to help with their classes and had to study elsewhere to get work done. I actually checked - it is a strange plot of land which every mobile provider claims they have coverage - but we've checked with a few people with different providers and indeed the coverage is so poor (and/or there are so many people) that you would be lucky to get more than a few kbps there, no matter the provider. No one is pointing their antennas in that direction. Due to bureaucracy (the complex is owned by an underfunded department in the city) / no one being able to pay, they couldn't get cable, fiber or any other lines run into the public housing complex. They only have landline (so dialup) but most providers are dropping dialup support these days...and too far for reliable DSL to the point where the kbps mobile internet was more worthwhile (at least they were also getting fast Internet when they were elsewhere in the city). I honestly didn't realize how bad it was until I ran into this particular person, but RDOF should really also cover these areas. It is a very small investment to potentially help a lot of people get out of poverty.

Starlink would be worthwhile short-circuit through all the red tape to run a single stupid cable into the complex, and maybe help give some people who are really trying to get into a better place in life. I actually mentioned this a while ago, and it would definitely be great if they had coverage, especially with a reduced setup fee.

[su27k]

https://twitter.com/DiaMariesbeat/status/1471888782509199366

My latest: Of the top 10 RDOF bidders, only Windstream & some members of the Rural Electric Cooperative Consortium have had winning bids approved by the FCC in a “ready to authorize” notice. That means no LTD Broadband, Charter, Starlink yet. #broadband

https://twitter.com/RonEagleson/status/1472428260285771784

"According to Starlink Services, it does not expect the FCC to act on its pending long-form applications until the second quarter of 2022" from Order - 3023580-Law 10-28-21 Starlink's petition for reconsideration. https://puc.pa.gov/docket/P-2021-3023580

[Lar]

su27k, what do you (or anyone) think that means?

[su27k]

su27k, what do you (or anyone) think that means?

I read a few documents in the docket, it seems the delay is because SpaceX is still negotiating with 3rd party vendor for providing VoIP service. RDOF requires winner to provide a phone service along with broadband.

[vsatman]

su27k, what do you (or anyone) think that means?

I read a few documents in the docket, it seems the delay is because SpaceX is still negotiating with 3rd party vendor for providing VoIP service. RDOF requires winner to provide a phone service along with broadband.

just VoiP telephony is not a problem.
For an ETS license it is needed, when a subscriber calls 112, the call is routed to an 112` operator  in the nearest settlement.
In classic wired networks, this is simple, unlike VoIP

[Lar]

I am going to assume that "112" is the equivalent of "911" in the US?  ... an emergency contact point local to where you are that can dispatch assets (police/fire/ambulance/etc) to your location.

This is a solved problem for VOIP and cell, I thought.

[niwax]

I am going to assume that "112" is the equivalent of "911" in the US?  ... an emergency contact point local to where you are that can dispatch assets (police/fire/ambulance/etc) to your location.

This is a solved problem for VOIP and cell, I thought.

112 is the international standard number, if you happen to be in the US while dialing 112, the network will route it to the nearest 911 operator.

[DanClemmensen]

I am going to assume that "112" is the equivalent of "911" in the US?  ... an emergency contact point local to where you are that can dispatch assets (police/fire/ambulance/etc) to your location.

This is a solved problem for VOIP and cell, I thought.

It's fairly well solved for cell, because the cell tower has some idea of the cell phone's location. It is poorly solved for VOIP, because it generally requires the VOIP user to voluntarily enter the location of the VOIP phone. Starlink (and other non-GEO constellations) actually have an easier technical problem to solve, because the Starlink user equipment must have a precise location for the terminal to operate at all, even closer than a cell tower knows the location of a cell phone. So, if someone just plugs a VOIP unit into a Starlink user terminal as just another IP device, you get the same problems as any other VOIP unit on any IP network. However, If SpaceX provides a more customized solution, then it can find the proper emergency call center associated with the latitude and longitude of the user station. This will remain true even if in the future the Starlink user terminal is mounted on a vehicle.

[vsatman]

I am going to assume that "112" is the equivalent of "911" in the US?  ... an emergency contact point local to where you are that can dispatch assets (police/fire/ambulance/etc) to your location.

This is a solved problem for VOIP and cell, I thought.

Starlink (and other non-GEO constellations) actually have an easier technical problem to solve, because the Starlink user equipment must have a precise location for the terminal to operate at all, even closer than a cell tower knows the location of a cell phone. So, if someone just plugs a VOIP unit into a Starlink user terminal as just another IP device, you get the same problems as any other VOIP unit on any IP network. However, If SpaceX provides a more customized solution, then it can find the proper emergency call center associated with the latitude and longitude of the user station. This will remain true even if in the future the Starlink user terminal is mounted on a vehicle.

Yes, you wrote everything correctly, but even if SpaceS has a mark where the subscriber is located, it must transfer this call to the local PSTN (Public Switched Telephone Network) provider who will transfer the call to the appropriate 911 police / fire department office. And here the question is on what conditions this local provider will agree to receive calls from SpaceX and for how many dollars to upgrade its network for this it want...

[DanClemmensen]

I am going to assume that "112" is the equivalent of "911" in the US?  ... an emergency contact point local to where you are that can dispatch assets (police/fire/ambulance/etc) to your location.

This is a solved problem for VOIP and cell, I thought.

Starlink (and other non-GEO constellations) actually have an easier technical problem to solve, because the Starlink user equipment must have a precise location for the terminal to operate at all, even closer than a cell tower knows the location of a cell phone. So, if someone just plugs a VOIP unit into a Starlink user terminal as just another IP device, you get the same problems as any other VOIP unit on any IP network. However, If SpaceX provides a more customized solution, then it can find the proper emergency call center associated with the latitude and longitude of the user station. This will remain true even if in the future the Starlink user terminal is mounted on a vehicle.

Yes, you wrote everything correctly, but even if SpaceS has a mark where the subscriber is located, it must transfer this call to the local PSTN (Public Switched Telephone Network) provider who will transfer the call to the appropriate 911 police / fire department office. And here the question is on what conditions this local provider will agree to receive calls from SpaceX and for how many dollars to upgrade its network for this it want...

Sure, but at the technical level all of those interfaces are well specified, well understood, and already in use for VOIP. This means that problems will be primarily administrative and contractual, not technical, except perhaps for cumbersome lethargic responses within the IT departments of the PSTN providers, where simply updating the interface database can take months.

[su27k]

SpaceX loses $900 million in rural broadband subsidies

paceX has lost its bid for nearly $900 million in rural broadband subsidies for its Starlink broadband service.

The Federal Communications Commission said Aug. 10 that SpaceX had failed to show it could meet requirements for unlocking the funds, which aim to incentivize expanding broadband services to unserved areas across the United States.

“We must put scarce universal service dollars to their best possible use as we move into a digital future that demands ever more powerful and faster networks,” FCC chair Jessica Rosenworcel said in a statement.

“We cannot afford to subsidize ventures that are not delivering the promised speeds or are not likely to meet program requirements.”

[su27k]

https://twitter.com/BrendanCarrFCC/status/1562433655578193921

The FCC’s abrupt decision to reverse an $885 million infrastructure award to Elon Musk’s Starlink is concerning.

For one, the decision is without legal justification.

For another, it will leave rural Americans waiting on the wrong side of the digital divide.

My statement:

[rubicondsrv]

https://twitter.com/BrendanCarrFCC/status/1562433655578193921

The FCC’s abrupt decision to reverse an $885 million infrastructure award to Elon Musk’s Starlink is concerning.

For one, the decision is without legal justification.

For another, it will leave rural Americans waiting on the wrong side of the digital divide.

My statement:

The thing is it will not impact starlink deployment at all, in fact the strongest argument against starlink subsidies is the network is getting built at the same rate with or without the money. 

that is not to say there might or might not be funny business involved in the decision, but to argue it will impact access to starlink is questionable at best.   
   

[neoforce]

The thing is it will not impact starlink deployment at all, in fact the strongest argument against starlink subsidies is the network is getting built at the same rate with or without the money. 

that is not to say there might or might not be funny business involved in the decision, but to argue it will impact access to starlink is questionable at best.   
 

Agreed it won't impact starlink deployment.  And I don't know how RDOF works, but was any of the money set aside for low income folks to get subsidies, as in this program:  https://www.fcc.gov/acp  Or was the money 100% for infrastructure?

If the ACP program would have gotten this money for some folks to get discounts on starlink, this decision certainly impacts people. 

[DreamyPickle]

I wonder if there's any connection between the RDOF program and the Starlink/T-Mobile announcement?

In theory ductaping a Starlink dish with a 5G basestation would be an extremely effective and fast way to boost connectivity, much cheaper than fiber.

After kicking out Starlink from RDOF they're no longer competitors so maybe they're going to partner.

[rubicondsrv]

Agreed it won't impact starlink deployment.  And I don't know how RDOF works, but was any of the money set aside for low income folks to get subsidies, as in this program:  https://www.fcc.gov/acp  Or was the money 100% for infrastructure?

If the ACP program would have gotten this money for some folks to get discounts on starlink, this decision certainly impacts people. 

in most of the areas covered cost subsidies is a minor consideration, the improvement from existing service is so great that even if large amounts of people cant afford the service it will be a huge improvement.

these are areas where the best case for improving service involves being close to an existing line and paying tens of thousands of dollars or more to the owner of that line to provide service to your house, if they are even willing to do anything.

they could charge $500 per month and still have large numbers of subscribers. 

[CT Space Guy]

Cross posted
Perhaps SpaceX is not qualified in the same way that Tesla was not qualified to be at the White House for the EV summit back in August

[oldAtlas_Eguy]

The losers here is not SpaceX but the RDOF qualified subscribers who lost their priority position in the receiving of a terminal. An interesting side note is that there is not much of a terminal waiting list anymore but a bandwidth/bit-rate at a location waiting list.

As more sats and Gateways are added the data rates will improve. The current situation is a temporary occurrence that was caused by a terminal build rate that outstripped the sat launch rate last year and keeping SpaceX hopping in expanding the capabilities of the existing and installation of new Gateways. Eventually the subscriber rate increase related to the sat deployment rate will start to widen vs contracting.

[su27k]

https://twitter.com/thesheetztweetz/status/1568352153026154504

Update: SpaceX filed a request to the FCC, asking to appeal the FCC's RDOF decision and saying the denial issued last month "is flawed as a matter of both law and policy."

https://fcc.gov/ecfs/filing/status/detail/confirmation/20220909251827409


SpaceX implying the FCC's turnaround on RDOF subsidies is an "improper attempt" under President Biden's administration "to undo" a decision made during former President Trump's administration, saying "it is hard not to see it" that way.

[launchwatcher]

Looks like the appeal was rejected:

FCC REAFFIRMS DECISION TO REJECT STARLINK APPLICATION
FOR NEARLY $900 MILLION IN SUBSIDIES

Applicant Failed to Meet Burden for Rural Digital Opportunity Fund

WASHINGTON, December 12, 2023—The Federal Communications Commission today
reaffirmed the Wireline Bureau’s prior decision to reject the long-form application of Starlink to
receive public support through the Rural Digital Opportunity Fund program, based on the
applicant’s failure to meet the program requirements.

...

https://www.fcc.gov/document/fcc-reaffirms-rejection-nearly-900-million-subsidy-starlink
Ruling (pdf) is here:
https://docs.fcc.gov/public/attachments/FCC-23-105A1.pdf

[eeergo]

Looks like the appeal was rejected:

FCC REAFFIRMS DECISION TO REJECT STARLINK APPLICATION
FOR NEARLY $900 MILLION IN SUBSIDIES

Applicant Failed to Meet Burden for Rural Digital Opportunity Fund

WASHINGTON, December 12, 2023—The Federal Communications Commission today
reaffirmed the Wireline Bureau’s prior decision to reject the long-form application of Starlink to
receive public support through the Rural Digital Opportunity Fund program, based on the
applicant’s failure to meet the program requirements.

...

https://www.fcc.gov/document/fcc-reaffirms-rejection-nearly-900-million-subsidy-starlink
Ruling (pdf) is here:
https://docs.fcc.gov/public/attachments/FCC-23-105A1.pdf

Reasoning excerpt:

[Precedents]
Among other things, the Bureau asked Starlink to explain why its network performance was below the required minimum speeds of 100/20 Mbps [redacted + context note: A Public Notice announcing that Starlink was in default was released concurrently] After reviewing all of the information submitted by Starlink, the Bureau ultimately concluded that Starlink had not shown that it was reasonably capable of fulfilling RDOF’s requirements to deploy a network of the scope, scale, and size required to serve the 642,925 model locations in 35 states for which it was the winning bidder.
[Latest resolution]
the Bureau followed Commission guidance and correctly concluded that Starlink is not reasonably capable of offering the required high-speed, low-latency service throughout the areas where it won auction support [...] by approving Starlink’s short-form application, the Bureau concluded that, based on the high-level information required in the short-form application, Starlink was reasonably capable of offering, at some level, the required service in at least one relevant area in each of the states in which it was approved to bid [...] the long-form application review [instead] determined whether the applicant could provide that service “associated with its winning bids,” i.e., in each of the areas where it ultimately won support
The most recent available evidence showed that "Starlink performance had been declining for download speed, upload speed and jitter test performance". In other words, it was not only failing to meet the RDOF public interest obligations, but also trending further away from them. [...] Unlike fiber and other technologies currently in use, Starlink did not point to examples where its technology was providing service at the required level in the United States.
Starlink recorded a median download speed of 64.54 Mbps in Q3 2023, a marginal decline quarter-on-quarter, but still an increase over the 53.00 Mbps recorded in Q3 2022. Even if the performance had improved though, that would still not demonstrate an ability to meet RDOF's performance standards, and it also does not show how Starlink would meet its RDOF obligations to a significantly larger customer base.

[edzieba]

Short version:
- Funding required providing uplink and downlink speeds above a certain threshold.
- Starlink could provide these speeds, but only by limiting subscribers per cell - fixed bandwidth per satellite, fixed number of satellites visible per cell, subdivide too much and you drop below that threshold per subscriber.
- Starlink had the choice to either go for the funding and cap subscriber numbers within the US, or keep subscribers uncapped and lose out on funding due to not meeting speed requirements.
- Starlink chose not to cap subscribers, thus could not guarantee the minimum speeds required for fund eligibility.

Whether that was the right choice is up to Starlink's accountants (i.e. whether the funds lost out on are more or less than the subscription fees lost out on by capping subscribers until more satellites can be rolled out), but they can't have their cake and eat it when it comes to bandwidth-per-sat vs. bandwidth-per-subscriber.

[thespacecow]

Short version:
- Funding required providing uplink and downlink speeds above a certain threshold.
- Starlink could provide these speeds, but only by limiting subscribers per cell - fixed bandwidth per satellite, fixed number of satellites visible per cell, subdivide too much and you drop below that threshold per subscriber.
- Starlink had the choice to either go for the funding and cap subscriber numbers within the US, or keep subscribers uncapped and lose out on funding due to not meeting speed requirements.
- Starlink chose not to cap subscribers, thus could not guarantee the minimum speeds required for fund eligibility.

Whether that was the right choice is up to Starlink's accountants (i.e. whether the funds lost out on are more or less than the subscription fees lost out on by capping subscribers until more satellites can be rolled out), but they can't have their cake and eat it when it comes to bandwidth-per-sat vs. bandwidth-per-subscriber.

Of course they can have their cake and eat it too, they just need to launch more satellites and increase the overall network capacity, which is what they're doing.

The latest ookla data is showing 79.04 Mbps in November, an substantial increase from the 64.54 Mbps in Q3 2023, and SpaceX is still two years away from the first milestone, where is the analysis to show they couldn't get to 100 Mbps in the next two years with the insane # of launches they're planning?

[ZachF]

The primary objective of this program was never bringing broadband to rural areas, that was just a possible side effect of the real objective: spreading funds to politically connected special interests.

[Nomadd]

 In other words, the government requirements would have prevented people from getting service instead of helping them.

[edzieba]

In other words, the government requirements would have prevented people from getting service instead of helping them.

Damned if you do, dammed if you don't. Set very low (100mbps down 30 mbps up IIRC) minimum service requirements, be accused of setting requirements too high, don't set minimum service requirements and be accused of GEO SATCOM/dialup/GSM providers receiving funds for doing little to nothing.

Sadly, regulatory capture means that proposing an actual solution (funding of last-mile link installation, and mandating Local Loop Unbundling) which would allow actual direct competition, would disrupt the current commercial monopoly model that US ISPs operate under, so is a non-starter without regulatory reform.

[DanClemmensen]

In other words, the government requirements would have prevented people from getting service instead of helping them.

Damned if you do, dammed if you don't. Set very low (100mbps down 30 mbps up IIRC) minimum service requirements, be accused of setting requirements too high, don't set minimum service requirements and be accused of GEO SATCOM/dialup/GSM providers receiving funds for doing little to nothing.

Sadly, regulatory capture means that proposing an actual solution (funding of last-mile link installation, and mandating Local Loop Unbundling) which would allow actual direct competition, would disrupt the current commercial monopoly model that US ISPs operate under, so is a non-starter without regulatory reform.

Yep. SpaceX's response will be to ultimately provide a better service for less money than even the subsidized providers. It'll just take a bit longer. Maybe two years?

[envy887]

In other words, the government requirements would have prevented people from getting service instead of helping them.

Damned if you do, dammed if you don't. Set very low (100mbps down 30 mbps up IIRC) minimum service requirements, be accused of setting requirements too high, don't set minimum service requirements and be accused of GEO SATCOM/dialup/GSM providers receiving funds for doing little to nothing.

Sadly, regulatory capture means that proposing an actual solution (funding of last-mile link installation, and mandating Local Loop Unbundling) which would allow actual direct competition, would disrupt the current commercial monopoly model that US ISPs operate under, so is a non-starter without regulatory reform.

The requirements were 100/20, and that is not "very low" when you consider that the vast majority of the areas Starlink bid on have zero providers offering service meeting the old 25/3 FCC standard. Starlink, even if it isn't quite at 100/20 on average yet, is still a massive upgrade.

[matthewkantar]

Favoring the perfect that has historically not come, over the good that is available right now screws the folks the program is supposedly trying to help. 

[cpushack]

In other words, the government requirements would have prevented people from getting service instead of helping them.

Damned if you do, dammed if you don't. Set very low (100mbps down 30 mbps up IIRC) minimum service requirements, be accused of setting requirements too high, don't set minimum service requirements and be accused of GEO SATCOM/dialup/GSM providers receiving funds for doing little to nothing.

Sadly, regulatory capture means that proposing an actual solution (funding of last-mile link installation, and mandating Local Loop Unbundling) which would allow actual direct competition, would disrupt the current commercial monopoly model that US ISPs operate under, so is a non-starter without regulatory reform.

The requirements were 100/20, and that is not "very low" when you consider that the vast majority of the areas Starlink bid on have zero providers offering service meeting the old 25/3 FCC standard. Starlink, even if it isn't quite at 100/20 on average yet, is still a massive upgrade.

And they had till 2025 to hit 100mb average, and are trending higher, yet the govt says they are trending lower (despite the data)

[r8ix]

There is some question, based on recent court cases (primarily involving the SEC), that the FCC ruling on its own decision may, in fact, be unconstitutional. In the SEC cases, higher courts have ruled that internal judication is an illegal transfer of authority from the judicial branch to the executive branch. Something similar could, in principle, be in play here.

[Asteroza]

If the FCC is gonna hold them to that standard, they're also checking on the other RDOF participants and their adherence to said standard, right?

[edzieba]

In other words, the government requirements would have prevented people from getting service instead of helping them.

Damned if you do, dammed if you don't. Set very low (100mbps down 30 mbps up IIRC) minimum service requirements, be accused of setting requirements too high, don't set minimum service requirements and be accused of GEO SATCOM/dialup/GSM providers receiving funds for doing little to nothing.

Sadly, regulatory capture means that proposing an actual solution (funding of last-mile link installation, and mandating Local Loop Unbundling) which would allow actual direct competition, would disrupt the current commercial monopoly model that US ISPs operate under, so is a non-starter without regulatory reform.

The requirements were 100/20, and that is not "very low" when you consider that the vast majority of the areas Starlink bid on have zero providers offering service meeting the old 25/3 FCC standard. Starlink, even if it isn't quite at 100/20 on average yet, is still a massive upgrade.

And they had till 2025 to hit 100mb average, and are trending higher, yet the govt says they are trending lower (despite the data)

Because Starlink service started out well above the threshold (400mbps achieved in some cases), and dropped down below it over time as service was rolled out to more subscribers and the fixed amount of bandwidth available per cell saturated. The only way average available bandwidth per user can rise back above the threshold again is either by throwing subscribers off the service not not letting them back on, significantly increasing the number of satellites in orbit (essentially, doubling the number of satellites of the same design is required to double the bandwidth), and/or replacing the existing satellites with the same number of new satellites with more bandwidth available.

So the question then becomes whether the FCC expect Space can roll out enough new and/or replacement satellites to fill the shells required to bring available cell bandwidth comfortably above the threshold (because if they're just barely meeting the threshold for existing subscribers, they can't onboard any new subscribers, defeating the purpose of the funding for rollout to new subscribers) by 2025. Presumably the threshold takes into account the ability to meet that threshold whilst servicing some minimum proportion of the population within each cell (because only being able to offer the guaranteed service to 0.1% - for example - of users per cell is the problem they're trying to solve). Their current determination is that SpaceX is not capable of meeting that, hence the judgement.

[RedLineTrain]

Their current determination is that SpaceX is not capable of meeting that, hence the judgement.

The Wireline Competition Bureau's current determination, mind.  These folks can be expected to be skeptical of wireless, which is why the Commissioners should have evaluated this for themselves, asked around the Space Bureau staff, and reversed the determination based on SpaceX's growing constellation buildout momentum.

It's indefensible on its merits.  That the vote was on a party line basis rightfully raises eyebrows.  The pound of flesh for this vote should be extracted.

[envy887]

If the FCC is gonna hold them to that standard, they're also checking on the other RDOF participants and their adherence to said standard, right?

Eventually they will be checking, but there is no penalty for underperforming until 2026.

[cpushack]

Citizens are petitioning the FCC's decision which has VIASat worried enough to file legal opinions against it.

https://www.pcmag.com/news/viasat-tries-to-stop-citizen-effort-to-revive-fcc-funding-for-starlink

[DistantTemple]

Tongue in cheek - but not impossible.
With SX's RDOF "cancelled", alternative providers will not be able to provide an appropriate quality of service - or perhaps any service to some of these unconnected citiczens.
In rural  locations somebody could become the "last mile provider" somehow using an RDOF award (money that would have gone to SX).
However that somebody (local government/last mile provider/Starlink competitor) would need to buy one of SpaceX's new "Private Ground Stations" .... thereby paying SpaceX to do the real work, and restoring it in a roundabout way to a rightful place in the program.
(Unlike several others on NSF I am NOT a communications engineer!)

[raptorx2]

Not sure if this was posted before?

OT RDOF - Starlink Plans to Join Affordable Connectivity Fund Subsidy Program

https://broadbandbreakfast.com/2023/12/starlink-plans-to-join-affordable-connectivity-fund-subsidy-program/

[raptorx2]

Citizens are petitioning the FCC's decision which has VIASat worried enough to file legal opinions against it.

https://www.pcmag.com/news/viasat-tries-to-stop-citizen-effort-to-revive-fcc-funding-for-starlink

Or a change in Administration.