Eutelsat OneWeb: Constellation - General Thread

[tonya]

Plus no GNSS chip set maker will bother to add One Web.
...

This doesn't make any sense since there are already chipset makers supporting Galileo or Glonass or BeiDou, too.

The UK system is also unlikely to be dual use anytime soon or possibly ever. It's a defence project which won't use consumer chipsets.

[GWR64]

This rumor for FT is interesting. Supposedly at least one of the bids includes a consortium with some UK gov't backing, to the tune of £500M.

https://twitter.com/Megaconstellati/status/1276380004578922496

It'll be interesting to see how this all plays out tomorrow. Is tomorrow still the deadline for all of the proposed bids? How much info (if any) about the bids will be public information? Will we even know who actually placed bids? Or will we only hear whoever had the best bid, once the bankruptcy judge rules on things? Anyone here with Chapter 11 experience care to weigh in on what next steps look like?

~Jon

The Financial Times article is dated June 25. When did the bidding period expire exactly?

[edzieba]

I recall that we declined because they wouldn't allow acess to 100% of the system.

The proposed arrangement was 100% access (including the military subchannel), but UK companies - the UK no longer being a member of the EU and rejecting integration with EU standards and leaving the EU security interoperation system - would not be eligible to bid for further work. The UK claimed that they would no longer have absolute certainty that that deal would not change in the future so rejected this arrangement (with everyone being pretty certain the real reason was because UK companies would not be getting the work).
Somewhat ironic that a 20% share in an international consortium would be considered as an adequate way to address the claimed reason, and the satellite construction arm being US located and partially owned mostly by European companies not addressing the 'no work for UK manufacturers' reason.

[Rocket Rancher]

So who wants to take a stab at what would need to be done to turn OneWeb into a positioning system?
As Robotbeat says, have the atomic clock external to the satellites, for starters. What could be achieved with the birds already on orbit?

I have a hard time figuring out what parts they even have that could be used as a baseline to work from. OneWeb doesn't actually own the bus and manufacturing, all of that was part of a 50-50 joint venture with Airbus. Considering Airbus is also a creditor in the bankruptcy, so gets a part of OneWebs share in the joint venture, using the bus amounts to buying it off Airbus just like any other. The bus itself can't easily be reconfigured to do anything else because it doesn't actually contain a communication payload in the traditional sense, they are designed as dumb analog repeaters. You can't go and add your own messages because there is no communcations logic - just a pipe from input to output. As far as user terminals, the latest we heard was that they were outsourced to someone else. The topic of launches has also been discussed to death - any cheap market rate they could get, pretty much, but nothing amazing either.

And I don't see why we should expect anything else. The history of OneWeb shows that they had to take every inflexible compromise they could get to arrive at a fragile but potentially workable solution. If they had succeeded it would have been because they had sacrificed every part of their system that didn't lead to the quickest functioning product for the sake of cost. But any way forward, even staying in the communication industry, would have been very hard work.

The statement " The history of OneWeb shows that they had to take every inflexible compromise they could get to arrive at a fragile but potentially workable solution." can imply that OneWeb has "cut corners". Does this refer to the finance part or the technical part of their system?

And are you familiar with Arrow?

[Kaputnik]

So are we in agreement that there is no way that the existing satellites could be used for position finding?

[gosnold]

I disagree with Elon, but when I defended this idea, I got excoriated on Twitter LOL.

One possibility for a GNSS payload is to do what the Japanese did (experimentally) on their local GNSS satellites: instead of having an on-board atomic clock, simply re-broadcast from an atomic clock at the ground station. This drastically reduces the cost, weight, etc of a GNSS payload, perhaps enough to use the OneWeb bus.

It's funny how simply proposing another possibility as reason why you can't dismiss something out of hand (without better knowledge) gets attacked so vociferously...

That might work for civilian use but not for some military applications: you need a ground station visible by the satellites, and that means if you are striking deep into enemy territory that the ground station has to be located inside or closes to their territory. That makes the system unreliable.

However the satellite application catapult claims that Onweb can provide PNT with the current design:
https://sa.catapult.org.uk/news/oneweb-the-opportunity-for-the-uk/

Our analysis concludes that, even without modification, the existing OneWeb satellites could provide a highly accurate timing service. And then, by upgrading the satellites, and providing an appropriate navigation ground component, it would be possible to significantly augment or even replace existing navigation systems.

It might not be as precise as Galileo or GPS though, but in some application it doesn't matter.

[Nomadd]

I disagree with Elon, but when I defended this idea, I got excoriated on Twitter LOL.

One possibility for a GNSS payload is to do what the Japanese did (experimentally) on their local GNSS satellites: instead of having an on-board atomic clock, simply re-broadcast from an atomic clock at the ground station. This drastically reduces the cost, weight, etc of a GNSS payload, perhaps enough to use the OneWeb bus.

It's funny how simply proposing another possibility as reason why you can't dismiss something out of hand (without better knowledge) gets attacked so vociferously...

I never did quite understand why a single reference signal couldn't be used for a system. Accurate, constantly updated orbital information shouldn't be that hard to stream to receivers along with the clock signal. It sounds easier than trying to keep a load of separate clocks perfect.
 But if Twitter says it's wrong, you shouldn't even consider it.

[niwax]

I disagree with Elon, but when I defended this idea, I got excoriated on Twitter LOL.

One possibility for a GNSS payload is to do what the Japanese did (experimentally) on their local GNSS satellites: instead of having an on-board atomic clock, simply re-broadcast from an atomic clock at the ground station. This drastically reduces the cost, weight, etc of a GNSS payload, perhaps enough to use the OneWeb bus.

It's funny how simply proposing another possibility as reason why you can't dismiss something out of hand (without better knowledge) gets attacked so vociferously...

I never did quite understand why a single reference signal couldn't be used for a system. Accurate, constantly updated orbital information shouldn't be that hard to stream to receivers along with the clock signal. It sounds easier than trying to keep a load of separate clocks perfect.
 But if Twitter says it's wrong, you shouldn't even consider it.

You don't really need to keep the clocks perfect, you need to keep the offsets between them well-known to figure out your position from multiple clock signals. I'd guess the reason they decided to put the clocks on the satellites is because it was easier to do than to figure out the exact offset of a terrestrial signal through atmospheric distortion and the like. Naively, you'd at least double your offset error compared to the current method. I'd also wager any theoretical correction would be near-impossible if your satellites are being thrown around in the upper atmosphere.

[nacnud]

That and clocks on the ground tend to be vaporised just when the military really would like them. 

[Nomadd]

That and clocks on the ground tend to be vaporised just when the military really would like them. 

The reference doesn't have to be on the ground and can be as redundant as you want.
 The system isn't always as well updated as it should be. Certain satellites being in the mix tend to cause more error than others. Took me a while to figure out it was the particular sats and not just the signal path of the moment.

[alexterrell]

The biggest source of error (when it is working correctly - and after reflected signals if they can't be removed) in GNSS is variations in the speed of light caused by different atmospheric conditions. (Variations in gravitational fields cause a smaller error, which can also be accounted for).

If you add a second link from base station to satellite, you double this error at least. That makes your system less accurate than what your opponents are using - probably Google Maps on smart phones.

You also need to send the clock signal from your ground station to the satellites. That is fine if you're trying to navigate around Salisbury Plain, but if you need to go to say Afghanistan, your ground signal has to be relayed via other signals. That won't be analogue signal amplification, but digital processing taking variable numbers of microprocessor clock cycles. These need to be counted and at the same time have a fixed length - needing an external atomic clock. This might be doable, and might even be able to run on the computers on the One Web satellites, but I wouldn't bet on it. 

You can also significantly improve accuracy by using an overlay system e.g. https://www.gsa.europa.eu/egnos/what-egnos

This relies on receivers with a known position to calculate an error value and broadcast this over mobile phone signals. However, using ground based clocks will make the error dependent on which satellites you are using, so the correction signal will vary from one handset to the other.

Plus, as mentioned, this will not be a normal free GNSS service, so why would chip set makers accommodate it? Then you end up with customised military chipsets at >$1000, compared to less than $1 for what is in your phone.

[alexterrell]

So who wants to take a stab at what would need to be done to turn OneWeb into a positioning system?
As Robotbeat says, have the atomic clock external to the satellites, for starters. What could be achieved with the birds already on orbit?

You may be able to do something, but I'm sure it would be a whole lot easier to just buy back into Galileo, which was definitely on the table until the UK decided it wasn't.

It would be cheaper to properly fund Loran C.

That would also give a foolproof integrity check and back up for all the navigation systems.

But Loran C is so last Century.

[Nomadd]

So who wants to take a stab at what would need to be done to turn OneWeb into a positioning system?
As Robotbeat says, have the atomic clock external to the satellites, for starters. What could be achieved with the birds already on orbit?

You may be able to do something, but I'm sure it would be a whole lot easier to just buy back into Galileo, which was definitely on the table until the UK decided it wasn't.

It would be cheaper to properly fund Loran C.

That would also give a foolproof integrity check and back up for all the navigation systems.

But Loran C is so last Century.

Loran only survived as long as it did for the same kind of non-technical reasons that some in the UK are pushing it's own positioning system for. That and having places to exile USCG ETs you were mad at. (I was fortunate enough to snag a 378 myself.)
 I would be seriously surprised if re-purposing those existing sats wasn't a complete time and money waste compared to just building what they need.

[TorenAltair]

[...]
 I would be seriously surprised if re-purposing those existing sats wasn't a complete time and money waste compared to just building what they need.

I agree. But as in many cases it seems to be a political decision. Starting from "we want to be independent of the Continental Europe" to "it is a UK company". Just remember Iridium or Globalstar and the US military backing. So I assume there will be enough political backing, perhaps some kind of pressure, that they get the bid. If it makes OneWeb more viable, that imo remains questionable.

[Rocket Rancher]

Canada's Telesat has put in a bid for OneWeb, source UK Telegraph. Auction on Thursday.

[Asteroza]

[...]
 I would be seriously surprised if re-purposing those existing sats wasn't a complete time and money waste compared to just building what they need.

I agree. But as in many cases it seems to be a political decision. Starting from "we want to be independent of the Continental Europe" to "it is a UK company". Just remember Iridium or Globalstar and the US military backing. So I assume there will be enough political backing, perhaps some kind of pressure, that they get the bid. If it makes OneWeb more viable, that imo remains questionable.

The ugly issue is terminal cost for a GNSS. You only have cheap chips because they make so many that also cover all the major ones (GPS/GLONASS/Beidou/Galileo). Getting the chip makers to add chip area to acommodate yet another GNSS system is going to be a hard sell.

EDIT

But, with the push towards SDR's paired with adaptive antennas, if the current chip is internally an SDR and the antenna can reasonably accommodate the frequencies, then it really just becomes a new software patch/firmware blob that hardware makers release in their update cycle (assuming the added software fits within the chip at runtime)...

Also, the major reason current chips support the big 4 is pressure from cellphone makers, who were pressured by governments. The start was Russia requiring mobile phones with GNSS capabilities to use GLONASS,and Apple didn't want to lose iPhone business there so they asked for support. Likely around the same time, Galileo and Beidou were on the horizon (and similar chinese and EU regulations were foreseen), so GNSS chip makers shoehorned in support around the same time I believe. With the smalller chip manufacturing node sizes at the time comping online at various semiconductor fabs, adding GLONASS to improve coverage is not that hard an ask as it was extant. Adding Galileo and Beidou is justifiable due to market size.

Nobody is going to support a british specific standard as the market is far too small, unless the standard itself is basically a straight copy with small frequency offsets, such as the semilocal SBAS systems like WASS or EGNOS.

[gongora]

Canada's Telesat has put in a bid for OneWeb, source UK Telegraph. Auction on Thursday.

Telesat has been very critical of OneWeb's design.  I wonder if completing OneWeb would just be a lot cheaper than building the Telesat design from scratch, and then they would do their own design on second generation sats.  Kinda doubt Telesat will outbid the field, but you never know what's going to happen with auctions.

[ChrisWilson68]

Canada's Telesat has put in a bid for OneWeb, source UK Telegraph. Auction on Thursday.

Telesat has been very critical of OneWeb's design.  I wonder if completing OneWeb would just be a lot cheaper than building the Telesat design from scratch, and then they would do their own design on second generation sats.  Kinda doubt Telesat will outbid the field, but you never know what's going to happen with auctions.

Maybe they don't have any plans to complete OneWeb's constellation but they think they can get the spectrum rights transferred over to their chosen design.  Or maybe they just want to buy it to kill it, to keep someone else from completing it and competing with them.  Or maybe they don't want to win the auction at all but want to drive the price up for whoever wins.

[gongora]

Canada's Telesat has put in a bid for OneWeb, source UK Telegraph. Auction on Thursday.

Telesat has been very critical of OneWeb's design.  I wonder if completing OneWeb would just be a lot cheaper than building the Telesat design from scratch, and then they would do their own design on second generation sats.  Kinda doubt Telesat will outbid the field, but you never know what's going to happen with auctions.

Maybe they don't have any plans to complete OneWeb's constellation but they think they can get the spectrum rights transferred over to their chosen design.  Or maybe they just want to buy it to kill it, to keep someone else from completing it and competing with them.  Or maybe they don't want to win the auction at all but want to drive the price up for whoever wins.

I'd say "buy it to kill it" is not an option for Telesat, they don't have that kind of money to waste.

[high road]

It started getting quite political, leading to comparisons of whether the UK or US government was dumber, which is not a road this thread really needs to go down

Ah, that's fair. Question does anybody know if there will be any more info released publicly today from the bankruptcy proceedings?

~Jon

Doesn't Oneweb owe money to Altius for the grapple fixtures they were going to use? I assumed that if any information would be available, you'd be the first to know.