Author Topic: SpaceX and OneWeb internet satellite comparison  (Read 119705 times)

Offline MikeAtkinson

  • Full Member
  • ****
  • Posts: 1980
  • Bracknell, England
  • Liked: 784
  • Likes Given: 120
SpaceX and OneWeb internet satellite comparison
« on: 01/18/2015 09:47 am »
Here is a comparison of the SpaceX/Musk and OneWeb/Wyler global internet satellite constellations. I'll update it as more information becomes available.

FeatureOneWebSpaceX
Aims
Number of users
Number of business users10%[2]
Suburban penetration10%[6]
% of long-haul internet traffic50%[2]
.
Constellation
Number of sats648[4]~900[11]4025[2]
Initial usable size
Version 1 full capability5 years[6]
Total bandwidth
FrequencyKu band[1]Ku band?, Ka band>24GHz [8]
Replenishment strategydon't replace, tolerate failures[6]
Evolution strategy2400 sat 2nd generation[5]Not all capabilities initially[6]
Generation timeNew versions every 5 years or less[6]
.
Satellites
Altitude1200 km[1]750 miles[3], 1100km[2]
Mass125 kg[1]150 kg[11]several hundred kg[2]
SizeOf a beer fridge[9]
Beams
Power
Bandwidth/beam
Total bandwidth8 gigabit/s[1]
Inter-satellite links
Lifetime5 years?
Station keepingHall thrusters[2]
Disposal
Technology levelconservative state of the artsophisticated[3]
Cost$1-2M [9], $400k [10]
.
Launch
1st launch2017[1]2018[11]
Launcher?F9R
Replenishment launcherLauncherOne[1]N/A ?
.
User Ground Segment
Uplink bandwidth max
Uplink bandwidth typical
Downlink bandwidth max50 megabit/s[1]gigabit level[6]
Downlink bandwidth typical
Antenna size36cm x 16cm[1]
Antenna typemechanical steering + phase array[1]phase array, micro-sec switching[6]
Cost$100-$300 depending on capability[6]
.
Network Interconnect
Uplink bandwidth max
Uplink bandwidth typical
Downlink bandwidth max
Downlink bandwidth typical
Antenna size
Antenna type
Location(s)
.
Development
Prime ContractorSpaceX
Design site?Seattle
Design engineers (Jan 2015)60[2]
Manufacturing sitemost likely US[1]First 10 Toulouse, rest USA[11]US[2]
.
Operations
ITU frequency allocationyesapplied for[6]
Ground link legalitycountry by country[6]
Sales organisation
.
Financing
Total funding$1.5B - $2B[1], $2B initial sat batch[5]$10B - $15B[2]
Initial constellation funding$1.7B from COFACE[12]?
Initial funding fromVirgin & Qualcomm[4]SpaceX[2]
Funding during designcommitments to $500M[12]$1B Google and Fidelity[7], SpaceX?
Majority funding??
.

References
[1]Space News http://spacenews.com/virgin-qualcomm-invest-in-global-satellite-internet-plan/
[2]@malderi twitter http://forum.nasaspaceflight.com/index.php?topic=36552.msg1317801#msg1317801
[3]Business Week http://www.businessweek.com/articles/2015-01-17/elon-musk-and-spacex-plan-a-space-internet#p1
[4]OneWeb press release http://www.businesswire.com/news/home/20150115005196/en/OneWeb-Announces-Plans-Launch-Satellite-Constellation-Bring#.VLuTCkesU3l
[5]CNBC http://www.cnbc.com/id/102340448
[6]Launch event video http://forum.nasaspaceflight.com/index.php?topic=36552.200
[7]SpaceX http://www.spacex.com/press/2015/01/20/financing-round
[8]FierceWirelessTech http://www.fiercewireless.com/tech/story/5g-proceeding-spacex-urges-fcc-protect-future-satellite-ventures/2015-02-22
[9]https://twitter.com/pbdes/status/575287187153010690
[10] https://twitter.com/pbdes/status/577853009373319181
[11]http://airbusdefenceandspace.com/newsroom/news-and-features/airbus-defence-and-space-selected-to-partner-in-production-of-oneweb-satellite-constellation/
[12]http://tmfassociates.com/blog/2015/06/11/round-1-to-wyler
Notes:
Strange that there is no announcement/press release from Qualcomm
« Last Edit: 06/15/2015 01:20 pm by MikeAtkinson »

Offline Saltvann

  • Member
  • Posts: 18
  • Liked: 1
  • Likes Given: 38
Re: SpaceX and OneWeb internet satellite comparison
« Reply #1 on: 01/18/2015 08:43 pm »
It's interesting that you have a source giving 750km and one giving a different number of 1100km. It's worth noting that 1100km is actually around 750 miles. Perhaps the Businessweek article got the units mixed up?

Offline MikeAtkinson

  • Full Member
  • ****
  • Posts: 1980
  • Bracknell, England
  • Liked: 784
  • Likes Given: 120
Re: SpaceX and OneWeb internet satellite comparison
« Reply #2 on: 01/18/2015 08:45 pm »
It's interesting that you have a source giving 750km and one giving a different number of 1100km. It's worth noting that 1100km is actually around 750 miles. Perhaps the Businessweek article got the units mixed up?

My mistake, Business Week said 750 miles.

Offline meekGee

  • Senior Member
  • *****
  • Posts: 14158
  • N. California
  • Liked: 14046
  • Likes Given: 1392
Re: SpaceX and OneWeb internet satellite comparison
« Reply #3 on: 01/19/2015 06:57 am »
You can add "cost to launch entire constellation"

For Wyler, it is 648 * $10M, or $6.5B


For Musk, it is 4025 / <sats-per-launch> * <FH reusable cost>

Since we don't know the numbers, using a WAG of 64 planes * 64 satellites, and each FH carrying one plane's worth (plus a dispenser, and maybe spares), so a total of 64 launches at <$80M, or $5.1B.

Less if reusability works well.

 
« Last Edit: 01/19/2015 07:07 am by meekGee »
ABCD - Always Be Counting Down

Offline Oli

  • Senior Member
  • *****
  • Posts: 2467
  • Liked: 605
  • Likes Given: 60
Re: SpaceX and OneWeb internet satellite comparison
« Reply #4 on: 01/19/2015 07:54 am »
So if I understand correctly there will only be communication with the ground and not between sats, I don't know how one can achieve across the globe latency of 20ms in such a case (like Musk said). When it comes to the throughput, with a bandwidth of 8gb/s per sat OneWeb could provide 50mb/s to ~160 users per "cell".

An interesting part of the spacenews article:

Quote
“We did it with O3b and we’ll do it again with OneWeb,” Wyler said, stressing that the O3b business of delivering high-throughput to corporate and government users has few overlaps with the lower-speed OneWeb, whose market is centered on individual consumers and airlines.

Attached images of the O3b (MEO) and OneWeb (LEO) networks below.

« Last Edit: 01/19/2015 08:24 am by Oli »

Offline ChrisWilson68

  • Senior Member
  • *****
  • Posts: 5266
  • Sunnyvale, CA
  • Liked: 4992
  • Likes Given: 6459
Re: SpaceX and OneWeb internet satellite comparison
« Reply #5 on: 01/19/2015 08:08 am »
Here is a comparison of the SpaceX/Musk and OneWeb/Wyler global internet satellite constellations. I'll update it as more information becomes available.

FeatureOneWebSpaceX
Financing
Total funding$1.5B - $2B[1], $2B initial sat batch[5]$10B - $15B[2]
Initial constellation funding??
Initial funding fromVirgin & Qualcomm[4]SpaceX[2]
.

This table fails to reflect the most important difference: The SpaceX operation will be part of a multi-billion dollar company, led by a multi-billionaire.  OneWeb is a tiny company.  The fact that they've gotten small investments (so small they're too embarrassed to say how big they are) from two larger companies doesn't make up the difference.  Many small companies get investments from other companies, particularly companies that are trying to sell them something.  It's not a coincidence that Virgin is a company desperate to sell small satellite launch service and Qualcomm is a telecommunications hardware company.  Investments by vendors are generally little more than a way to pay a small deposit to get the foot in the door on the off chance that a little company is a success.

The SpaceX operation has access to two orders of magnitude more funding.

Offline MikeAtkinson

  • Full Member
  • ****
  • Posts: 1980
  • Bracknell, England
  • Liked: 784
  • Likes Given: 120
Re: SpaceX and OneWeb internet satellite comparison
« Reply #6 on: 01/19/2015 08:26 am »
You can add "cost to launch entire constellation"

For Wyler, it is 648 * $10M, or $6.5B


For Musk, it is 4025 / <sats-per-launch> * <FH reusable cost>

Since we don't know the numbers, using a WAG of 64 planes * 64 satellites, and each FH carrying one plane's worth (plus a dispenser, and maybe spares), so a total of 64 launches at <$80M, or $5.1B.

Less if reusability works well.

I won't be adding computed values to the table, they make too many assumptions. Here you make the assumptions of single sats on LauncherOne for the OneWeb constellation (where Wyler has said they will design a multi-satellite dispensor), for the SpaceX constellation we don't know the number of planes or that FH will be used to launch the satellites.

Offline MikeAtkinson

  • Full Member
  • ****
  • Posts: 1980
  • Bracknell, England
  • Liked: 784
  • Likes Given: 120
Re: SpaceX and OneWeb internet satellite comparison
« Reply #7 on: 01/19/2015 08:31 am »
So if I understand correctly there will only be communication with the ground and not between sats, I don't know how one can achieve across the globe latency of 20ms in such a case (like Musk said).

As far as I can tell from the statements both networks will have inter-satellite links at least in their eventual configurations. It is possible that initial (test) sats may not have those links or that they are not required to work to give a usable constellation.

Offline MikeAtkinson

  • Full Member
  • ****
  • Posts: 1980
  • Bracknell, England
  • Liked: 784
  • Likes Given: 120
Re: SpaceX and OneWeb internet satellite comparison
« Reply #8 on: 01/19/2015 08:48 am »
Here is a comparison of the SpaceX/Musk and OneWeb/Wyler global internet satellite constellations. I'll update it as more information becomes available.

FeatureOneWebSpaceX
Financing
Total funding$1.5B - $2B[1], $2B initial sat batch[5]$10B - $15B[2]
Initial constellation funding??
Initial funding fromVirgin & Qualcomm[4]SpaceX[2]
.

This table fails to reflect the most important difference: The SpaceX operation will be part of a multi-billion dollar company, led by a multi-billionaire.  OneWeb is a tiny company.  The fact that they've gotten small investments (so small they're too embarrassed to say how big they are) from two larger companies doesn't make up the difference.  Many small companies get investments from other companies, particularly companies that are trying to sell them something.  It's not a coincidence that Virgin is a company desperate to sell small satellite launch service and Qualcomm is a telecommunications hardware company.  Investments by vendors are generally little more than a way to pay a small deposit to get the foot in the door on the off chance that a little company is a success.

The SpaceX operation has access to two orders of magnitude more funding.

I agree that investments in OneWeb are at present small, whether they are embarrassed about them or not is speculation, and I want to avoid speculation here as much as possible. OneWeb say that they will appoint a prime contractor in a couple of months, by then they will have to have the financing lined up (IMO most of the financing won't come from Virgin or Qualcomm - your speculation about foot-in-the-door financing is probably correct). OneWeb having financing lined up does not mean that much is needed now, the major expenditures are still a couple of years away.

SpaceX has access to internal revenue, but nowhere near enough to finance a $10B. They will need to go for outside funding of some sort, but probably not this year or next.

Offline Oli

  • Senior Member
  • *****
  • Posts: 2467
  • Liked: 605
  • Likes Given: 60
Re: SpaceX and OneWeb internet satellite comparison
« Reply #9 on: 01/19/2015 09:08 am »
So if I understand correctly there will only be communication with the ground and not between sats, I don't know how one can achieve across the globe latency of 20ms in such a case (like Musk said).

As far as I can tell from the statements both networks will have inter-satellite links at least in their eventual configurations. It is possible that initial (test) sats may not have those links or that they are not required to work to give a usable constellation.

Thanks.

As I see it the only advantage a MEO or LEO constellation has over GEO sats is the latency. GEO sats can service specific regions with high demand (unless LEO sats which fly over the ocean and uninhabited territory most of the time) and do not need adjustable ground antennas (unless they're on planes and ships).

I agree that investments in OneWeb are at present small, whether they are embarrassed about them or not is speculation, and I want to avoid speculation here as much as possible.

Wyler has brought SES aboard for O3b, so I suppose he has some experience in finding investors for his plans.

Offline guckyfan

  • Senior Member
  • *****
  • Posts: 7438
  • Germany
  • Liked: 2332
  • Likes Given: 2891
Re: SpaceX and OneWeb internet satellite comparison
« Reply #10 on: 01/19/2015 10:24 am »
Why would you assume Falcon Heavy for launch. It is more likely IMO that fairing size is the limiting factor and they will use Falcon 9.

I would also assume only 2, maybe 3 inclinations to launch into. For coverage of the whole globe they can use phasing, easily done for different trajectories from one launch. I understand that initially the polar region would not be served. In that case I am confident, 2 inclinations are enough.

I also see no reason why OneWeb would not use Falcon for launch, with their dispensers. They already said, that the Virgin launcher is for replacements.


Offline QuantumG

  • Senior Member
  • *****
  • Posts: 9238
  • Australia
  • Liked: 4477
  • Likes Given: 1108
Re: SpaceX and OneWeb internet satellite comparison
« Reply #11 on: 01/19/2015 10:53 am »
So if I understand correctly there will only be communication with the ground and not between sats, I don't know how one can achieve across the globe latency of 20ms in such a case (like Musk said).

I got the exact opposite impression. i.e., he said

Quote from: Elon Musk
Let's say you want to communicate from Seattle to South Africa. If you look at the actual path it takes, that's extremely convoluted. It'll follow the outline of the continents. It'll go through 200 routers and repeaters and the latency is extremely bad. Whereas, if you did it with a satellite network, you could actually do it in two or three hops. Well, maybe four hops. It depends on the altitude of the satellites and what the capacity and things are. But basically, let's say, at least an order of magnitude fewer repeaters or routers and then going through space at 50% faster speed of light. So it seems from a physics standpoint inherently better to do the long distance Internet traffic through space - source

Sounds like space-to-space to me.
« Last Edit: 01/19/2015 11:11 am by QuantumG »
Human spaceflight is basically just LARPing now.

Offline meekGee

  • Senior Member
  • *****
  • Posts: 14158
  • N. California
  • Liked: 14046
  • Likes Given: 1392
Re: SpaceX and OneWeb internet satellite comparison
« Reply #12 on: 01/19/2015 11:08 am »
Fair enough about not using computed values.

Still, whether in the table or not, we should track launch costs.

FH makes sense since it is cheaper per pound (cost, not price!) in the reusable mode, and since even one plane is heavy enough for it.  Of course if they choose to, they can also launch to multiple inclinations, it is just extra dv.
ABCD - Always Be Counting Down

Offline MikeAtkinson

  • Full Member
  • ****
  • Posts: 1980
  • Bracknell, England
  • Liked: 784
  • Likes Given: 120
Re: SpaceX and OneWeb internet satellite comparison
« Reply #13 on: 01/19/2015 11:29 am »
So if I understand correctly there will only be communication with the ground and not between sats, I don't know how one can achieve across the globe latency of 20ms in such a case (like Musk said).

As far as I can tell from the statements both networks will have inter-satellite links at least in their eventual configurations. It is possible that initial (test) sats may not have those links or that they are not required to work to give a usable constellation.

Thanks.

As I see it the only advantage a MEO or LEO constellation has over GEO sats is the latency. GEO sats can service specific regions with high demand (unless LEO sats which fly over the ocean and uninhabited territory most of the time) and do not need adjustable ground antennas (unless they're on planes and ships).

Latency is one problem with GEO sats. Another is that they are to address anything other than a niche market they have to be very large (antenna, power bandwidth).

Anyway GEO and MEO constellations are outside the scope of this topic, which addresses the LEO constellations of SpaceX and OneWeb.

Offline MikeAtkinson

  • Full Member
  • ****
  • Posts: 1980
  • Bracknell, England
  • Liked: 784
  • Likes Given: 120
Re: SpaceX and OneWeb internet satellite comparison
« Reply #14 on: 01/19/2015 11:32 am »
I would also assume only 2, maybe 3 inclinations to launch into.

As far as I know other large LEO constellation have used a single inclination. There has been no information released on the OneWeb  and SpaceX inclinations that I have seen.
« Last Edit: 01/19/2015 11:33 am by MikeAtkinson »

Offline quanthasaquality

  • Full Member
  • *
  • Posts: 146
  • Liked: 6
  • Likes Given: 2
Re: SpaceX and OneWeb internet satellite comparison
« Reply #15 on: 01/19/2015 11:50 am »
I guess this is the successor thread to   "SpaceX - now a satellite vendor?"

This reminds me of 2 approaches to wireless in a city. Use some big cell phone towers, or use lots of 802.11 hot spots. The big cell phone towers strategy won out.

I don't think SpaceX has the ability to design a wireless data transmission standard, or implementation. Qualcomm, which is backing OneWeb, probably does. Now, if SpaceX teamed up with Microsoft, for the money, and maybe programming expertise, and Nokia, for the wireless stuff, then yeah, I bet it could work. Maybe Elon's going to be stealthy, and get backing from Paul Allen, and steal away employees, for SpaceX Helsinki.

As for long-haul internet traffic by satellite, I don't think it will be competitive. A fiber backbone is like a high speed rail line. It is very expensive to build, but once you have it, you can send lots of people, with a low per unit cost. I guess today, a single optical fiber can transport a few dozen different frequencies of light. Each frequency of light has a bandwidth of several gigabits/sec. The physical tunnel, that the physical right of way is needed for, could have over 1 thousand optical fibers. That comes out to about a few hundred terabits/sec. If you want to get really hardcore, you could stick 100 thousand optical fibers into the Chunnel. I don't think anyone needs that kind of bandwidth, but it can be done.


I guess Elon has everything to win by doing a vaporware launch of a giant satellite internet constellation. If Elon can get the satellite internet hype going strong enough, someone with money might bite, to build any satellite internet constellation. That will mean more launches. The only one out there that can launch satellites at a cost close to SpaceX, is Ukraine/Russia, so probably more business for SpaceX.

Offline MikeAtkinson

  • Full Member
  • ****
  • Posts: 1980
  • Bracknell, England
  • Liked: 784
  • Likes Given: 120
Re: SpaceX and OneWeb internet satellite comparison
« Reply #16 on: 01/19/2015 11:58 am »
FH makes sense since it is cheaper per pound (cost, not price!) in the reusable mode, and since even one plane is heavy enough for it.  Of course if they choose to, they can also launch to multiple inclinations, it is just extra dv.

I think that fairing volume will be the limiting factor, possibly 20-30 satellites per launch (which would be 6-9 tonnes for 300 kg satellites).

FH payload to LEO when reusing all the 1st stage cores has not been published by SpaceX. The cost and price are unknown for such a large multi-launch order either for F9R for FHR, it is in my opinion unwise to assume that either one or other is cheaper for this application.

It is also unwise in my opinion to assume that the satellite division will be charged less than external customers for launches. They could charge the same for many reasons, not least because the satellite division may be spun off into a separate company.

Offline MikeAtkinson

  • Full Member
  • ****
  • Posts: 1980
  • Bracknell, England
  • Liked: 784
  • Likes Given: 120
Re: SpaceX and OneWeb internet satellite comparison
« Reply #17 on: 01/19/2015 12:00 pm »
I guess this is the successor thread to   "SpaceX - now a satellite vendor?"

I want this to be a thread specifically comparing the two systems, using hard information rather than speculation.

IMO there should be two other threads devoted to discussion and speculation about the individual systems.

Offline Oli

  • Senior Member
  • *****
  • Posts: 2467
  • Liked: 605
  • Likes Given: 60
Re: SpaceX and OneWeb internet satellite comparison
« Reply #18 on: 01/19/2015 02:55 pm »
I don't think SpaceX has the ability to design a wireless data transmission standard, or implementation.

The system is not targeting mobile, that should be clear from Musk statements. I suppose the market will be developing countries with little fiber infrastructure as well as the airline/shipping industry. Cars could also be a candidate, but then you have again the wall penetration problem (tunnels, cities, mountains (?)). In sparsely populated regions without cell coverage it could be interesting. I'm convinced too that it cannot compete with fiber in the long term, but not everybody has fiber to his house like I do :). It might take a while for developing countries to catch up.

So if I understand correctly there will only be communication with the ground and not between sats, I don't know how one can achieve across the globe latency of 20ms in such a case (like Musk said).

I got the exact opposite impression. i.e., he said

Quote from: Elon Musk
Let's say you want to communicate from Seattle to South Africa. If you look at the actual path it takes, that's extremely convoluted. It'll follow the outline of the continents. It'll go through 200 routers and repeaters and the latency is extremely bad. Whereas, if you did it with a satellite network, you could actually do it in two or three hops. Well, maybe four hops. It depends on the altitude of the satellites and what the capacity and things are. But basically, let's say, at least an order of magnitude fewer repeaters or routers and then going through space at 50% faster speed of light. So it seems from a physics standpoint inherently better to do the long distance Internet traffic through space

Sounds like space-to-space to me.

Well you could do the hop over a ground station if there is cell overlap between sats. But I agree that would be silly for a LEO constellation.

« Last Edit: 01/19/2015 03:02 pm by Oli »

Offline meekGee

  • Senior Member
  • *****
  • Posts: 14158
  • N. California
  • Liked: 14046
  • Likes Given: 1392
Re: SpaceX and OneWeb internet satellite comparison
« Reply #19 on: 01/19/2015 03:25 pm »
FH makes sense since it is cheaper per pound (cost, not price!) in the reusable mode, and since even one plane is heavy enough for it.  Of course if they choose to, they can also launch to multiple inclinations, it is just extra dv.

I think that fairing volume will be the limiting factor, possibly 20-30 satellites per launch (which would be 6-9 tonnes for 300 kg satellites).

FH payload to LEO when reusing all the 1st stage cores has not been published by SpaceX. The cost and price are unknown for such a large multi-launch order either for F9R for FHR, it is in my opinion unwise to assume that either one or other is cheaper for this application.

It is also unwise in my opinion to assume that the satellite division will be charged less than external customers for launches. They could charge the same for many reasons, not least because the satellite division may be spun off into a separate company.
I'm taking this over to the regular thread then.
ABCD - Always Be Counting Down

Tags:
 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
0