The problem is actually that most of those systems are on considerably less advanced tech than Starlink and as a result only have a fraction of the capacity.
Thus each piece of capacity costs more.. thus the very high costs of Iridium and Viasat.
Their profit margins aren't actually that good because their costs are so high compared to their capacity and the costs are so high that demand simply doesn't materialise - people just do without.
Starlink will change this game because of their drastically increased capacity (assuming they get sat-sat links working). Until another mega-constellation comes online I fully expect them to do to satellite Internet what they did to the launch market.
Iridium was the first network to successfully operate inter-satellite links more than 20 years ago. I wouldn't call it "considerably less advanced".
Iridium and Starlink operate in completely different bands (L vs Ka), with orders of magnitude more bandwidth available on Ka than on L. That (and the fleet size, which differs by 2-3 orders of magnitudes) is where the significantly lower bandwidth stems from.
Iridium was a marvel of engineering no-doubt but Starlink as a constellation is in a different league entirely.
They also followed different engineering principles, Iridium was engineered to perfection while Starlink is hacked until it works and paper over the kinks with sheer scale and many iterations.
Worth mentioning I have nothing but admiration for the original Iridium constellation, it was decades ahead of it's time, it's just so unfortunate they didn't see that broadband needs were going to dominate communications in time or maybe Motorola in it's original glory would still be around today.
big difference also is that iridium was designed from the outset to address handheld size customer terminals like an original iridum 9505 phone, starlink is something that needs a considerably larger fixed phased array. rf laws of physics being what they are I don't see a ku/ka-band starlink handheld phone or anything similar to the iridium L/S-band hockey puck sized antennas any time soon.
the business model was clearly not the best, as witnessed by its spectacular bankruptcy, and the acquisition by the 2nd corporate incarnation of iridium.
Starlink has plenty of room for increasing their antenna size on the satellite size to something that a handheld could use. I was surprised how small the phased array on the sat side is. Increase that to make it similar to the size of the satellite (not counting solar array) and with the larger v2 satellites, and they'd have much more gain (allowing the handheld antenna to be smaller). They essentially already use 5G bandwidth.
Isn‘t part of what allows much hire bandwidths for Starlink due to spatial frequency reuse, which requires mandatory beamforming on both sides of the connection?
There are many satellites overhead at any given time (the current LEOs and MEOs usually have just one), so terminals need to be able to limit their gain to a small angle not only for gain reasons. Steering in handheld applications seems very difficult.
Of course, they could dedicate parts of their spectrum to "one sided beamforming" and compete with Iridium that way (no idea if the L-band has any other advantages over Ka when used like that).
Actually in exactly the same way. The nice thing about phased arrays is they steer electronically, so you can aim them extremely fast and compensate for any movement. That's how 5G works and that's also how Starlink on moving vehicles works.
a flat panel phased array still has considerably more gain in the direction it's aimed, vs a randomly oriented omnidirectional (physical) antenna in a phone. i can guarantee you starlink doesn't work well when the array elements aren't aimed correctly in the general direction of where the satellites are.
the path loss and need for more gain at LEO satellite distance is considerably greater than talking about terrestrial cellphone networks in bands <2500 MHz.
5G phones already use a flat panel phased array, just with fewer elements (and less gain, but that can be compensated by using a lot more elements on the satellite side and by operating at much lower bandwidth). And yes, I understand the difference between lower frequencies and higher. 5G can also use lower frequencies but the main bandwidth advantage (over LTE/etc) is in the higher frequency, the mm-wave, roughly 25-50GHz, a frequency band that overlaps that used by the Starlink phased arrays.
5G and Wi-Fi don't use MIMO for terminal side spatial multiplexing though, as far as I know.
In MU-MIMO (available in newer versions of 802.11, for example), the base station transmits to and receives from multiple mobile devices that are located at different spatial angles. But in the mobile devices themselves, the gain from that steering is not high/fast enough to allow for multiple (relatively) fixed base stations in the same space.
In other words, a stationary/slowly moving antenna array on one side of the channel can target individual moving users due to angles changing slowly over time and the antenna array being quite sophisticated, but moving users with smaller antennas and angles varying over a much shorter period of time can't do the same.
Think about it: From a satellite's point of view, your angle varies much more slowly and predictably than the other way around, e.g. when taking a turn in a car or on a rocking boat (that's why high-gain antennas usually have to be gimbal-stabilized).
Starlink's orders of magnitude higher bandwidth, as far as I can tell, stems from requiring slowly and predictably changing angles on both sides, giving it m:n (relatively) independent spatial channels rather than just 1:n as is common for a one-sided omnidirectional approach. (That's also why Iridium and Globalstar wouldn't be able to scale significantly better in low-gain mobile-client applications by just launching more satellites.)
Sorry to interject between two people who clearly know what their talking about.
I've been reading books on satellite communications and have some idea of all the technical terms you guys mentioned as well as relationship between 5g and phase array etc.
Where do you get your knowledge from regarding these issues in such detail? Do you guys work in this area? Are there any reading materials you recommend in this field that are relevant and updated to keep track of these tech?
this is completely incorrect. viasat's technology is far more advanced than starlink. just because it's not LEO doesn't mean it's simple.
and starlink doesn't have dramatically increased capacity. they have a moderate increase over the busy areas geo incumbents have. most of their capacity is over water.
The end results of the link budgets required for geostationary to low cost consumer grade terminals means that now matter how good viasat's tech is, the end result is going to be mediocre. Have you actually lived on the far end of a consumer grade highly contended oversubscribed ku/ka band vsat terminal?
Viasat's modems and rf chain stuff are about as good as can be expected within the very constrained BOM budget, antenna gains and link budgets involved.
I've had their service for 4 years straight in the past. it's not as good as cable, but it's nowhere near as bad as you say if you're not playing real time games.
the constrained BOM is the very reason why the terminal cost is reasonable for this business. it's well known that SpaceX is selling the terminal at a huge loss to gather customers in the short term.
Even if viasat's consumer service was handing out high quality steel 1.8 meter elliptical offset ku band dishes with norsat PLL LNBs and 8W BUCs to consumers, the path loss and modulations required would mean that a given section of contended (let's say, 10MHz of a transponder) service to many terminals would still have significantly lower speeds and greater oversubscription to be economically viable compared to what end users see right now on starlink.
that's simply not true. they have transponders that should deliver 10Gbps to a single user IF they were willing to pay for both the service and terminal. SpaceX is limited in the exact same way, and their terminals + satellite costs are much higher for the number of them.
10Gbps to a single terminal using exactly what modulation and code rate, and channel size?
Or course you can achieve high speeds over geostationary if you throw entire transponders full of MHz at the problem. Which has a proportionally huge monthly recurring cost to control that transponder space. Or as a total percentage of the satellite's transponders dedicated to your network.
Yes satellite tx power from the twta is more powerful than it used to be. You still need a huge ass earth station to start doing 16apsk/32apsk reliably.
I'm not sure what point you're trying to make here. starlink is also multiple smaller beams within a single beam for frequency reuse. a single beam does not get more bandwidth (in Gbps) than a viasat terminal.
> You still need a huge ass earth station to start doing 16apsk/32apsk reliably.
16apsk has been used for about a decade now on GEO satellites. I'm not sure why the gateway size matters when we're talking about user speeds.
If you don't understand why in a geostationary based link, the antenna size at either end matters a great deal for overall link budget, gain, Eb/No and what modulation you can use at a certain MCS and code rate before it gets too blurry in a QAM eyeball chart, there's no sense in talking about satellite communications with you.
I note you didn't answer my question because you have no idea of what modulations, channel size and such are required to actually push 10Gbps through a satellite link to a single terminal. I think you're going off Viasat's marketing material where they're claiming aggregate throughput of an entire satellite or something.
Show me the exact hardware configuration of modems you think are capable of 10Gbps by geostationary and how much transponder MHz it needs.
I literally just said that 16APSK has been used for a decade on Geo satellites. I don't know if you aren't aware what modulation means, but that is the modulation. it does not matter at all what size the gateway or the user antenna is. 16APSK (actually 32APSK is used as well) and the carrier size tells you all you need to know about the speeds, unless you've never done a link budget or worked in this industry.
I said if they wanted to they could sell a 10Gbps plan, because that's what the link budget allows. they don't, of course, because it's not profitable. SpaceX also could in theory, do that, but they also don't. your original claim is that SpaceX somehow has far greater bandwidth to an area than Geo is patently false.
> the path loss and modulations required would mean that a given section of contended (let's say, 10MHz of a transponder) service to many terminals would still have significantly lower speeds and greater oversubscription to be economically viable compared to what end users see right now on starlink.
if your definition of economically viable is getting VC money and government subsidies pumped in so that someone else other than the subscriber is paying for the service, then I agree with you. but that's not the metric most people use when they discuss being profitable
Thus each piece of capacity costs more.. thus the very high costs of Iridium and Viasat.
Their profit margins aren't actually that good because their costs are so high compared to their capacity and the costs are so high that demand simply doesn't materialise - people just do without.
Starlink will change this game because of their drastically increased capacity (assuming they get sat-sat links working). Until another mega-constellation comes online I fully expect them to do to satellite Internet what they did to the launch market.