A constellation of ~ 40,000 :eek: low earth orbit satellites for internet. It should greatly improve latency issues for rural folks such as my self.
Apparently, the satellites are networked by laser, and as they orbit you are passed off from one to the next. Currently, my sat is in geosynchronous orbit some 22,000 miles above. These new sats are going to be much, much lower (370 miles) and thus should improve latency.
No debate, Just wonder if anyone else is following this, and your opinions.
I would like better internet now that I am working from home, but doubt it will come into fruition before I end up back in my cube. Not sure how all this is going to play out. I may be allowed to work from home from now on.
Star link is only one of several plans to deploy thousands of communications satellites in low earth orbit. I’m really surprised they got approval, since it’s going to cause so many problems for astronomers.
I can’t figure out how much data the satellites can handle.
40,000 satellites. I read somewhere each satellite can handle 20 Gbps. So the entire starlink network can handle 800 Tbps.
Which sounds like a lot, but thats about 10 Mbps speeds for 80 million people simultaneously. And there are times of day when internet demand is highest.
It’s probably going to be a while before we really understand the dynamics of their network. That said:
I believe the 20 Gbps figure was for the early satellites. Later ones are more like 100 Gbps. And they plan on further upgrades from that.
All ISPs oversubscribe their bandwidth. A 10x ratio is mild; many are more like 100x. So an average 10 Mbps per customer is still a great experience.
80 million people would be a crazy customer count. At $80/mo, that’s $77B/yr in revenue. And really, a lot more if you add expensive commercial and military contracts. The network isn’t going to be cheap, but that’s a heck of a lot of money. I don’t think even SpaceX is being that optimistic yet. 10M customers would still easily pay for the network.
The satellites are designed to burn up completely in the atmosphere. This was a significant design constraint, and part of the reason the current batch of satellites don’t have the laser interlinks is because the mirror system doesn’t yet meet this standard.
The satellites also have active propulsion: they’ll stay up as long as they have propellant. They can be disposed of manually, but will enter the atmosphere on their own if control is lost somehow.
OneWeb was going to be a major competitor, but they’ve gone bankrupt. Amazon has their Kuiper project but as usual with Bezos projects, it’s a complete mystery as of yet.
SpaceX has been pretty good about working with astronomers. Mostly the satellites won’t be visible once in orbit. And a few adjustments like changing the angle of the solar panels and installing a sunshade for the antennas should make a significant difference during deployment.
Kinda funny, one of our SysAdmins checked a coworkers ping rate at home and asked him if he was bouncing signals off the moon. No, only a 1/10 of the distance to the moon.
How big is each of these satellites, and how many can they pack into one launch? They can’t be too small, if they’re putting station-keeping thrusters on them (trying to imagine a CubeSat with a thruster).
They’re taking 60 up per launch and they’re small enough to jigsaw that many into a standard Falcon 9 fairing, although obviously with various panels and antennae stowed for transport.
And how many customers will they get who just dearly want to give the middle finger to their local cable company? I’m lucky that I have a choice between two of Satan’s minions, but most of us operate on the Soviet system–one choice, and no incentive to be better.
While I can see that rural people might appreciate low latency Internet at home, I’m not sure how much of an overall difference it’ll make or how pressing a need it is to have low latency Internet for home use in rural areas. I don’t mean to be dismissive but rural home use is not a use case for which there is likely to be much of a price premium or extensive demand.
Where there is unquenchable thirst and a lot of money to be made is smartphone bandwidth. Linking cell towers and being able to rearrange comms networks in real-time, that could be something. Orbiting at 600km and presuming a cell tower at 25m, that gives you a radio horizon of more than 3000km. If the cell towers communicate with satellites using lasers, that still gives you about 2800km of visual horizon. Even if, practically speaking, it’s just 1/3 of that, that’ll mean a lot of cell towers in possible direct contact with a lot of satellites which enables connections to be shifted quickly and flexibly. E.g.: If cell towers in London and Paris both see the same 1000 satellites and data consumption in Paris goes up comparatively with London, you can take satellite bandwidth from London and shift some of it to Paris. If Berlin then needs more than Paris, you can shift satlinks in real-time. Unless there’s something I’m not getting, you can do this all over the cell-tower networks and respond to spikes in demand in a way that’s fine-grained and instantaneous.
Sorry about double post, my 5 minutes were up: Not having to see WiFi hotspots as oases in a desert could change the way people use their pocket computers, one of the most ubiquitous and powerful forms of information technology humans have so far created. I can see that changing a lot but I don’t know how.
Also, drones could be more widespread and more capable if satellite/cell-tower connectivity gets faster/cheaper.
It could also be useful for various types of research if you can saturate any place on earth with sensors and have them upload a lot of data cheaply.
You can see there are two separate stacks of 30, and that they’re quite flat. They’re probably around 3.5x1.5x0.3 m in dimensions, and weigh 250 kg (approximate numbers are easy to figure out since they’re right at the limit of Falcon 9’s capabilities [in reusable mode], both in dimensions and mass).
Cubesat thrusters exist. The ones I know of take a significant fraction of a U, so a 1U is out except as a stunt, but a 3U or larger could certainly make use of one.
Starlink sats are much bigger, but still in the smallsat range. They use a Hall effect thrusters using krypton, not the usual xenon. Although somewhat lower performance, it is much cheaper per unit mass. And the satellites are so cheap (estimates are ~$250k each) that propellant costs are not negligible. I wonder if they’ll eventually switch to argon or even water as time goes on and their production costs decline further. Even krypton isn’t too cheap if your satellite is, say, $25k.
There are so many propulsion systems available for CubeSats and other small satellite systems that the annual SmallSat Conference dedicates an entire session to it (and it’s one of my favorite sessions even though it is often poorly attended). You can find propulsion systems that will fit into a fraction of a 1U form though anything that is warm gas or ionized is going to require enough power that it will practically only work in a 3U or larger envelope. For satellite constellations that require precision orientation or alignment, propulsion may be mandatory.
Water is a poor selection for a propellant because of the high heat of enthalpy. For a room temperature liquid monopropellant ion engine methanol or benzene would be a better choice despite the higher molecular weight, but for a Hall Effect thruster sublimating solid iodine is actually a good compromise of cost and performance.
