Slashdot Mirror
Elon Musk's Next Mission: Internet Satellites
An anonymous reader writes: According to a report in the Wall Street Journal, Elon Musk is looking at a new project: smaller, cheaper satellites that can provide internet access for people all across the world. "Mr. Musk is working with Greg Wyler, a satellite-industry veteran and former Google Inc. executive, these people said. Mr. Wyler founded WorldVu Satellites Ltd., which controls a large block of radio spectrum. In talks with industry executives, Messrs. Musk and Wyler have discussed launching around 700 satellites, each weighing less than 250 pounds, the people said. That is about half the size of the smallest communications satellites now in commercial use. The satellite constellation would be 10 times the size of the largest current fleet, managed by Iridium Communications Inc. ... The smallest communications satellites now weigh under 500 pounds and cost several million dollars each. WorldVu hopes to bring the cost of manufacturing smaller models under $1 million, according to two people familiar with its plans."
These will not be high latency. If you have 700 satellites more-or-less evenly distributed around the globe (say from 60S to 60N latitude) and you want a minimum of 45 degree elevation to the nearest satellite, they can be lower than 400 miles altitude, or 600 miles away. Assuming that the system will bounce signals from the satellites to a distributed network of fiber connected ground stations, latency should only be 10ms more than a pure cable transmissions.
Previous satellite internet to geosynchronous satellites are nothing like this.
I agree with other commenters that this is pretty unlikely, but SpaceX and Tesla were quite unlikely to succeed as well.
I love Mondays. On a Monday, anything is possible.
This suffers from the same problems that Iridium had:
* The people in the world with money to buy this already have good Internet access.
* The system doesn't work until it's global: you need to pay for the entire system before you get customers.
Land-based networks can build out a region at a time, starting in the wealthiest areas, creating paying customers who provide the capital for the next phase of expansion. Satellite systems are egalitarian, which sounds nice but is a problem: if you need 700 satellites to cover the globe but can only afford 350, you get global coverage that only works half the time, which nobody wants to pay for. And you have to set your asking price lower than what the poorest community that can't afford cell service can pay, which is a very low limbo bar to get under, and getting lower all the time.
Nope, read up on the Iridium system, which already exists. You don't need satellite tracking, you can use an omnidirectional antenna to communicate with low earth orbit. You just need more power.
That said, Iridium ping times are horrible, but that's more a function of 1980s technology than the speed of light or information theory.