Hacker News Reader

So... the easy part of this is the resolution of the focal plane array. Gigapixel CMOS is easy/cheap. Then there’s the diffraction limit of the optics. They’ll have to be huge to get a small GSD. But the biggest physical limitation is always getting the data back down to earth. How have you grown a link budget to match your resolution?

View on HN

Great observation! One of the biggest challenges for high resolution satellite imaging, even beyond the optics, is getting all those pixels down. Fortunately, the ground station infrastructure has blown up recently with Big Tech offerings like AWS Ground Station and Azure Orbital, plus legacy providers like KSAT, so there is plenty of access to get data down through Ka-band.

Just getting a rough idea here: 10cm gsd at 200km elev with 500nm wavelength requires a 1m aperture to hit the diffraction limit. If you use 40dbi phased arrays at 100W you get -40dbm at the receiver and hit the noise floor at about 100gbps, which is 40km2 of raw photos per second of downlink. So i guess it’s not whole earth observation, but definitely a lot of capability for pre-tasked observation. But then we have to ask, why satellites rather than aerial drones? A satellite is on a relatively fixed trajectory that hits any given target for 11 minutes of every 90, and the same with the ground station if you’re lucky enough to have both on the same trajectory. What is the use case where this is cheaper or superior (aside from no-fly zones)?

The people you're snooping on don't hear any buzzing quadcopter rotors overhead.

Reply on news.ycombinator.com