Hacker News Reader

Comment by 1024core | Hacker News Reader

They conveniently tested it in SoCal.

> Specifically, rain and humidity could result in very high attenuation for terrestrial MMW links and could severely compromise their availability.

I imagine this will be useless in most equatorial parts of the world, and probably even south-east US.

View on HN

Baby steps. They weren't sure if they could even fly a solar plane, so they figured that out first. Then they weren't sure if they could even communicate with the thing once it's up in the air, so they've just figured that part out. Next step will be to figure out if they can fly and communicate in different weather conditions. And maybe they won't figure that out, but at least they didn't "imagine it was useless" and not even try.

Sure, but there are plenty of parts of the world where this actually would be extremely useful. Complaining that a new technology isn't useful in 100% of cases is such an unhelpful position to take. Finding a solution that works in 10% of the world means the next solution only needs to worry about 90%. No one ever claimed that there would be one single solution to the problem they're attacking.

It's unfortunate that they didn't delve deeper into the atmospheric attenuation since the whole point of the project is to deliver internet from UAVs. Seems like the next test should involve a foggy day in San Francisco instead of a "air-to-ground bidirectional link".

some tangentially related systems have fallbacks to different wavelengths that penetrate better in sub-optimal conditions. I don't know if it's applicable here, but I imagine it could be.

In recent years they substantially fixed Ku-Band communications in poor weather with some new algorithms and consequent hardware/protocols. I suspect this would be the next step for this technology.

Not really. The fundamental problem is still there where the attenuation is high at certain frequencies in C/Ku/Ka. The way they mitigate it is to use different modulations and code rates in each frame so you don't need to bring the quality of the entire link down due to one user. However, the tradeoff is that the system capacity goes down fairly sharply as a result of this.

I guess you missed the part where they say they set up a duplicate at a data Center that gets a lot of rain (Prineville, OR in this case)

I didn't read the statement where they discuss this, but Prineville, OR does not get a lot of rain. It is in the eastern Oregon high desert, very near Bend, and it gets more than 200 days per year that are sunny or mostly sunny, the highest number of clear sky days in the state.

Reply on news.ycombinator.com