What inspired you to go down this route? Nobody has used millimeter-wave technology in this way.
Purely market conditions is probably a good way to describe it right now. If you draw the curve out and say, OK, X amount of bandwidth is necessary now ... either you can develop systems that are super spectrally efficient and try to go for, like, 100 Mbps ... or you can say, let's get higher in frequency, where you can get access to a ton more bandwidth.
You have other challenges, but spectral efficiency is just an incredibly difficult game to play. It comes down to innovating in base bands and radio, which is a much longer cycle. More importantly, you have no consumer industry support, compared to using existing consumer baseband technology and apply[ing] smart RF on different bands. That gives you a ton of agility on different strategies.
A lot of companies in the past have tried to do a different push from a radio-technology perspective. Ultimately you're seeing a lot of effort in millimeter wave today. The FCC is pushing it for 5G. Cellular guys have been looking at it from a 5G perspective. ... A month ago higher-level spectrum was opened up.
So you were ready for it?
We've been tracking it, we're not geniuses. We're good at looking at the set of things that are coming together both from a radio-technology perspective and other modulation improvements. The piece we're really adding is saying, How do we combine this into a front-end technology?
Two years ago we were starting to noodle where things were going. In the Aereo days we understood that at some point, the access network is going to be a choke point, either by technology or by politics.
It's interesting that you're aiming for unlimited bandwidth. That's something the wireless carriers offered when mobile broadband started happening, but ultimately they couldn't deliver. What makes you think you'll be able to do it?
Again, that's the spectral-efficiency issue. In all fairness, wholesale bandwidth isn't necessarily that large of a problem. [Carriers] essentially monetize non-wholesale -- really they plant costs by putting caps in. It's an opportunity cost question. If I start giving you unlimited bandwidth right now and if I need to build out my network better, where's the money going to come from?
We're playing in an area where the transport cost is largely de minimis, because we control both ends of the equation from a technology perspective. It makes a lot of sense to take that approach.
Will Starry be compatible with multiple routers? Or is the idea that you'll have to use the Starry Station?
Ideally ours, but we're not going to be a company that's going to restrict anyone from doing anything. [The company later told me it will give Starry Station and Point units to every subscriber.]
In terms of the range, how does Starry compare to a cellular base station? [Starry says their Beam stations can reach between 1 and 2 kilometers (0.6 to 1.2 miles).]
Cellular base stations tend to be much larger, although most LTE stations these days are getting built out in a couple of kilometer ranges, largely from a capacity-issue perspective. It depends on the market conditions. If they're in a semi-suburban or rural environment, cellular will reach miles.
Is Starry's network affected by walls, foliage and other obstructions?
Yes. It won't go through walls. In fact, it's very optical in its nature. It uses walls almost as specular surfaces, so [signals] bounce and almost create a multi-path, which helps with total capacity.
It looks like your window has to be slightly open to install Beam Point. Will that change?
It's kind of beta-ish right now. There will be a two-part solution that you can mount outside and bring in a coaxial cable.
How does weather affect reception? Will Starry equipment have issues with snow like satellite dishes?
There's actually a heater built into [Starry Point]. It puts about 25 watts out.
So how will rain affect things?
The modulation these days is so sophisticated, it will actually downgrade the connection ... but the flip is, since you're using millimeter wave the channels are much larger. So you don't care about spectral efficiency as much since your total throughput is still a function of the bandwidth that you get.
Will there ever be support for attaching something to a traditional outdoor TV antenna?
[Laughs.] No. But at the end of the day, look, a lot of UHF frequencies will get harvested for cellular connections. As they should be.