Software, technology, sysadmin war stories, and more. Feed
Tuesday, December 18, 2012

Augmenting satellite broadcasts with in-car data links

I like coming up with ideas which also have "oh yeah, that would be useful" side-effects. Here's one that I hope has already been considered by the people who are responsible for the future of satellite radio.

Right now, you pretty much have a one-way stream from one or more satellites down to the receiver in your car. I believe they do some interesting time-spreading stuff so that you can actually lose the entire feed for a second or two and not hear the gap in your music just because you drove under a bridge.

Still, there are a few areas where I can pretty much guarantee that it's going to drop out every time I drive through. One of them is where 237 crosses highway 85 in Mountain View. I'm not sure if it's tree cover, some particularly tall obstruction to the south, or just something local splattering all over the band, but my radio always skips in that area. There are others like this around the Bay Area, and they're all mildly annoying because they're predictable. Some of them seem to be correlated to cell towers, but not all of them are immediately obvious.

I assume that most cars in the future will come with their own "always on" data service which is used to do things like "OnStar" and possibly send down fresh traffic data. It would be pretty interesting if the radio could use that pipe to request a quick "fill" for its buffer to keep you from hearing a gap.

The general concept here is that having all of the cars streaming music over their glorified telemetry channel wouldn't scale, so the satellite broadcasts would handle most of that. It would just be for little bits and blobs here and there when there's a localized signal disruption. That shouldn't introduce nearly as much load to the system.

The neat side effect I mentioned earlier comes from the realization that these cars probably would have a halfway decent GPS fix at all times. It becomes a "simple matter of programming" to just have it report its location any time it loses the satellite signal. This would turn into a swarm of signal strength issue reports at the provider, and then they'd know exactly where to set up a repeater or go look for leaky or unauthorized transmissions in their band.

It would be in the satrad company's best interest to deploy localized repeaters in the hottest spots to cut down on the telemetry system's load. Every byte they can serve on their usual service maps to many more which don't need to be pushed over the car to car links in that area.

If done correctly, the actual customers wouldn't have to know about the acrobatic feats of data-flinging that these systems are performing behind the scenes. They can just turn on their music and rock out.