Receiver Design for the Future: Webinar Q&A

September 3, 2015  - By
Image: GPS World
Image: GPS World

The September article Receiver Design for the Future is based on a GPS World webinar, which sprang from a presentation at the Stanford PNT Symposium. Listener questions and Greg Turetzky’s answers during the webinar are provided below. Greg Turetzky is a principal engineer at Intel responsible for strategic business development in Intel’s Wireless Communication Group focusing on location. He has more than 25 years of experience in the GNSS industry at JHU-APL, Stanford Telecom, Trimble, SiRF and CSR.

 

Is dual frequency expected to be seen in smartphones this year?

If what you’re saying there is L1, L2, L5, the answer is absolutely not. But if what you’re saying is GPS, Galileo, BeiDou, which are all in different bands, then I think we are already seeing tri-bands between the GLONASS band, the GPS band and the BeiDou band. I expect to see that continue from a multi-constellation standpoint, rather than multi-frequency on individual constellations.

How do you see antenna design changing and developing relative to receiver design?

If you go back to that slide, the answer is the opposite — antenna design has been getting worse and worse in order to shave cost and size and is being made up for in silicon design. A really good example is, most GPS receivers that we build for mobile phones aren’t optimized to work at -140 dBm, our standard normal outdoor power, because we never see that. The antennas that we typically work with are 8, 10, 12 dB down, so no matter what, we never see anything above -140.

So I think the answer is the opposite — no one is trying in my market to make better antenna design, they’re trying to make them even smaller and even cheaper. Especially if you think about getting into wearables and button-sized things, you need a GPS antenna, a Bluetooth antenna and a Wi-Fi antenna in a button. That’s the problem, and still leaving room for performance. So basically we’re being asked to make up for that in the receiver design.

One of your slides showed GPS with 100% penetration, and SBAS was one of the next highest bars on that chart, outdated as it was, even though it’s only less than a year old. What are the benefits of SBAS in a commercial receiver?

The issue with that fact is we like geostationary satellites, because they’re easy to find and they’re useful from a visibility standpoint. But the data demodulation of those is even more challenging than GPS because of the additional coding schemes that go on top of it. It’s very difficult for us to demodulate off the SBAS satellite, so we primarily use them for ranging and for autonomous operation when we’re not aided.

In aided operation, we use them less, because we get the data that we need off the Internet, right off a feed, or it comes in off of a satellite, which is much more useful. So SBAS stuff is there because it doesn’t add a lot of cost of difficulty to the receiver design, but it’s not a crucial part of the operation anymore — I’d say with the exception of QZSS, which has a large impact in its regional operating area.

Let’s continue the trend in questions towards multi-constellations. What’s the strategy to switch on another GNSS constellation in case of a GPS problem for a future receiver?

It’s a really good question, but it comes from a premise that we would switch something on that was normally off. The general strategy that’s being followed is to use everything that’s available all the time, so that we can use the methodologies of essentially autonomous RAIM on a receiver where we now have 8, 10, 12 signals coming into the receiver. It’s pretty obvious right away when something has gone wrong, and so it’s not so much time when we can flag when to switch on. It’s more a time when we can switch off. If we see a systemic problem in multiple satellites, then we may use that in our definition, but it’s not from a switching-on standpoint.

So, basically what we’re doing is we’re keeping everything on all the time and relying on autonomous RAIM capabilities from the fact that we’re tracking 14 or 16 satellites at a time with all this extra compute horsepower, because now I’m running embedded CPUs at hundreds of megahertz, where back at SiRF when we could get 15 megahertz, we were happy. So we have a lot more compute horsepower on the mobile side to do autonomous RAIM.

This article is tagged with , and posted in OEM

1 Comment on "Receiver Design for the Future: Webinar Q&A"

Trackback | Comments RSS Feed

Inbound Links

  1. WireTalk.net | Receiver Design for the Future | September 3, 2015