How about this:
If we have all day to track 100 positions and looking for the cheapest solution.
We install 100 remote receivers (micro + RF) and sample the IF (20 to 30 MHz) for 10 to 20 seconds (100 MBytes)? We can transmit this with a slow serial link to a central station.
We can then process this 10G data in DGPS mode, using the Philips's software solution.
Will this work?
OK, I guess we need to use a mobile rover, using the new L2C signals. Someone has done this without the L2C signal, they claim to have cm accuracy. There are currently 2 or 3 birds with L2C, so we need to time the daily sampling windows.
The plan is to have two sets of dual frequencies RF front ends (L1=1.5GHz L2=1.2GHz). Using a micro, we can sample and store a couple of seconds of the 1Mhz sub-carrier signals. Differential measurements can be taken with both mobile and stationary (fixed reference position) samples together.
The samples would be WiFi'ed to a central computer for post processings. We can even archive the 1G bytes samples on DVD (3 to 4 days of samples), in case we need to go back and recalibrate the data.
Any suggestions or comments?
Any WAAS enabled GPSR would give you 2 meter accuracy on a good day if you averaged for a couple of minutes. This assumes that that is good enough (you don't sat what accuracy you need) and that you are in an area with a decent sky view and can receive WAAS (you give no indication where you are). Waypoints will store the location(s).
Yes, we have WAAS enabled. The sky is clear. We want 0.1 meter accuracy. Someone did a terrain mapping within a couple of cm with DGPS, using L2 (military band) sub-carrier corrections. I think that's how they make survey grade GNSS.
I believe we can make use of the new L2C (civilian band) codes for similar result. We don't need to process the data immediately, just to log enough information for later processing. We can tap into the L1 signal from a standard GPS. My question is whether we can use a different crystal and get the L2 signal?
Yes. In fact, if you chose your frontend wisely, you only need to sample about 10MBits/sec - 2 bits @ 5MHz. It's not an easy job though. You may want to talk to NXP (Ex-Philips Semiconductors) first, before investing a lot of work in this.
However, unless you are planning to make thousands of mobile devices, it's probably cheaper (and lower power) to find a complete GPS chipset which outputs pseudo-ranges.
Kind regards,
Iwo
The problem is finding a chipset for L2 (1.2G). Some internal filters tuned to L1 (1.5G) would not work.
In additional, we want to use the NIST chip atomic clock (E-15 precision), instead of the TCXO crystal (E-10 precision). This would elimate the carrier phase unknown and simplify the PLL.
The atomic clock chip contains laser tuned cesium particles in silicon wafer. It does not seem too difficult to make. A Differential mode Atomic Global Positioning System (DA-GPS) would be perfect for us. Is anyone working on the atomic clock, other than the NIST reseach team? Thanks.
Have something to add? Share your thoughts — no account required.
Ask the community — no account required