Open Source GPS ION GNSS 2006 Conference review, to do list, some hardware/software ramplings

Guys,

While I reviewed the papers at the session we had at ION GNSS 2006 I didn't tell you how the rest of the week

went for the project. Rick Niles and I taught a seminar for Navtech GPS on OpenSource GPS. We were very pleased

with the attendance and even had some people well known in the GPS community stop by.

Doug Baker's company GPS Creations along with OpenSource GPS had a small booth at the conference. Doug had

many of his boards available and using two computers (one running DOS on a PC104 and the other running Linux) we

were able to connect up antennas to see if we could track any satellites. We were a bit skeptical since we were told

that re-radiation of the GPS signals was not going to be allowed and it looked like the signals would have to travel throuh

a good amount of building material. But, we were occasionally able to track satellites above 50 degrees elevation

and come up with a navigation solution that appeared to make sense.

The traffic at the booth was great and we had a chance to meet up again with old friends and hopefully made a few new

ones.

This is a to do list that was generated for a handout we gave out at ION GNSS 2006. Unfortunately we ran out of them

on the second day of the conference.

* General improvements

** Get it to work using interrupts from the card (must have newer version of the card with these interrupts available)

** Allow data message to output every six seconds with full frame information and keep user side working in sane manner.

** WAAS support

- Need previous problem solved at a minimum

- Step 1: Add WAAS PRN codes and search for them

- Step 2: Verify you can track a WAAS satellite

- Step 3: Data pull off changes since it's 5x data rate.

- Step 4: Fix user side for this data change

- Step 5: Implement DO-229C on user side.

** Add a matrix math library (file) so we can do Kalman filters and other higher level math stuff.

** Add weighting to the least squares solution routine

** Add capability of using or generating differential corrections

** Provide option to use 3rd order tracking loops (currently only 2nd)

** Modify acquisition method to allow better low signal performance (longer integration time, Tong detector?)

* Linux Specific

** Figure out an elegant solution to number of doppler search bins bug when you don't have an almanac. Dilemma: The number of search

bins is large if doing cold search in device driver and automatically reduces when channel is assigned. However, userland cold search

assigns channels and messes up the whole idea. Obvious solutions are to either not do cold search on user side at all (this might break

the DOS version) or allow user side to control number of search bins. Don't want to modify the interface.

** Write another user side program that uses the device driver to allow for more testing of interface and show flexibility. Perhaps, re-use

some other open source GPS libraries from other groups to make these easier.

** Add support for gpsd so we have a moving map via something like gps drive.

** Allow the user program to set the system time if it has authority to do so. (root permission?)

** Integrate with other kernel time setting code to allow for extremely accurate time control. (this could be really useful for certain

applications)

* DOS Specific

** Get PCI bus probe in assembly working for DOS and Borland C 2.01

** Get the DOS version to use the RTC instead of timer interrupts (mostly just for consistantly with Linux version.)

** Get program to compile using DJGPP compiler

Some hardware we are looking forward to using in the future:

Talking to the GP2021 over a USB connection.

An FPGA version of the GP2021 and then hopefully more advanced FPGA versions to track L5 and the Galileo signals.

An IF A/D data dump to allow a project of Open Source GPS to work as a software receiver.

Additional software projects:

All software receiver

Software linkage between OpenSource GPS and GPStk

Open internet data source for broadcast ephemeris data

Open internet wide area differential corrections.

If you have any additional ideas or comments please post them for discussion.

Best regards,

Cliff

_______________________________________________
opensource_gps mailing list
[email protected]
http://lists.psas.pdx.edu/cgi-bin/mailman/listinfo/opensource_gps