Thursday, May 3, 2012

Improving Galileo's Accuracy with only 18 Satellites

I'm just back from the European Navigation Conference in Gdansk Poland, where I presented a paper on improving the accuracy of the Initial Galileo navigation constellation.  Europe now has all 18 satellites funded to reach their Initial Operating Capability (IOC), currently slated for 2015 sometime.  With only 18 satellites, there is 100% coverage (at least 4 satellites in view at any time, globally), but the orientation of those satellites will not always be ideal.  In fact, there will be times during the day when Galileo will suffer from Dilution of Precision (DOP) spikes for times longer than 30 minutes.  This will be seen as Galileo outages by receivers.  These times are dispersed throughout the day; other times during the day navigation accuracy is not bad.
In my paper, I outline how the Galileo Control Center (GCC) can improve average user accuracy by improving the upload times for each satellite.  Essentially, the more frequent the uploads of predicted satellite orbital and clock states, the more accurate the Galileo navigation satellites will be.
I use the notion of a Signal-In-Space User Range Error (SISURE) that defines how accurate each Galileo satellite is.  The lower the SISURE, the more accurate a user's position.  When the GCC uploads its predictions more frequently, the SISURE will be smaller.  This is because the SISURE is a difference between the actual satellite position and time and the predicted satellite position and time, uploaded to the satellite and broadcast to the users for use in calculating their positions.  The older the predictions are, the higher the SISURE is, the larger the user's error is.

Below is a picture of 24 hour average navigation accuracy with each Galileo satellite having a SISURE of 2 meters.  Note that because of the geometry of the 18 satellites, there are large mid-latitude bands that suffer more from DOP spikes.  These will go away when more satellites are launched.  The average accuracy across the globe for this configuration is 5.7 meters.

Now, if the GCC can maintain an upload schedule that limits the SISURE to 0.5 meters, a more accurate result can be seen:

In this scenario the average global accuracy drops to 1.9 meters - a big improvement.  Notice also how the DOP spikes are minimized in the mid-latitude bands.

This is only one of many other methods that can be done to improve accuracy with an initial Galileo constellation.  Working to keep the user's navigation error small will gain a lot of trust for Galileo in the user base, making it a better system from the start.

Thursday, February 16, 2012

LightSquared's Approval Revoked!

LightSquared's conditional approval to operate ground-based 4G LTE transmitters (cell towers) was revoked by the FCC this week.  This is great news for GPS users.  GPS has an incredibly weak signal and can be easily interfered with, or jammed.  4G services by their nature are high powered relative to GPS signals, and even though the frequency bands used by both were not the same, they were next to each other - causing unacceptable interference to GPS receivers.
LightSquared secured their spectrum quickly through the FCC, and interference issues were not taken into account by either party.  Because GPS is the equivalent of a national utility used by both civilians and the military, the FCC and LightSquared both screwed up by ignoring the interference possibilities. Additionally, GPS receiver manufacturers could have put filters in place (though not completely effective), to alleviate some of the interference from any outside transmissions.  However, there have been no realized threats to the spectrum - why would the receiver manufacturers build something into a receiver, at their expense for a problem that didn't exist?
The good news is, your GPS receiver, my GPS receiver, the FAA's NextGen air traffic control system, banks, power companies, farmers, cellular service providers and all other users of GPS can now stop worrying about whether their GPS receivers will work when LightSquared turned on their system.
GPS receiver manufacturers should take note - you got lucky this time, but now IS the time to start building in safeguards and fortifying your defenses against jamming, unintentional, or intentional.  There will be other spectrum contenders to come.

Sunday, February 12, 2012

PRN 26 off the air - what happened?

On February 9, GPS PRN 26 was taken out of service.  Notifications were sent to GPS users via UNUSUFN NANU 2012006 (Unusable until Further Notice notifications).  When these types of notifications are sent out, it means the Air Force wants this satellite of the air immediately.
Here's a look at PRN26's atomic clock behavior for three days prior to it going out.  Erratic clock behavior is one of the likely causes of an outage and is typical of an aging clock.  According to Richard Langley's GPS status page, this is a Rubidium atomic clock and was turned on in July of 1992 - which makes it almost 20 years old - 10 years beyond its design life!  This is the second oldest operating clock in the entire GPS constellation.  (The oldest clock belongs to PRN 32, turned on in December of 1990)
Ideally, this pink line will be kept as close to zero as possible.  The amount of error in shown in this plot is incorporated into your GPS receiver and can be seen as errors in your position determination.  The US Air Force adjusts the satellite's clock values to keep this line as close to zero as possible.  A typical adjustment can be seen about 12 hours into Feb 07.  Normally the clock values move around a bit, but about 6 hours into Feb 08 the clock jumped dramatically.  Each dot on the line represents 15 minutes, and clock errors of several meters (1 meter = 3.34 nanoseconds) in that short of a time are unusual.  You can see that the clock is settling down between jumps but keeps having erratic episodes until, around 1800 hours on Feb 09, it starts changing by 1 meter every 15 minutes.  Finally, the Air Force took it off the air at 1917 UTC.  
There's always a balance between keeping a satellite on the air (more satellites in view means a better position determination for you) and turning off a satellite that is harming your position.  So, as the Air Force watches each 15 minute update, they have to decide to keep it on, or turn it off.  Good call on this satellite and let's hope PRN 26 is healthy again soon.

Friday, January 6, 2012

LightSquared facing the death knell?

Has the death knell for LightSquared arrived? The latest legislation signed by President Obama has a portion dedicated to GPS interference by ground-based data services. For ground-based data systems to exist, they must prove that they do not interfere with GPS. With Sprint putting its investment in LightSquared on hold, LightSquared is facing a major challenge. Here's the link to the eWeek article on the new legislation:
LightSquared GPS Interference Issue Faces Congressional Oversight - Enterprise Networking - News & Reviews -, and here's a link to the Sprint status with LightSquared: Sprint waits on LightSquared.

'via Blog this'

Tuesday, November 8, 2011

How long can you use an almanac?

Note: this is a reissue of a Nog originally published in 2008.
When you turn on your GPS receiver - what happens? Well, lots of stuff.  But primarily, GPS receivers work in a two stage process:
  1. Look for available satellites to track
  2. Do everything else
In this article, I want to focus on step 1, we'll get to step 2 later.  This is probably common knowledge, but for the record I want to state it.  The GPS receiver uses an almanac downloaded from a single GPS satellite to help it determine what satellites are above the horizon as it searches for signals.  Makes sense to not look for satellite signals that aren't even visible - it decreases your time to first fix (TTFF).  Since the receiver is simply determining whether a satellite is above the horizon, the almanacs don't have to be very accurate.  They are typically a coarser version of the precise ephemeris broadcast by each individual satellite. 
What I do want to discuss is how long you can use an almanac for analysis.  Your receiver will usually download a new almanac when it sees that a new one is available, so it always has the freshest data.  When you do analysis though, sometimes you may not have the latest almanac (or the one correct for the time of analysis - a related problem).  So, is it ok to use any old almanac for analysis?  Since they are not the most accurate ephemeris representations, should I be using them for analysis anyway?  I'll answer these questions and show some interesting graphics that bring the point home.

Saturday, August 27, 2011

Recalled SVN-35 Not Doing Well

On August 16th, 2011, according to the 2SOPs Satellite Outage File (Outage Calendar link) and NANU 2011-062, PRN-30 (SVN-35) was set healthy, bringing back a satellite previous removed from the constellation.  SVN-35 had been labeled a spare, making it reusable in case it was needed.  That’s just what happened  over the last few months.  According to InsideGNSS:
SVN-35, also a Block IIA satellite, had been decommissioned from active service back in 2009 to make room in the constellation for the launch and eventual deployment of the latest new GPS Block IIR vehicle.
Even while it was removed from the almanac of the active constellation, SVN-35 maintained accurate timing and navigation signals; so, when the need arose for a spare, 2 SOPS analysts knew just where to go.
However, over the last few days, SVN-35 is showing it’s age.

Friday, August 12, 2011

GPS OCX PDR Still Pending - 106 issues noted

Raytheon has 106 Preliminary Design Review (PDR) Issue Notices or PINs to work through - issues that were raised at the June OCX PDR. A team within Raytheon has been working them, to a completion date at the end of August.
An independent OCX PDR review team from the SMC, chartered by Lt. Gen. Pawlikowski, examined plans for completion of the PINs and stated that Raytheon is on track to complete them.
From and article in Inside GNSS: GPS OCX PDR Still Pending against Backdrop of Defense Cuts | Inside GNSS