SparkFun Forums 

Where electronics enthusiasts find answers.

Have a good idea for a new product for SFE or Olimex? Let us know!
By Cannibal
#118629
Hi All,

Many of you may not be aware of this, but there is a Soviet/Russian equivalent to the GPS system which works in much the same way. In the financial turmoil of the mid 90's the system suffered and fell as low as 6 satellites in operation, however since 2000 a modernization and re-population effort has been in progress that has seen the constellation restored to 22 operational satellites, all of them modernized dual frequency units.

In the navigation lab I work at, I use GLONASS all the time and have to say that in challenged environments, such as an area with tall buildings, having GLONASS can be a life saver. Even out in the middle of a field there are times of day where you can see as few as 6 GPS satellites in view, so having an extra 4-5 from GLONASS can make the difference between a barely usable position solution, and having an excellent idea of where you are. This week when I was doing field work we had periods of 21 satellites in view, and this is before the EU's GALILEO system is even launched (this chip also supports that system when/if it shows up on the scene).

With the rebuilding of the GLONASS constellation nearly complete, big name manufacturers have started to produce GPS+GLONASS high-sensitivity modules for use in vehicle navigation, or advanced handsets.

One of these is the 'Teseo II' now available from ST Microelectronics.

http://www.st.com/internet/com/TECHNICA ... 295372.pdf

As a GNSS (global navigation satellite systems, not just GPS) enthusiast I'd love to see sparkfun build a module.

I'd also like to offer my assistance in design consultation - I am an Engineering PhD in a satellite navigation research lab who has designed RF systems in the past - just fire me an email if you have any questions.

Thanks in advance.
By Arachnivore
#118693
Wow! I've never heard of GLONASS and I used to work for a firm that builds Geo-location systems. Now it makes sense why you would want a receiver with a few dosen channels. GNSS sounds pretty amazing. Do you know what kind of accuracy you can get by using multiple constellations?

That chip is really neat too, GNSS + an MCU could be really handy.
By Cannibal
#118823
Do you know what kind of accuracy you can get by using multiple constellations?
I wish I could give you a straight forward answer to this one, but as usual the best I can do is 'it depends'.

If you're using a single frequency receiver (L1 GPS 1575.42 MHZ) out in the middle of a field, you will get a slightly better position solution using GPS with WAAS enabled in most situations. This is because with WAAS enabled your User Equivalent Ranging Error (UERE) is usually below a metre.

If you go to glonass-ianc.rsa.ru then click on english, then on the GPS tab, then on the monitoring station link on the left, you can see a real time chart of the signal in space ranging errors to each of the GPS satellites in view. The typical way to determine how accurate your position solution will be given these ranging errors is to multiply the highest value by your Dilution of Precision (DOP). So if you have good satellite geometry and have a DOP of 2.0 and a UERE of 2.5 metres, you would expect a position bound error of 5.0 metres.

If you throw GLONASS into the mix, their ranging signals aren't quite as accurate and while you could bring your DOP down to 1.5, you might end up with a UERE of 5.5, resulting in a worse overall position solution.

Where GLONASS starts to become very handy is if you have a restricted environment, around buildings, where your DOP with GPS alone will be much higher and the trade off of less accurate ranging signals versus better observation geometry makes sense. Our lab made a GPS+GLONASS training system for the Canadian ski team and found that in the mountains the masking was bad enough that with GPS alone there were literally times of day when the system was on a knife edge of 4 visible GPS satellites - GLONASS saved us here. This is incidentally why they have recently re-maneuvered 3 of the GPS satellites into different orbits - to optimize coverage in mountainous regions like Afghanistan.

If you work in a mode called 'differential' mode however, whereby you have a base station close to where you are taking your measurements, both systems quickly become capable of centimetre level accuracy. This is because both base station and rover, if within a few km of each other or even a few 10's of km, will observe almost exactly the same orbital error effects, satellite clock error effects, and atmospheric effects. Suddenly with these errors cancelled out you can determine where you are (relative to the base station) to within cm using either GPS or GLONASS, but ideally both together since then you have more robustness to blockage.


Another area where GLONASS starts to become very useful, but isn't directly related to this conversation is if you are trying to counteract the effects of the ionosphere. The ionosphere is a charged layer of atmosphere that is dispersive at RF frequencies, meaning that different signals will travel through it at different speeds, depending on their carrier frequency. GPS broadcasts a civil signal at L1 (1575.42 MHz) (L1CA) and has 7-8 satellites that now broadcast a civil signal on L2 (1227.6 MHz) (L2CM+L2CL). GLONASS has 22 healthy satellites, 21 of which broadcast L1(1602 MHZ) and L2 (1246 MHZ) civil signals. This means that if you wish to use a strong, openly available GNSS signal to measure and correct for the ionosphere you have 3 times the number of observations available from GLONASS than from GPS. This is a result of the very long lifetimes of the GPS satellites versus the short lifetimes of the GLONASS satellites - in effect the longevity of the GPS satellites has slowed the pace of modernization.

Keep in mind though that there aren't any low cost or high sensitivity modules that use either L2C or the GLONASS L2CA, so that's not something we really have to worry about yet.

To bring my long rant about how accurate you can be with multiple constellations.... why it really depends... ;)