Nokia has tacitly admitted the GPS units it builds into a small number of its mobile phones are having problems quickly calculating their locations and has fallen back on data sent over the mobile phone network to mitigate the issue. But wait a minute - it's not a bug, it's a feature. Nokia is spinning the approach as a "new" …
Not just Nokias
I've tried out 2 Orange SPV M700s, the models with a Magellan GPS device. They were ok in getting a 2D lock, but both are crap when calculating elevation, consistently reporting a figure about 60m too high.
Fine for SatNav, but no good with MemoryMap: When mountain climbing, it's worse than useless.
The T-Mobile equivilant, with a Sirf3 GPS, is much better
What this isn't
Please note, this is not DGPS, nor will it ever have that kind of accuracy. In fact, a questionable initial fix will make things worse, not better. I'd almost say that this is "we cocked up GPS. Here's a work around that we'll be calling a feature. Good luck and God speed. And whatever you do, don't make a left turn in the middle of that bridge."
A-GPS has to be seen to be believed - fix in 10 seconds
The A-GPS on the latest N95 firmware is absolutely miraculous, transforming the GPS from mediocre to by far the best I've ever seen.
It usually now only takes 10 or 15 seconds to get a fix, even in difficult circumstances.
Initial GPS lock
The delay in initial GPS lock is because the GPS receiver has to receive an almanac from the satellite that tells the GPS receiver where every satellite should be at any given time. A-GPS simply provides the almanac via an alternative method meaning that the GPS receiver in the phone can get a proper fix much more quickly. That's why GPS receivers re-acquisition time is always much quicker than initial acquisition.
Re: Not DGPS
Correct this is not Differential GPS, it DOES have the potential though.
IF nokia's server was in contact with fixed position reference GPSs and forwarded the relevent corrective data to the phone, you'd have a slightly lagged, but still perfectly valid DGPS system.
Initially locating a GPS receiver through another tracking system is hardly new, a quick google finds a number of references tieing it to INS dating back over the last decade.
This is a case of a company actually using one feature of their product to enhance another, thereby getting the most "bang" from the available components, most people would call that "Good Engineering"
What are you on about?
Anyone that knows anything about mobile phones has known the GPS implementation in the Nokia N95 and E90 is just a bit of an extra in the chipset.
If you can't see Sirf chipset or similar in the specs list you won't get a good GPS experience.
N80 w LD-3W
Will this fix help my N80 with an LD-3W in any way? Is there a plan to implement anything in the nokia maps app itself?
I've been considering upgrading to an E90... does it already have a-gps?
Re: What are you on about.
Anyone who knows anything about GPS knows that at heart it falls into downshift/demodulation,timing and then common or garden number crunching. Once the RF and timing exercises are accomplished it doesn't matter if the number crunching is performed by a dedicated DSP or as a process in the phone's OS.
... how it works.
I just updated my N95 software after reading this and I'm impressed. The built-in GPS now seems as responsive as the bluetooth sirf receiver I have. It managed to pick up a position indoors, which it wouldn't do before, so outside I expect it will be even better.
This, and the release of Mail For Exchange support for N95 is restoring my faith in the device, and Nokia.
All I need now is for tomtom to pull their finger out and issue an update for TT Mobile.
We were achieving sub 15 sec A-GPS fixes on the 3 UK network with the NEC 616 3G phone in 2004 !!!!!