[FOSS-GPS] High accuracy positioning with low cost GPS devices:a FOSS project

[Abdelrazak Younes]

Hello John,

On 10/10/2008 19:37, John Morris wrote:
> I did a zero baseline test with the LEA-4T, and the carrier phase looked
> pretty good. The LEA-4T needs some interpolation tricks to get things to
> work.

What do you mean by 'interpolation tricks'? Do you make use of future 
measurements in order to help cycle slip detection? I am using a very 
simple cycle slip detector based on range rate step detection (I've put 
the deytection threshold to 6 m/s²). In the future I'd like to switch to 
a sequential algorithm (cusum). I'd like to keep my code working in 
real-time but, considering that I am working with 10Hz raw data, I guess 
I can afford waiting a couple of sample before using the carrier phase...

In my experience, the carrier phase is starting to be usable with a CNO 
of 36dBHz in static positioning mode. By 'usable', I mean no cycle slips 
every three seconds. And I found that the accuracy is not very good, 
about 10cm/s (one sigma) at 40dBHz. With dynamic positionning in a urban 
environment, things are much, much worse...

I'd like to hear about your experience. My goal is to have the best 
possible standalone navigation solution (including SBAS). Right now I 
manage to get a position that is better than the LEA-4T but not by a big 
margin.

>   The other inexpensive receiver is Thales AC-12. This one provides
> carrier phase within a couple mm. My understanding it is the same GPS engine
> which is used in the Promark II for doing single frequency RTK. (BTW, is
> there a cheaper LEA-4T than the u-blox demo unit?)
>    
Don't know, maybe in the Chinese or Taiwanese market.

Abdel.