CellHunter nearness questions

EdorFaus edorfaus at xepher.net
Wed Sep 16 19:20:50 CEST 2009


Hi,

I've been doing some cellhunting, and have some questions about the 
"near" (OCoGn) locations that I haven't been able to find answers to.

I'm asking here instead of in private because I figure others might be 
interested as well.


First, how near is near?
Or stated differently, how far apart do locations have to be, to not be 
so near each other that they don't give my group any points?

(Giving this in relative GPS coords would be fine (I'm assuming that's 
what the check is based on) - as that's what I'm looking at (CH logs).)


Second, are previous "near" locations counted when determining if a 
given location is too close to existing ones?

As an example of what I mean, say I'm in an area with no preexisting 
data, and only one GSM cell. If (for argument's sake) locations have to 
be 15 meters apart to not be too near, and I move in a straight line, 
registering a location every 10 meters, will I score any points at all 
(other than for the first location, which would be NCG)?


Third, is the OCoGn data valuable at all? Is it useful for anything?

If not, it might as well be pruned before it is submitted...
(Unless we want the OCoGn count to go up I guess.)


Especially the second question is interesting, as it has an impact on 
what data should be submitted to maximize the score (if the example 
would only give me points for the NCG, I could have gotten more points 
by not submitting all the data gathered, or with a different order).


--
Frode



More information about the community mailing list