Process logic for aggregating cells into towers they share


#1

Hello! I am interested in creating an algorithm to use this data to create a fairly reliable database of actual tower/antenna geo coordinates. In looking through the data and the map visualizations, it is clear that the triangulation method used introduces some inherent position error - I believe I read that in some rare cases it can be as much as 100-200m.

Understanding that many of the cells in this database which appear within 100m or so of each other probably share the same tower, would it be reasonable to assume that an algorithm which aggregated all cells within 100m of each other, and perhaps choosing the centroid of their coordinates as the approximate location of the tower, would be a reasonable approach for rural and suburban areas in the US? I realize this would not likely work in dense urban environments.

Is it also safe to assume that if a cell appears in the data and is widely separated from other cells in the database (let’s say more than 200m) that the cell is really present and is not a false positive? I ask because I see cells on the map in our area that do not appear to be present when I visit the location. So I assume either 1) a false positive based on bad readings or triangulation errors, or 2) an antenna/cell which is actually present in that approximate location, but is hidden (building, etc.).

Thanks!


#2

Hello William

If you are successful in creating an algorithm that works to map cell towers globally could you please contact me. I work for an international cell tower company this data would be instrumental for us to run competitive market analysis. So far we haven’t been able to find a good tool for this.

Thank you!


#3

CellMapper will help you.


#4

Hello-

I have come up with an algorithm that provides clear, reasonably accurate cell tower locations based on processing Opencellid data. I’ve processed cell tower locations for the U.S. only at this point, but I see no reason why it wouldn’t work in other parts of the world as well since I believe the Opencellid data is captured and process in the same fashion globally. Let me know if you’re interested and I’d be more than happy to discuss how I could be of assistance.

Best Regards,

Nick Damenti

m: +1-516-695-3003


#5

Hello-

As I mentioned to William on this thread, I’ve come up with a method to provide accurate cell tower locations based on OpencellID data. Let me know if you’d like to discuss.

Best Regards,

Nick Damenti

m: +1-516-695-3003