If you have a list of cell tower ids and their respective start and end time (i.e. the time the phone was registered at that tower), is there an easy to cluster the ids and visualise the cells and their clusters in a diagram?
Thanks
Not without some knowledge of location. Minimally at start time and at end time. If you were travelling in a line for example you could estimate the distance of the tower from the midpoint of those. If you just have time data but no location data there is nothing you can map.
Edit: To simplify, if you have no location data, you can't map much of anything, you could simply index transitions - which towers are certainly adjacent. If you jumped from tower A to B to D to B to C to F to C to D to E to A, then you can map them in graph form:
A -- B -- C -- F
| | //
E -- D
This is only the connections that are definitely there, in the example F could be connected to D, or B could be connected to E, but you can only map the transitions you are sure of.
With location data for the mobile device, you could map the points of influence of the towers. From these points you could derive probable areas of influence (circular).
Conversely with location data for the towers, you could map probable transition spots. These could be connected into polygons of influence.
With no location data you can only make speculative graphs.