The Google Maps Geolocation API actually lets you define multiple gsm cell towers into its request. But it will just return the location and accuracy based on the first cell(serving cell), doing nothing meaningful about the neighbour cells, rssi and timing advance values.

So this is an attempt to do something a bit more meaningful...

Use pip:

pip install geocell

Or clone the repo and:

pip setup.py install

The key will be used in:

• Google Maps Geolocation API for requesting cell locations
• Google Maps Javascript API for drawing maps

If you don't have a google maps api key. Go here and follow the instructions to get your key.

Set your api key using either of the following methods:

• Set your key as the GOOGLE_MAPS_API_KEY environment variable
• import geocell then geocell.api_key="<your api key>"

>>> import geocell
>>> sample_cell = {"mnc":2,"mcc":286,"cid":51861,"lac":54110}
>>> geocell.locate(sample_cell)
{'location': {'lat': 40.7018889, 'lng': 29.8912834}, 'accuracy': 3250.0}

We need the rssi values and multiple cell information to make an estimation

>>> sample_cells = [
...     {"rssi":-82,"mnc":2,"mcc":286,"cid":51861,"lac":54110},
...     {"rssi":-85,"mnc":2,"mcc":286,"cid":16116,"lac":54110},
...     {"rssi":-93,"mnc":2,"mcc":286,"cid":0,"lac":54108},
...     {"rssi":-94,"mnc":2,"mcc":286,"cid":38344,"lac":54110},
...     {"rssi":-97,"mnc":2,"mcc":286,"cid":52555,"lac":54110},
...     {"rssi":-98,"mnc":2,"mcc":286,"cid":51857,"lac":54108},
...     {"rssi":-99,"mnc":2,"mcc":286,"cid":39684,"lac":54110}
... ]
>>> geocell.estimate(sample_cells)
{'location': {'lat': 40.70356939393244, 'lng': 29.88564243119295}, 'accuracy': 251.3839360809747}

You will notice that this function takes couple of seconds to return, the reason is that the Google Maps Geolocation API requests take some time. But the cell requests are cached internally in the module so the next time you request the same cell, the value will be returned instantly. If you like, you can also set geocell.is_multiprocess = True to do multiple requests, just don't use multiprocessing inside a interpreter, it doesn't like it.

Let's see it in action now:

>>> geocell.estimate(sample_cells, "map.html")
{'location': {'lat': 40.70356939393244, 'lng': 29.88564243119295}, 'accuracy': 251.38393
60809747}

Looking at the generated map:

• Light yellow circle is the serving cell
• Red colored are the neighbour cells
• The blue lines are the flight path showing estimation done at each step. Starting point of the flight path is the center of the serving cell
• The marker shows the end of the flight path and the last estimated location

Let's take a closer look:

>>> geocell.estimate(sample_cells, "map_estimate.html", cell_display="estimate")
{'location': {'lat': 40.70356939393244, 'lng': 29.88564243119295}, 'accuracy': 251.38393
60809747}

Now the map shows only the estimated location(marker) and estimated accuracy(radius). The arrow points to the actual location.