Cell tower positions in map seem to represent individual measurements

[mvglasow]

What steps will reproduce the problem? Open the map, display cell positions and zoom in on the Baltic Sea between Helsinki and Tallinn. (Currently visible only when you zoom in, as the data is quite fresh.)

What is the expected output? What do you see instead? I would expect some averaged cell tower positions from measurements taken on the ferry from Tallinn to Helsinki and back. Instead I see two clearly distinguishable “breadcrumb trails” with dots spaced evenly apart (some brighter than others, as my device occasionally switched to 2G and picked up a bunch of cells). This looks to me as if there’s one dot per measurement, not one dot per cell.

What version are you using? On what operating system?

Please provide any additional information below. Measurements were made with a single device on the ferry from Tallinn to Helsinki and back. I’d expect to have picked up at least some cells on both legs of the journey, which should then show up as averaged positions between the two trails—instead I am seeing two clearly separated trails.

[wish7code]

Hey Michael,

the values in our database look valid (i.e. distinct cell ids). It's 90km distance, so at first I assumed that part of the readings might come from maritime base stations installed on the ferries itself. This could potentially explain the 'both legs of the journey', as in that case the cell station would travel across the sea too.

But even in the middle of the sea we see different cell ids. So I could imagine we've got a measurement bias here: That is, we've lot of different measurements, but our ferry captain is on GPS auto-pilot, so all measurements collapse on the two legs along the (auto-piloted) ferry passage only.

Just an assumption: Other ideas?

Cheers Toby

[mvglasow]

That looks a bit different from what the renderer is producing, see screenshot:

A mobile base station aboard the ship is unlikely, as I collected somewhere around 100 cell IDs along the whole leg of each trip, and remember seeing both Estonian and Finnish cells, with the cut-over occurring somewhere in the middle—thus they seem to be cells on land. (For a mobile base station I would expect to see the same cell ID all along the trip.) I also don’t remember ever picking up LTE cells (the card I had in my phone, or its contract, do not seem to support LTE). For the records: as far as I can remember, I was roaming on Elisa EE and Elisa FI, didn’t ever see any other carriers.

Just a guess: could it be that the image you posted includes OpenCellID data, while the map on the web site doesn’t?

What puzzles me is that both legs are so clearly visible on the rendered map, with not a single cell in between—cells I’d picked up twice, once on each leg, I would expect to show somewhere between. So the renderer image looks like either there was not a single cell that I’d picked up twice, or no averaging of cell positions took place.