How exactly do we define a “cell” in the OpenCelliD data? I find there are many more cells than can be feasibly possible when I look at OpenCelliD data, therefore I’m wondering if my interpretation/definition is incorrect.
Background: I recently downloaded data from December 2022 and I’ve been having problems validating against truth datasets, so I was hoping for some guidance from the OpenCelliD community please.
Existing knowledge: As just one example, I know there are ~40,000 sites in the UK spread across four operators, and essentially two main networks (given site sharing).
OpenCelliD: With >1.7 million cells estimated for the UK, this would suggest more than 40 cells per site, which is clearly not correct. If I had a three-sector site I would class this as three cells, but this assumption can’t be true.
Possible interpretations:
- Is a ‘cell’ technically classed as something else? For example, would a multi-band antenna be counted as multiple cells?
- Does a MIMO-enabled asset (e.g. 4G, 2x2, 4x4, or in 5G, 64x64 or 128x128) begin to count as multiple cells?
- Do we have double counting in the data?
I would very much appreciate some advice and guidance please from people who have a better understanding of this data than I do.