Speedtest (Download Performance) vs NBN (Technology Maps)

This dataset merges two public datasets:

- Speedtest network performance data for Australia (Q3, 2020), as loaded to AURIN geo-analysis platform (88,000 approx locations)

- NBN mapping of Technology data (inc FTTN, FTTP, FTTC, FTTB, HFC, Wireless, Satellite); complete map of Australia, colour-coded by technology, in WMS or KML format. KML format used in this instance.

to produce an intersection datasets (Result: 319337 rows × 26 columns), including:

- data includes LocID, download speed, nbn tech, lat, Lon, SA2, SA3, SA4 (see ABS link below).

But many techs in one Speedtest block (600m^2) so have to untangle.

- a ten line sample (CSV) included.

- see image as provided by AURIN - QGIS.png.

** Versions

v2.Load updated Jupyter Notebook v1.1 and locations csv, which shows location breakdown by NBN Technology (pivot table export).

v1. Initial load, including Jupyter Notebook and human readable, geojson.

** METHOD: Advice from AURIN:

In order to join these two maps, you will need to perform a spatial join based on the two layers. It is possible to do this with geopandas.sjoin(), which by default performs an intersection join - that is, any portion of a matching polygon from the second layer is considered a match to join on. More information about spatial predicates is here in case you're looking for a different spatial relationship.

In this link, I've supplied a notebook (OoklaNBN-AURIN.ipynb) that collects the datasets from the AURIN API and data.gov.au, combines the several KML NBN layers into one, and joins them with the Ookla 2020 Q3 dataset. In order to use it, you will first need to input your AURIN API credentials into the first cell.

The spatial join occurs in the final notebook cell and writes its output to a geopackage (OoklaNBN.gpkg) which I've included also, as the script can take some time to run. You will notice that each Ookla cell now may be represented by many records, this is due to there being more than one overlapping NBN technology polygon. As one Ookla grid can cover many technology zones, aggregating these may be useful depending on how you approach your analysis.

Regards AURIN.

The authors acknowledge the facilities and scientific and technical assistance of the NCRIS-enabled Australian Urban Research Infrastructure Network (AURIN)”

Thanks Evan Thomas, AURIN - ORCiD: https://orcid.org/0000-0001-7564-4116

*** Preliminary Analysis

1. Count of Tech Type

Tech count

Fibre to the Basement (vectored or non-vectored) 14147

Satellite 15946

Fixed Wireless 23663

Fibre to the Curb 36843

Hybrid Fibre Coaxial (HFC) 41175

Fibre to the Premises 93732

Fibre to the Node 93831

2. Mean of Tech Type

NBN Tech Type Aus mean(mbps)

Satellite 48.1674

Fixed Wireless 49.1477

Fibre to the Node 75.1639

Fibre to the Premises 83.71

Fibre to the Curb 86.8433

Hybrid Fibre Coaxial (HFC) 87.1248

Fibre to the Basement (vectored or non-vectored) 116.324

- non-commercial (NC)

- reuse must be share-alike (SA)(add same licence).

This restricts the standard CC-BY Figshare licence.