Mapping Toronto’s Mobile Network
I wanted to know whether Ontario's mobile network could draw the province without roads, boundaries or a basemap—and how much that picture could tell me about who actually gets online.
The federal terrestrial spectrum site extract contains 755,444 rows tagged Ontario. At first that looked like three-quarters of a million places. It is not. Antennas, frequencies and technologies repeat the same location across rows. Group those rows by coordinate, then keep only points inside Ontario's official cartographic boundary, and the file resolves to 16,016 reported installation coordinates. The map below shows the 15,404 at or south of 49°N, where most of those records cluster.
at or south of 49°N
Without a municipal outline, the coordinates still roughly sketch the part of Ontario where most people live. The Windsor–Toronto–Ottawa corridor becomes a dense band. Thunder Bay and Sudbury appear as separate constellations, with smaller strings of points elsewhere across the province.
These are records submitted by licensees, not measurements of coverage, speed or use. But as a picture of where infrastructure has been reported, the shape is hard to miss.
Toronto region
Zooming in further to the Greater Toronto Area, the dense band resolves into multiple centres rather than one continuous glow. Toronto is the brightest cluster, with Hamilton and Oshawa anchoring either side and smaller concentrations between them. I thought the clusters looked oddly detached from the population field beneath them. They were not: on a five-kilometre grid, population and reported-coordinate counts were strongly related. The installations do not reproduce population block for block, but the regional network largely follows where people live.
Toronto
Inside Toronto, the points cluster most strongly downtown, with North York and Scarborough appearing as secondary centres. When I added selected road lines beneath them, some strings seemed to follow familiar corridors.
The maps make the infrastructure look almost legible. They also make it easy to assume that a dense-looking network means people have equal access to the internet. The dataset contains no signal strength, speed, price or home-access data. To ask who actually gets online, I needed a different dataset.
Statistics Canada asks people aged 15 and older whether they have internet access at home, excluding people whose only home connection is mobile data. About 95 percent of Ontarians inside a census metropolitan area or census agglomeration reported home internet access in all three survey years. Outside those areas, the estimate was 91.0 percent in 2018 and 88.3 percent in 2022. The point-estimate gap was 4.0 percentage points in 2018 and 7.2 in 2022, but the table alone does not establish that it widened.
Home-internet access inside and outside CMAs and CAs
- 201895.0%91.0%4.0 pp
- 202095.9%90.0%5.9 pp
- 202295.5%88.3%7.2 pp
The chart measures reported home access, not whether a provider offers service at a particular address. It shows the limit of the first map: infrastructure can be dense enough to draw the province while reported access still differs across it.
The first map was convincing enough that I almost mistook legibility for access. The survey is the correction: reported infrastructure can sketch Ontario without telling me who actually gets online at home.
How the maps and chart were madeAdministrative installation coordinates and survey estimates answer different questions and are not joined at the record level.
The July 7, 2026 ISED terrestrial extract contains 755,444 rows carrying the Ontario province code. Rows at identical WGS84 coordinates were grouped, retaining the set of reported radio technologies and summing the number of source rows. This produced 16,799 distinct reported coordinates before geography checks.
Coordinates were retained only when they fell inside Statistics Canada's 2021 Ontario cartographic boundary. That removed 783 out-of-bound coordinates and left 16,016. The broad map includes the 15,404 retained coordinates at or south of 49°N. The Toronto-region map uses a fixed rectangular viewport containing 7,580 coordinates. The Toronto close-up retains 3,785 coordinates inside the union of the City's 158 neighbourhood polygons. A coordinate is highlighted as reporting 5G when any grouped row contains 5GNR or 5GDSS.
The map unit is one distinct reported coordinate, not one physical tower, antenna, licence, carrier, cell, coverage area or measured connection. ISED's physical-cell identifier is not a tower identifier, and several technical records can share one location. The extract is administrative data supplied by licensees; it does not measure performance or coverage.
The Toronto close-up underlays selected roadway features from the City of Toronto's Toronto Centreline dataset. Local, collector, arterial and expressway features are included; laneways, trails, railways and waterways are not. The roads provide visual context only and are not joined to installation coordinates.
The Toronto-region map underlays 2021 Census population density for 10,504 dissemination areas intersecting the fixed viewport. Population density is log-scaled to the 99th percentile for display. For the regional comparison, I assigned each area's population to its area-weighted representative point and counted reported coordinates in fixed local-equirectangular grid cells. Spearman's rho between population and coordinate count was 0.67 at 2.5 kilometres, 0.85 at 5 kilometres and 0.95 at 10 kilometres. At the raw dissemination-area scale, sparse zero-coordinate areas made the result highly sensitive to inclusion, so I did not use that comparison in the prose. This compares 2021 population with July 2026 administrative installation records and is descriptive, not evidence about coverage or demand.
The home-access chart reproduces Ontario estimates from Statistics Canada Table 22-10-0134-01. The population is people aged 15 and older in the ten provinces, excluding full-time institutional residents and attempting to exclude reserve units. Mobile-data-only home access is excluded. The 2022 95% confidence intervals are 94.9–96.1 percent inside CMAs or CAs and 85.2–90.8 percent outside. No significance test for change in the gap is reported.
Contains information licensed under the Open Government Licence – Canada.
Contains information licensed under the Open Government Licence – Toronto.
Statistics Canada, Population and dwelling counts: Canada, provinces and territories, census subdivisions and dissemination areas, 2021 Census of Population; and Access to the Internet at home by geography, 2018, 2020 and 2022. Reproduced and distributed on an “as is” basis with the permission of Statistics Canada. This does not constitute an endorsement by Statistics Canada of this product.
Sources and further reading
- Innovation, Science and Economic Development Canada. Download SMS data
Source for the July 7, 2026 terrestrial spectrum licence site extract.
- Innovation, Science and Economic Development Canada. Terrestrial Spectrum Licence Site Data Extract—Field Descriptions
Definitions for the site, technology, coordinate, structure and identifier fields.
- Innovation, Science and Economic Development Canada. FAQ: Spectrum Licence Site Data Upload
Reporting scope, update cadence and the repeated-row structure of station data.
- City of Toronto. Toronto Centreline
Road geometry underlaid beneath the Toronto close-up for orientation.
- City of Toronto. Neighbourhoods
Official polygons used to limit the close-up to coordinates inside Toronto.
- Statistics Canada. Population and dwelling counts: Canada, provinces and territories, census subdivisions and dissemination areas
2021 population density and land area for the dissemination-area field beneath the Toronto-region map.
- Statistics Canada. Access to the Internet at home by geography
Ontario point estimates and confidence intervals for 2018, 2020 and 2022.
- Statistics Canada. 2021 Cartographic Boundary Files
Ontario boundary used to reject out-of-province coordinates and dissemination-area boundaries used for the population-density underlay.