Glossary
ASN-level aggregation
ASN-level aggregation rolls network measurements up to the autonomous system number, the routing unit used alongside country rollups. APNIC Labs aggregates its dashboards by country, UN region, organisation group and individual ASN across roughly 75,000 ASNs; Hurricane Electric publishes 147,000+ ASN pages.
What is ASN-level aggregation?
An autonomous system number identifies a network under one routing policy - an ISP, a cloud provider, a university. Aggregating to ASN therefore groups users and infrastructure by operator rather than by geography, which is why measurement teams publish both cuts.
APNIC Labs states its dimensions explicitly: measurements are 'aggregated by country, UN region, organisation group (RIRs, G8, G20) and individual ASN' across '~75,000 ASNs'. OONI Explorer makes its censorship measurements 'filterable by country, ASN, test and date'. At the registry end, Hurricane Electric's BGP Toolkit carries roughly 147,000+ ASN pages with per-ASN, per-prefix, per-IP, per-IXP and per-country views, and PeeringDB tags its interconnection records asn, peering and interconnection.
Why does ASN-level aggregation matter when choosing a dataset?
Because country averages hide operator reality. A national IPv6 adoption or censorship figure can be dominated by one or two large networks; the ASN cut shows whether a behaviour is market-wide or specific to a handful of providers you may or may not care about.
It is also the join key between datasets. Interconnection records in PeeringDB, BGP objects in Hurricane Electric's toolkit and measurement series from APNIC or OONI all resolve to the same ASN identifier, so a pipeline built on ASN-level aggregation can merge them without fuzzy geo-matching - provided each source's ASN coverage matches your target market.
How do you evaluate ASN-level aggregation in a source?
- Check declared ASN universe size. APNIC covers about 75,000 ASNs in its aggregation; Hurricane Electric lists roughly 147,000+ ASN pages - know which universe a claim assumes.
- Confirm the grouping dimensions. Country, UN region and organisation-group rollups (RIRs, G8, G20) alongside raw ASN let you move between views without reprocessing.
- Test filterability. OONI-style filtering by country, ASN, test and date is the minimum for time-series work.
- Verify registry linkage. Per-ASN pages should connect to per-prefix, per-IP, per-IXP and per-country views so routing context travels with the number.
Related terms
- internet-censorship-measurement - OONI's measurements, filterable by ASN.
- protocol-adoption-measurement - APNIC's adoption dashboards aggregated by ASN.
Sector context: interactive-media-services data.
Frequently asked questions
How many ASNs do measurement datasets cover?
APNIC Labs aggregates across roughly 75,000 ASNs by country, UN region, organisation group and individual ASN; Hurricane Electric publishes around 147,000+ ASN pages in its BGP Toolkit.
Why aggregate internet measurements by ASN instead of country?
Because operators, not countries, control routing behaviour. ASN-level cuts separate one large network's effect from a nationwide trend, and give a stable join key across PeeringDB, BGP and measurement datasets.
Datasets containing this field
Datasets containing ASN-level aggregation
6 datasets carry asn-level aggregation in the catalog. Open one, count the fields, judge for yourself.
APNIC Labs Measurements & Dashboards
Domain Name Industry Brief & DNIB Data
quarter / reporting period · total_domain_name_registrations · com_net_domain_name_base …+3 more
Hurricane Electric BGP Toolkit
IANA Root Zone Database
OONI Explorer – Global Internet Censorship Measurements
Every listing shows the field dictionary, sample rows, and coverage before you commit. API, files, or your warehouse. Daily, weekly, or hourly.
Get sample rows