README | Introduction | Datasets ⮕ Tasks | Notebook
The ASN customer cone is a metric used to gauge the size and influence of an Autonomous System (AS) within the global routing system. It represents the complete set of ASes that traffic can reach from a given AS by following only provider-to-customer (p2c) links.
We define an AS A's customer cone as:
- AS A itself
- Plus all ASes reachable from A by following only p2c links in observed BGP paths.
In other words, an AS's customer cone contains itself, its customers, its customers' customers, and so on.
This construct embeds an assumption that ASes in the customer cone for AS A pay AS A—either directly or indirectly—for transit. We denote the size of an AS's customer cone as the total number of ASNs in its customer cone set, providing a coarse metric of that AS's footprint in the routing system.
You will download CAIDA's May 2026 ASN Customer Cone (20260501.ppdc-ases.txt.bz2) file to the data directory. You will then complete tasks in the notebook that divide the ASNs into tiers based on their customer cone size.
CAIDA provides its ASN customer cone as part of the CAIDA AS Customer Cone (Serial-1) dataset. Follow the link below and download the file labeled 20260501.ppdc-ases.txt.bz2 and copy it to your data directory.
In this module, the size of an AS's customer cone is measured as the number of ASNs in that set.
In the ppdc-ases.txt.bz2 file, lines that start with a '#' are a comment. All other lines start with a single ASN followed by a list of ASNs in its customer cone. The cone size for a given AS is the number of space-separated tokens on its line minus 1 (the first token is the AS itself).
# This is a comment
# 23's customer cone size is 3 and includes (23,4,1)
23 23 4 1
# 1's customer cone size is 1 and includes only itself
1 1
CAIDA's AS-to-Organization-Mapping Dataset provides a list of organizations with ASNs and the set of ASNs assigned to those organizations. CAIDA infers these mappings based on Regional Internet Registry WHOIS records. The notebook downloads this file automatically and stores the results in data/as2org.jsonl.
| name | type | values | description |
|---|---|---|---|
| score | int | 1..120793 | organization sorting score (min..max) |
| orgId | string | 96894 | organization ID (number of unique IDs) |
| orgName | string | 91067 | organization name (number of unique names) |
| country | string | 239 | organization's headquarters country (number of unique codes) |
| source | string | 7 | Internet Registry source (number of registries) AFRINIC, APNIC, APNIC,JPNIC, ARIN, JPNIC, LACNIC, RIPE |
| members | [string] | 1..1008 | member ASNs (min..max list size) |
| changed | date | 1991/06/12 2026/04/01 |
last time the information was changed in WHOIS |
| date | date | 2026/04/01 2026/04/01 |
current record date |
| ts | date | 2026/04/17 2026/04/17 |
database record timestamp |
Files that end with the extention JSONL are JSON Line Files. After the notebook has downloaded the data/as2org.jsonl file, copy a single of the uncommented lines into a JSON Formatter. You will be able to see an example row from the dataset.
README | Introduction | Datasets ⮕ Tasks | Notebook