At which addresses can we trigger Iran's censorship middleboxes from outside?
Much of the GFI's filtering acts on traffic in both directions, so a probe sent from outside towards an address in Iran can provoke filtering on the path to it, and that response is what we record. A result below means bidirectionally-triggerable censorship infrastructure sits on that path: not that the address is itself actively doing censorship, nor that its operator did anything, but that someone using it would be censored. Read it as a floor rather than a portrait: a user there would meet at least this, and possibly filtering that only responds to traffic originating inside. An address with no result may still sit behind such filtering.
Censorship coverage by network
Censored-address counts over time inside the largest autonomous systems in Iran. Blocking is not applied uniformly across the country, so an operator can differ from the national picture. Mechanism:
Top 10 networks by censored IP address count
Each figure is the share of that network's address space in Iran in which we observed the mechanism; follow the network number to RIPEstat for its registered allocation. Each mechanism is counted on the most recent scanning batch it produced data. A 0 means we scanned that network for that mechanism and triggered nothing: either it is not filtered, or the filtering there cannot be triggered from outside.
These figures are a lower bound on censorship: we measure from outside Iran, so anything our probes cannot reach is missing rather than absent. See Q&A for what these measurements do and do not show.