Deep-Diving Through Iran’s Internet Blackout
Over 7.4 million remote observations of Iran’s internet don’t lie. Here we deep-dive through the moment the lights went out.
Last Friday, to the surprise of everyone, the Iranian government announced a 50% increase in gasoline prices and swiftly imposed a strict rationing system.
Amidst an already dire economic situation in Iran with an unemployment rate of 14% and an eye-watering inflation rate rate of around 40% ,
the regime’s announcement almost immediately sparked widespread civil unrest.
Information about the ongoing unrest quickly spread through social media and Iranians used navigation apps to coordinate road blocks with their cars (“car protests”) in Tehran.
To prevent further mobilisation of demonstrators, the government completely shut down the country’s internet on November 16.
Unfortunately, the Iranian regime is no rookie when it comes to disrupting the country’s internet during politically sensitive times.
There are many well documented cases of government interference with citizens’ ability to access the internet as a means to deter political dissent, such as during the contested 2009 presidential elections or the 2012 one year anniversary of the green movement protests.
As in the past, and in contrast to a number of autocratic regimes, the Iranian government openly admits to its citizens that the current outage is a government ordered shut-down, promising that it will turn the internet back on “soon”.
However, in line with the findings by Oracle’s Internet Intelligence, the Monash University IP-Observatory’s analysis confirms that the scale and intensity of this shut-down is unprecedented in Iran’s history.
While during previous events, the Iranian regime mainly resorted to tampering with internet speed or the blocking of social media sites and news blogs, this time, it has implemented a full-scale blanket shut-down.
9pm, 16 Nov 2019: The hour Iran’s Leaders flicked the switch
From our remote scanning infrastructure, the sudden, wide-spread, and near total shut down of access to the internet at around 9pm, 16 November 2019 IRST (Iran Standard Time), was all too clear to see.
Below, we provide a dynamic visualisation, at hourly frequency, of Internet Connectivity at 73 counties across the country that we have statistically reliable measurements for.
Internet Connectivity is an index from 0 to 100 that we calculate from measurements of online internet addresses known to be located within a given county (see details below).
A score of around 70 to 100 indicates normal internet, below 40 indicates severely degraded internet access, whilst scores of 5 or less imply that the region is effectively offline.
At 12pm midday on the 16th November 2019, all 73 counties had connectivity scores of 70 or more.
By 11pm, we counted zero.
However, at the end of the visualisation, one county — Bam, Kerman — stubbornly appears to buck the trend. We return to this point below.
7pm, 17 Nov 2019: An Information Blip?
Next, we look at the time-series of the event across 30 provinces of Iran.
The initial period of fluctuations between connectivity scores of 70 to 100 are normal in these kind of data: internet connected devices and machines tend to go off-line during the early hours of the morning.
However, the outage event of the 16 Nov is unmistakable.
But what’s also interesting is the partial return of the Internet in eight Provinces at around 7pm on the 17th Nov 2019.
Provinces included in this connectivity spike included: Ardebil, Golestan, Markazi, Mazandaran, Qazvin, Qom, Semnan and West Azarbaijan.
However, we note that none of these Provinces rose to a connectivity score of greater than 60 during this blip, and all returned to a severely degraded state soon after.
The Curious Case of Bam, Kerman Province
The IP-Observatory conducts high-frequency scans at very fine spatial resolution, which enables us to get a highly disaggregated picture of any shut-down.
As we noted in the first, dynamic visualisation above, despite the ongoing nature of the nation-wide shut-down which started on November 16th, a small pocket in the country’s south-east is still online.
This is the area around the historical city of Bam, Kerman Province.
Since the outage event, Bam’s connectivity score has rarely been below 70.
Bam was hit by a devastating Earthquake in 2003 that killed 26,271 people and injured up to 30,000.
As part of the international relief effort coordinated by UNDAC, Télécoms Sans Frontières (TSF)deployed an emergency telecom centre that was funded by the European Commission.
While the main goal of TSF’s mission was to install calling operations, it also established internet connection for the emergency organizations.
We cannot establish with our measurements alone if the ICT infrastructure laid during the UNDAC mission is responsible for Bam’s apparent dodging of the regime’s efforts to quash the internet, but it is clear that they are an outlier in a nation that has otherwise been subjected, by its own leaders, to a crippling informational darkness.
Methods | Connectivity
To generate the data behind these observations, we combine a commercially available geo-located IP database with our powerful scanning technology which measures the online or offline status of millions of Internet addresses globally every hour.
Our observational methodology uses the most basic Internet messaging protocol that is widely used billions of times a day to establish routes for your email, tweet, or share. After developing a carefully selected set of Internet addresses (IPs) to measure, we periodically send them one of these tiny messages, essentially asking, ‘Are on you online?’. These online/offline answers form the basis for our ‘connectivity’ indicators.
In this visualisation, we give the average count of online IPs at a given location prior to the outage a benchmark score for what we call ‘connectivity’ of 100, and then base subsequent measurements on this standard. For example, a connectivity score of 32 means that we observe only 32% of the online IP connections of the benchmark period. We consider a connectivity score of less than 30% as a major disruption, with a score of less than 5% effectively offline.
Importantly, the IP Observatory has no access to any content being shared, viewed, visited, or generated by a user at a given IP, and, all IP Observatory activity works in aggregates of thousands of randomly sampled measurements across geo-spatial sub-regions.
Our visualisation of the internet crack-down in Iran follows our coverage of the Hong Kong protests, the Venezuelan pro-democracy movement and electricity outages, shut-downs in Ethiopia, the ICT impact of Cyclone Fani and Hurricane Florence, and the 2018 Russian Presidential elections.
The mission of the Monash University IP Observatory — ‘internet insights for social good’ — is to monitor the availability and quality of the Internet during critical events such as elections, natural disasters or conflict to provide. The observatory was founded by Klaus Ackermann, lecturer in Econometrics and Business Statistics, and Simon Angus, and Paul Raschky, Associate Professors in Economics. The observatory is a project of SoDa Laboratories at the Monash Business School, and tweets @IP_Observatory.
