ISI's Networking Group Creates Interactive Internet Outages Heatmap

December 7, 2017

Researchers at the Information Sciences Institute (ISI) Analysis of Network Traffic Lab are creating a geographic internet outages map to report the time, location, and severity of internet service outages around the world.

The project is led by ISI research director John Heidemann, a professor of computer science at USC’s Viterbi School of Engineering, in collaboration with ISI’s Yuri Pradkin and Dominik Staros.

The tool currently maps nine months of global internet conditions around the clock—from March 2017 to November 2018—including the internet impact of hurricanes Harvey, Irma, and Maria. 

The internet plays a pivotal role ensuring timely communication during natural disasters—ISI’s interactive map shows networks experienced significant connectivity issues due to high winds, torrential rain, and other havoc wreaked by these recent hurricanes. 

 In addition to helping scientists and policymakers monitor, measure, and understand internet connectivity – and how it can be affected by events such as extreme weather – the map is also an important public awareness tool.

“We want to make this tool user-friendly and interesting so that people without a technical background can view and access this data,” says Heidemann. 

“The internet is important in our lives—for many people and businesses, it is as important as any other critical infrastructure, so transparency is crucial. Getting this information into the hands of those developing disaster response plans is also vital.”

Outages reflect real-world problems

There are four billion IPv4 internet addresses in the world. To gather data for the map, Heidemann and his team pinged about four million address blocks every 11 minutes looking for outages. These raw observations were then mapped onto a geographic grid.

Using the map to explore internet outages is as simple as panning to the selected region, hitting “play” and “walking” through the timeline. Circles indicate regions affected by outages, while color changes indicate the percentage of network outages in the area. Users may also zoom in on specific regions.

“When Harvey hit, we could see a large percentage of networks north of Corpus Christi were out,” says Heidemann. “A few days later, we could see very large numbers of network outages in Houston. The scariest thing is when you see circles that are both large and red. Our tool shows the long-term effects of Hurricane Maria on Puerto Rico, where power outages have taken large parts of the internet offline for weeks.”

In addition to measuring the impact of natural disasters, Heidemann hopes quantitative measurements of storm damage could help first responders identify affected areas, particularly in neighborhoods that have lost access to communications services or power. These measurements could also provide information about the speed of service restoration.

The team’s work is ongoing—first, they will set their sights on implementing real-time capabilities. 

“In 2018, we hope to have near real-time reporting so that we can help people understand the scope of the problem within an hour of losing internet connectivity,” says Heidemann. 

The research is funded by the Department of Homeland Security (DHS) and Michael Keston, a real estate entrepreneur and philanthropist who, with his wife Linda, created ISI’s first endowed directorship position in 2015.

“We are very fortunate that Michael Keston stepped forward and helped make our outage data more accessible,” says Heidemann.  

“In addition to the DHS’s support of the basic research and ongoing data collection, this generous support will help us share our research with the world. Outages reflect problems in the real-world, so it is vitally important to understand problems to know where to allocate resources and, in the long run, improve internet reliability.”

The website was developed by Dominik Staros, ISI web developer and owner of Imagine Web Consulting, based on data collected by ISI researcher Yuri Pradkin. It builds on prior work by Pradkin, Heidemann and USC’s Lin Quan in ISI’s Analysis of Network Traffic Lab.