New York City is the city that never sleeps, and if a new study from the University of Southern California’s Viterbi School of Engineering is any indication, it’s home to plenty of active Internet connections around the clock. But there are plenty of places around the globe where the Internet actually does sleep at night.
While the United States and parts of southern Africa and Western Europe display pretty much constant Internet connectivity, countries such as Armenia, Georgia, and Belarus follow a diurnal usage pattern that sees Internet usage peak as the day progresses and then taper off at night.
As interesting as that might be, it’s far more fun to look at the data visually, as evidenced by the animated GIF above that shows how patterns of Internet usage change while the night sweeps across the map. For reference, pinkish/reddish blocks indicate higher than mean Internet usage, while blue blocks represent lower than average activity.
The data was compiled by researchers at USC Viterbi’s Information Science Institute. Over the course of two months, ISI’s Lin Quan, John Heidemann, and Yuri Pradkin studied patterns of Internet usage by pinging 3.7 million IP address blocks—representing almost 1 billion of the world’s roughly 4 billion IPv4 addresses—every 11 minutes.
Given the data, the researchers suggest several conclusions. For one thing, diurnal usage seems to be correlated with countries with lower gross domestic product. That’s not particularly surprising, given that countries with stronger economies are often home to broadband Internet connections that are always on, rather than dial-up links which might be shut down at night. Likewise, countries and regions that attempt to conserve more electricity by turning off equipment at the end of the day may also display diurnal usage patterns.
As for practical applications of this research, one place it may come in useful is in establishing a daily baseline for Internet activity—a sort of background radiation reading, if you will. That might allow for easier identification of certain anomalies, such as network outages, and help avoid confusing them with natural patterns of use. Although the study doesn’t suggest it, it seems like it may also be handy for identifying areas where energy could be saved by minimizing constantly active Internet connections that might be more able to lie dormant during off hours.
At least the Internet’s lack of sleep doesn’t seem to be having the same ill effects as it does on people, although long nights of browsing Wikipedia, playing online games, and watching cat videos on YouTube has likely contributed to plenty of drooping eyelids the next morning.