The Telegraph reports that a recent study claims that Google makes its websites faster than rival services by using code to dominate online networks. Researchers from Cargenie Mellon University and software firm Nefell Networks claim in the study that a Google algorithm can dominate almost half of a network’s capacity when there are multiple online services vying for connections.
The report claims that Google-owned websites such as YouTube could stream videos faster than other competitors such as Netflix if users sharing the same connection try to access both sites simultaneously. This algorithm can be used by other services meaning that Google is most likely not the only group doing this, but the research focuses mostly on Google’s role in managing online traffic.
Internet services use algorithms known as “congestion control algorithms” or CCA’s to share bandwidth efficiently across a network and to “throttle” or slow down certain websites or services when the capacity of the connection is stressed.
Researchers found that Google’s algorithm which is called BBR take up 40 percent of internet capacity when multiple services are competing for traffic. Researchers state when the algorithm was active, other services would be granted less than 4 percent capacity.
Google does “open source” the algorithm meaning that other services can take advantage of it. Other websites use different CCAs meaning that they could face slower internet speeds. Researchers state that Google’s algorithm, which went public in 2016, takes a disproportionate share of Internet traffic. Justine Shery of Carnegie Mellon’s computer science department states: “By any traditional version of [network] fairness, BBR is not fair.”
Google has stated that it will be address issues of fairness in a new version of the algorithm being launched soon called BBR V2. Read more about the study here.