Free Mobile last in 5G: the operator criticizes the Arcep study
Arcep recently published its study on the performance of mobile networks and it turns out that free mobile is in last place, whether for 5G or for other media. The operator criticizes the methodology used by the telecoms policeman.
Free Mobile does not like Arcep’s measurements for its 5G test
It is on Twitter that Free Mobile, using its account @Free_1337 dedicated to its network, decided to tackle the study on the quality of 5G. Arcep noted in particular that the average speed of Xavier Niel’s operator was 31 Mb / s, against 142 Mb / s at Orange, 85 Mb / s at SFR and 71 Mb / s at Bouygues Telecom.
“These disparities result from the measurement protocol. The download speeds are very different from the most used tools, Speedtest, RNCMobile and NPerf, as in real use ”, starts Free. The group explains that the Arcep protocol consisted of using a single TCP Cubic single-threaded connection to measure the download time and then deduce the throughput. “What is therefore measured is the rate at which the Cubic congestion control algorithm self-regulates its speed over a single connection”, indicates Free before giving more details in a tweet:
4/11 TCP CUBIC, from 2008, is based on packet loss but can add delays and degrade performance. TCP BBR, from 2016, which becomes the majority, uses estimates of bandwidth and round trip time (RTT) and approaches the theoretical optimum
– Free 1337 (@Free_1337) November 22, 2021
Free Mobile defends itself by ensuring that its network “Has been optimized for use with several connections (multithreaded)”, noting in passing that several applications use several connections, including those for video playback. The operator says it can go up to 200 connections.
Any changes for the next study?
In addition, Free Mobile talks about the servers used to do the tests:
8/11 but with servers using TCP BBR and the measurement application opening one or more TCP connections
(table with mono and multithreaded results). pic.twitter.com/Wyo2J2f6q0– Free 1337 (@Free_1337) November 22, 2021
9/11 Conclusion: when the servers use TCP BBR and / or the measurement application opens several TCP connections, the Free Mobile speeds are similar or superior to those of the competitors.
– Free 1337 (@Free_1337) November 22, 2021
The operator concludes by explaining that in real use, “Free Mobile subscribers have high speeds in 4G, and even more in 5G”. He adds : “ARCEP’s measurement protocol does not reflect the actual uses of mobile subscribers and the speeds of Free Mobile subscribers. The next test campaigns will have to evolve to be closer to the realities in the field ”.