Upgrading to T-Mo’s 5G SA: Big bucks for
T-Mobile US made a splash with the announcement earlier this month that it had turned up Standalone 5G in an area encompassing 1.3 million square miles of its network. Signals Research Group wasted no time in getting out to test that shiny new network, although it required an expensive device and a hefty software upgrade: nearly $1,500 for the Samsung Galaxy S20 Ultra and an 800 MB software update.
Among its findings: SA 5G offered “some incredibly positive attributes” including reduced latency (partly from reduced signaling overhead) and modest but “meaningful” improvements to throughput.
SRG pointed out that the way T-Mobile US has designed its SA and NSA networks mean that the distinguishing factor in whether a SA-capable device connects via SA or NSA is based on the carrier’s mid-band LTE coverage — so if you already live in urban and suburban areas with robust mid-band LTE, you probably won’t connect to 5G SA much at this point.
“In other words, the phone takes advantage of the full capabilities of SA in areas with poor or no mid-band LTE coverage (likely rural areas) and it takes advantage of the full capabilities of LTE with 5G NR NSA in areas with good mid-band LTE coverage (likely urban and suburban areas),” SRG said, adding that “This is a great and prudent strategy, since otherwise the consumer would not benefit from the full potential of LTE + 5G NR in areas with good LTE coverage and likely ample use of LTE carrier aggregation.”
However, SRG also found that the SA device “could use NSA when it should be using SA and it could use SA when it should have been using NSA.” The testing and analysis firm also had to disable LTE Band 12 (700 MHz) to both smartphones in order to make sure that the phones attached to the 5G New Radio bearer.
5G SA offered some modest but “meaningful” speed improvements over 5G NSA. Interestingly, SRG found that on the Galaxy S20, there was “substantially less signaling” in SA mode than when the device was in 5G NSA — meaning that the device accomplished tasks such as connecting the network and handing over from cell-to-cell faster and with less latency.
SRG also said that based on its drive tests, it is “starting to question” how much 5G NR with SA in a low band frequency (like 600 MHz) will actually extend 5G NR coverage in real-world deployments, compared to 5G NR with NSA where the deployment hinges on the availability of LTE. 5G NR SA in the low-band only extends coverage “if and only if there is an area where the low-band 5G NR signal reaches a point that isn’t covered by a mid-band LTE frequency,” the company noted.
T-Mobile US, it should be pointed out, has claimed that turning up 5G SA in its 600-MHz spectrum increased its 5G footprint by 30%.
SRG used Accuver Americas’ XCAL-Solo and XCAL-M drive test tools to log the data and its XCAP post processing tool, along with Spirent Communications’ Umetrix data platform to generate the high-bandwidth data transfers used during its testing.
Read SRG’s report here.