U.S. 5G network tests show early growing pains – RootMetrics

This article originally appeared on FierceWireless by Bevin Fletcher.

n the early days of 5G launches, initial tests indicate U.S. carriers are experiencing some growing pains with inconsistent coverage and speeds, according IHS Markit’s RootMetrics latest report.

That’s not to be unexpected, as new technologies need time to mature and reach their potential, RootMetrics noted in its 5G First Look report. Still, tests in the U.S. cities of Atlanta, Chicago, and Dallas show 5G availability and speed vary widely among the country’s four major carriers, and in terms of consistency, are well behind that of operators in South Korea.

“While there is justifiable excitement surrounding 5G, initial test results suggest that several early growing pains must be worked out,” wrote RootMetrics in its report.

Sprint was the only carrier that had a meaningful 5G footprint across all three U.S. test cities and maximum 5G download speeds all hit at around 200 Mbps. In Chicago, Sprint’s 5G median download speed was 123.5 Mbps, meaning a user could download a 600 MB video in about 40 seconds, compared to 3 minutes on LTE. However, results were mixed in Dallas where Sprint’s maximum 5G download speed of 196.7 Mbps was slower than 200.4 Mbps peak over 4G LTE, and its median 5G download in the city was an incredibly slow 1.2 Mbps.

“Our tests showed that Sprint’s 5G network in the downtown area of Dallas experienced network problems while using 5G for download tasks and/or transitioning between 5G and LTE, which likely impacted the carrier’s results,” RootMetrics wrote.

In terms of meaningful coverage, it’s not that surprising Sprint has more availability given it’s using 2.5 GHz spectrum, unlike high-band millimeter wave spectrum used by the other three carriers. Sprint’s sub-6 GHz spectrum is known to deliver enhanced speeds with more reliable coverage because signals can penetrate objects like buildings, while mmWave delivers super-fast speeds, but its range is very limited because of poor propagation characteristics.

Notably, testing found Sprint’s 2.5 GHz spectrum had “remarkably similar” propagation as the 3.5 GHz spectrum used by South Korean operators KT, LG U+ and SK Telecom.

“This similarity in spectrum potentially portends good things for Sprint’s 5G network because speeds in South Korea were incredibly fast in general and much faster than what we found in the US,” wrote RootMetrics in its report.

Finding a 5G signal for AT&T, T-Mobile, and Verizon across all three cities posed challenges, with 7.1% 5G availability for Verizon the highest found among those three carriers in the test cities. For T-Mobile and AT&T, 5G connections were available roughly 3% or less of the time. Atlanta was the best city for Sprint’s 5G availability, with a 5G connection slightly more than half of the time, while the carrier’s lowest availability, found in Dallas, was still nearly 25%.

Compare that to 5G availability rates for each South Korean operator ranging from 42.2% to 45.6%. Peak and median 5G speeds were also higher and more consistent across South Korean operators, with U.S. 5G deployments comparatively only “in their infancy”, according to RootMetrics.

It should be noted the U.S. test results were collected between late July and mid-August, shortly after 5G networks had officially launched, so continued improvements ahead are likely and already underway.

RootMetrics 5G graph

(RootMetrics)

On Verizon’s Chicago network, for example, 7.1% of tests were connected to 5G (compared to 32.2% for Sprint), but it still had the most mmWave 5G among all carriers. A Verizon representative in late September told FierceWireless that the carrier had doubled the number of mmWave 5G small cells in its Chicago market since launching there in April, which added density and capacity, as well as extended coverage.

Verizon also delivered the fastest maximum 5G download speed of 1.1 Gbps in the Chicago market, and it was the only carrier to hit about the 1 Gbps mark in any city.

T-Mobile, meanwhile, produced strong maximum 5G download speeds of 257.7 Mbps in Atlanta, but median speeds of only 20.9 Mbps. In Dallas, maximum 5G downloads clocked a measly 18.5 Mbps on T-Mobile’s network. However, those low figures could be a result of T-Mobile’s very limited 5G coverage, as RootMetrics noted tests in Dallas connected to 5G only 1.3% of the time and 3.1% of the time in Atlanta. Still, LTE performance in Dallas was strong, with maximum download speeds of 220.8 Mbps.

T-Mobile is using millimeter wave spectrum in initial launch cities, but like AT&T, has promised to roll out broader 5G coverage using sub-6 GHz spectrum.

As for AT&T, users can expect faster 5G speeds but only if they can find a signal. Median 5G download speeds in the city hit 256.1 Mbps, more than 13 times the speed of LTE downloads, while peak speeds were the second-fastest found and hit just under 670 Mbps. That said, RootMetrics could only connect to AT&T’s 5G network in Dallas 2.6% of the time, though the company was encouraged by excellent reliability and speed when 5G was available.

RootMetrics’ results are based on the company’s own drive testing and data collection, rather than user-initiated tests. Across the three cities this includes more than 20,500 samples collected, over 500 miles driven and about 40 indoor locations visited between late July and early August, testing on a range of devices including the LG V50 ThinQ 5G, Samsung Galaxy S10 and Samsung Galaxy S10 5G.