The analysts at Signals Research Group (SRG) aren’t making any predictions about the future success of open Radio Access Network (RAN), but their tests of the Dish Network in Las Vegas show some inconsistencies in performance.
The illustration on the cover of the June 22 report sums it up. It depicts two versions of Dish Chairman Charlie Ergen: One where he is seemingly calm, cool and collected using the equipment of long-time infrastructure vendors Nokia and Ericsson, and another who looks nervous using unproven equipment from open RAN vendors. The title of the study: “Hold 'Em or Fold 'Em.”
In what may be Dish's biggest gamble yet, the company is building what it describes as the “world’s most advanced cloud-native 5G open RAN network,” which means it’s using open RAN gear from a lot of vendors as opposed to handing over its network build to infrastructure veterans like Ericsson and Nokia. Incumbent U.S. wireless operators use equipment from Ericsson, Nokia and Samsung and they’re dabbling in open RAN, but none of them are committed to open RAN like Dish.
SRG’s tests were done over two days in late May. They used two Motorola Edge+ smartphones on the Dish network and two Samsung Galaxy S20 smartphones on the T-Mobile network. Tests were conducted in 5G Standalone (SA) mode to make it easier to compare results between the two networks as well as Non-Standalone (NSA) mode when conducting the voice calls. That meant voice calls on the T-Mobile network were VoLTE and while on the Dish network, they were Voice over New Radio (VoNR) calls.
Scores on the T-Mobile VoLTE network were consistent and overall higher than on the Dish VoNR network. Calls on the Dish network varied, such that a poor voice quality call could occur immediately after a great voice quality call and while stationary in good RF conditions.
“Given what we know and what we suspect, we believe there are equal opportunities for all parties to improve the network performance,” wrote SRG in its report summary. “The overall network performance wasn’t terrible, and there were some indications of quite good performance, but there was no consistency in the performance and consistency is something that consumers expect.”
SRG said their biggest concern with the downlink performance was the on-again/off-again use of carrier aggregation (CA). “Since we weren’t able to log chipset data on the Motorola Edge+ smartphones we don’t have great visibility into what was happening. We believe part of the problem was poor RF conditions, but it wasn’t the only explanation since CA functionality got triggered/disabled even with good RF conditions while stationary,” the study said.
To conduct the tests, SRG collaborated with three test and measurement companies: Accuver Americas, Spirent Corporation and Rohde & Schwarz.
Boils down to coverage
SRG CEO and founder Mike Thelander said they tested both voice and data in the Las Vegas market, which was Dish’s first to launch and where it uses 600 MHz as a primary channel.
“Definitely, Dish has a lot to do to improve their coverage, both in terms of actually getting the signal there but also making sure it’s a good signal quality,” he told Fierce. “Bottom line on the drive testing, definitely the poor network coverage had a big impact on testing, both on the downlink and the uplink. But there were other issues, I think, like carrier aggregation that were separate from what was happening with the radio conditions.”
Earlier this month, Dish said it met its deadline for reaching 20% of the U.S. population as required by its deal with the U.S. government. But analyst Roger Entner, founder of Recon Analytics, said that’s the easy part. Meeting next year’s deadline to reach 70% of the population also won’t be too difficult, but the hard part will be meeting the coverage requirement for 75% of the population in each of the 400 or so Partial Economic Areas (PEAs) by mid-2025.
And even though Dish has network service agreements with both AT&T and T-Mobile, the latter of which was the subject of headlines this past week, Thelander said it was fairly common for the calls SRG was making in Las Vegas to roam onto the AT&T network.
“You’re driving along and you’re testing, and you’re on the Dish 5G network and then all of a sudden, you’re on the AT&T network,” he said.
Surely there are places both inside buildings and outside where calls will connect, “but their coverage, quite frankly, isn’t that good,” he said.
New Street Research has estimated that Dish’s new wireless network covers just 125,000 square miles with less than 4,000 cell sites to meet the 20% nationwide buildout requirement.
Speaking with Fierce this week, Dish Wireless President and COO John Swieringa would not confirm any specifics. He said Dish is due to make a July 14 filing with the FCC about its 20% network buildout.
Of course, deploying a 5G network is just part of what Dish is doing. It’s also building a wireless retail business, and it’s hinted at big things in the enterprise space. But how fast it all happens is anyone’s guess. Thelander recalls that Dish initially was supposed to launch quite a while ago.
“Dish is doing 5G standalone. They’ve got Amazon doing the cloudification of their core network. They’re doing carrier aggregation. They’re doing Voice over NR. That’s a very high degree of difficulty,” Thelander said. “On the one hand, you’ve got to cut them some slack. But on the other hand, consumers don’t care if they’ve got it hard. They just want a good experience and a consistent experience.”
One could even conclude what it’s accomplished is impressive given all the challenges. However, “if you’re paying for the service, you don’t care. What you care about is getting a good user experience… It can be good,” especially if you’re stationary and in a good RF environment. But it can equally be “quite bad,” he said.
It’s an open question as to how much time it should take. “It could be one software upgrade, and a lot of these problems are solved,” Thelander said. “But the network coverage thing, that’s real and that’s not going to be easy.”