LTE's Immaturity in the US
The study, called "The Mother of all Network Benchmark Tests" -- which tested mobile broadband networks across the U.S., including 3G, LTE and mobile WiMax -- found striking performance differences between the LTE networks of AT&T Inc. (NYSE: T) and Verizon Wireless and concluded that most of the variance was due to the network equipment used and the immaturity of the LTE ecosystem, rather than the number of users on each network.
"LTE is in its infancy," said Mike Thelander, CEO and founder of Signals Research. "[There are] certain deficiencies that some vendors have that clearly manifest itself in what markets they're deployed in … It's because we're at the early stages of LTE and they make design trade-offs to get the equipment out to the market."
Both operators use LTE equipment from Alcatel-Lucent (NYSE: ALU) and Ericsson AB (Nasdaq: ERIC).
To illustrate, Thelander shared some insights from the tests he conducted in Houston using measurement kit from Accuver. The difference between the LTE networks of AT&T and Verizon Wireless in that market was like "night and day," and AT&T was better, he said.
But he explained that in other markets the results would be reversed and would favor Verizon Wireless, depending on which vendor's equipment had been deployed. This also means that LTE performance can vary among each operator's own markets. [Ed note: Light Reading Mobile did ask which vendor was better, but Thelander wouldn't specify.]
"I knew there were meaningful differences, measureable differences, and nuances," said Thelander. "But I wasn't expecting to see something so dramatic that you wouldn't even need test equipment to identify the supplier."
And these differences signify just how young and immature LTE technology is today. Performance varies depending on infrastructure supplier, chipset supplier and devices as well as on how much an operator has optimized its network.
"There are lots of inefficiencies in how [the vendors have] implemented LTE today," he added. "Over time, those inefficiencies will go away."
For example, Thelander pointed to differences in how vendors support multiple users simultaneously as well as differences on uplink speed.
In Houston, the average uplink speed on AT&T's network, which was measured as 15.2 Mbit/s, was up to three times better than Verizon's, Thelander said. (See AT&T LTE Test-Drive Results Revealed.)
Thelander stressed that the performance differences between the AT&T and Verizon networks had little to do with how many users were on the network at any given time. Considering that AT&T commercially launched LTE services in September and Verizon's services have been available since December last year, it would be natural to assume that AT&T's network performed better in Houston because it has fewer users. (See Verizon Sells 1.4M LTE Devices in Q3.)
But that's not the case. Thelander said that "90 percent of the difference has nothing to do with different traffic levels on the network."
Myriad factors affect mobile broadband network performance. And the conclusion to be drawn from this study is that as the supplier ecosystem matures and as operators optimize their networks, the big differences found in today's LTE networks in the U.S. will not be as great and overall performance will improve in the coming months even as more users are added to the networks.
— Michelle Donegan, European Editor, Light Reading Mobile