Experts at the sharp edge of 5G development have identified a surprising problem: there's too much agreement.
Five years away from its expected commercial debut, a broad industry consensus has emerged about 5G's key features, in contrast to the dissension that surrounded 3G and 4G.
But NTT DoCoMo Inc. (NYSE: DCM), perhaps the world's most ambitious 5G operator, worries that the absence of a rival technology could cause its timetable to slip. The Japanese cellco is at the center of a national effort to offer the first commercial service in time for the Tokyo Olympics in the summer of 2020.
NTT DoCoMo has signed up eight vendor partners to expedite development and is already running trials, including a successful 4.5Gbit/s test with Ericsson.
Takehiro Nakamura, managing director of DoCoMo's 5G Laboratory, admits he has "concerns" about the timetable. "Competitiveness is important to promote systems development," he says. "We had a good competitor in 3G -- 3GPP versus 3GPP2. In the LTE case we had WiMax."
"Thanks to those competitors, we developed the new systems quickly," he adds. "Everybody could focus. But now, there's no competitor and many players have many directions."
He adds that while 2020 is important for Japan, other regions "have a very relaxed schedule -- not 2020, maybe 2023 or 2025."
Erik Dahlman, senior expert in radio access technologies at Ericsson AB (Nasdaq: ERIC), who took part in the 3G and 4G standards development, admits there is "some truth" in Nakamura's views.
He says while the issue of operator requirements, such as the 125-page NGMN white paper, has been helpful, the industry needs "some kind of push" to stimulate the process.
For more on 5G, visit the dedicated 5G section here on Light Reading, and register to attend the upcoming "Building America's 5G Ecosystem" event in NYC.
Korean operator SK Telecom (Nasdaq: SKM) is also anxious about meeting an Olympics deadline. The Korean government aims to launch a demonstration version of 5G at the 2018 Winter Olympics, much as China launched a pre-commercial version of TD-SCDMA in 2008.
Choi Chang-soon, senior manager at SK Telecom's 5G Tech Lab, says the timetable will be "the tricky part." His small team -- comprising just eight staff -- is hoping to set up a 5G test bed by year-end.
Despite the anxiety over the timing, the Asian cellcos are benefiting from the agreement about the likely shape of 5G, namely: LTE at the core; massive MIMO to take advantage of very high frequencies; latency as low as 1-2mms; and much greater energy efficiency.
As Dahlman puts it, 5G is a different beast. "3G was very much about new radio and 4G was a new radio technology, but 5G is much more than about radio," he says. "5G is about the networked society -- cars, traffic safety, washing machines, industrial robots, everything."
Next page: Radio wrangling
Radio wrangling
The radio part is still important, though, and there is division on two important radio issues. First, does 5G need a new radio technology for the very high frequencies? And second, should there be another one for IoT?
SK Telecom's Choi believes the use of high frequencies will require a fresh radio access technology (RAT). "We need to have a new RAT that fulfils the requirement of 5G, which is higher capacity and lower latency," he says. "And also to support beam-forming, which is also important to increase capacity."
A number of technologies are being put forward, such as Alcatel-Lucent (NYSE: ALU)'s UF-OFDM and Huawei Technologies Co. Ltd. 's OF-OFDM and SCDMA. (See Alcatel-Lucent, Huawei Pitch 5G Radio Technologies .)
DoCoMo has been experimenting with those technologies and others, including Filter Bank Multicarrier (FBMC) and Universal Filtered Multicarrier (UFMC). Nakamura says a new wave form may be required, but he thinks the best solution for higher frequencies may well be an OFDM-based single-carrier solution across 1GHz.
None of these have been verified, he points out. So it could be some time before the emergence of a front-runner, let alone a final agreement.
For IoT, industries such as wearables, healthcare and logistics cannot wait until 2020, and already a number of non-telco IoT networks have been formed. One of them, SigFox, in fact counts SK Telecom and NTT DoCoMo as investors.
The issue is whether to create a new technology for 5G or to evolve from LTE-M, an IoT standard now under development. (See Huawei Promises 4.5G & LTE-M in 2016.)
Tied up with this is also the question of what spectrum the IoT network will use.
"If they eventually use a new RAT for IoT, they might be able to use a lower frequency band," says Choi. "If they still keep using the LTE-M, then they're going to use the legacy spectrum."
Jim McGregor of Tirias Research believes it will come down to a choice between a "stripped-down LTE and LTE-lite protocol [or] a new radio access technology. The LTE-lite is the more likely solution."
But he also believes the 5G spec is "really more of a line in the sand than a hard and fast change in standards."
It's too early to determine what will make it into the 5G spec, and the specifications keep evolving. "Many of the proposals may actually be integrated into future 4G releases."
— Robert Clark, contributing editor, special to Light Reading