Wanshi Chen is on the hot seat for 5G.
The chairman of the 3GPP’s RAN1 committee is tasked with delivering by the end of the year a draft for the next-generation cellular radio. The spec will form the blueprint for silicon needed to make the first standard 5G connection.
On one side, carriers and their vendors are calling for the specs ASAP so they can test and launch 5G services as early as next year. On the other side, as many as 800 engineers are showing up at meetings of Chen’s group, submitting as many as 3,000 proposals per meeting in hopes of getting a feature in the spec.
“Some sessions have run as late as 1 a.m., but a typical day is 12 hours,” said Chen, a principal engineer at Qualcomm who was elected chair of RAN1 in August after nine years attending meetings, four of them as a vice chair.
“We only have two [plenary] meetings to go and tons of stuff to work out. It’s hard to predict how late the meetings will run … I hope we can make it. In the last meeting, people tried to emphasize the sense of urgency.”
In an effort to increase their chances of finishing on time, engineers agreed at that meeting two weeks ago in Sapporo, Japan, to postpone until June at least 10 features originally in the spec. “I expected more reduction of the scope … it’s not to the level I’d like to see … [the still-large feature set] makes it difficult for me to get things done,” he said.
(For a full list of proposed and postponed features, find and click on document RP-172108 at this 3GPP page.)
The idea is to capture in the December draft everything required in hardware. “Anything after December has to be optional … with no hardware impact, but it’s hard to be 100% sure we’ve done the full due diligence … different features have different interest levels from different operators and vendors,” he said. “It’s hard to converge.”
Given the uncertainty, Verizon and KT (formerly Korea Telecom) launched separate efforts developing their own specs. KT defined in June 2016 its Pyeongchang spec named for the county where it aims to provide 5G-like services during the Winter Olympics in February.
For its part, Verizon rallied Cisco, Ericsson, Intel, Nokia, Samsung, and others around its 5GTF in late 2015. The spec aims to be the foundation for a last-mile wireless service for consumers that Verizon hopes to switch on next year.
“We had to have something to test … the 3GPP timing is still suspect,” said Sanyogita Shamsunder, executive director of 5G ecosystem planning at Verizon, in a brief interview on the show floor of the Mobile World Congress Americas earlier this month.
“We will track [the 3GPP work] and we want to work with the ecosystem, but we don’t want it to be the long pole, so we will continue to develop 5GTF,” she said, noting that she still has an option of using either one as the basis for planned 5G fixed-wireless access services next year.
Efforts at Verizon and KT helped motivate a consensus earlier this year to accelerate the 3GPP effort. At that time, a majority of stakeholders agreed to move the date for completing the first draft of the radio spec to December 2017, up from June 2018.
“We think we will see real 5G in mobility in 2019 … a year ago or less, it would have been 2020 or 2021,” said Rick Corker, head of Nokia in North America, one of Verizon’s top vendors.
Major developers of 5G silicon such as Ericsson, Intel, Nokia, and Qualcomm have all started work on chips. They need the final spec before they can freeze feature sets and start implementing their designs. To date, they have been providing carrier systems using FPGAs for their trials.
“The amount of pressure to have a spec to let silicon be developed is enormous,” said Michael Murphy, chief technology officer for Nokia North America. “With all the comments and change requests, it’s very, very challenging for the chair to manage.”
The tight deadline cuts into the time that engineers have to run simulations in the lab between 3GPP meetings, work that sometimes spawns new proposals. “As engineers, we want to get it right,” said John Smee, who works on 5G at Qualcomm Research. “These days, everything is simulated and evaluated.”
“Just from an engineering perspective, I’d rather have the deadlines more relaxed,” said Chen, who is taking a practical approach to the challenge.
For example, he asked participants from multiple companies to team up on proposals, hoping to spark strategic compromises. He also started assigning engineers to act as lead representatives for features “to manage and summarize what needs to be addressed and lead offline discussions so online talks can be more focused,” said Chen.
“Discussions can get out of control. The last meeting was my first as chair, so I provided a lot of my thoughts on how RAN1 should be managed, how online and offline discussions should be carried out, and how contributions should be written.”
“The key thing is making meetings more efficient and contributions more self-contained with the background needed for good solid proposals,” he added. “People want to hold on to their own proposals, but when it comes to compromise, there’s a good spirit of being flexible.”
To manage his own stress, Chen tries to start his day with a run of about six miles.
“I’m a very good runner,” said Chen, who finished the Boston Marathon in three hours and eight minutes back in April, about the time that the 5G schedule was kicked into high gear. “This pressure has cut into my mileage a little, but it’s still an effective way to relieve the stress.”
在线留言询价
型号 | 品牌 | 询价 |
---|---|---|
MC33074DR2G | onsemi | |
RB751G-40T2R | ROHM Semiconductor | |
BD71847AMWV-E2 | ROHM Semiconductor | |
CDZVT2R20B | ROHM Semiconductor | |
TL431ACLPR | Texas Instruments |
型号 | 品牌 | 抢购 |
---|---|---|
STM32F429IGT6 | STMicroelectronics | |
BP3621 | ROHM Semiconductor | |
BU33JA2MNVX-CTL | ROHM Semiconductor | |
TPS63050YFFR | Texas Instruments | |
ESR03EZPJ151 | ROHM Semiconductor | |
IPZ40N04S5L4R8ATMA1 | Infineon Technologies |
AMEYA360公众号二维码
识别二维码,即可关注