By Eric Griffith
July 12, 2004
This week in Portland, Oregon is the IEEE 802 Plenary Session, which includes a meeting of the 802.11 Working Group (WG) to define all the various Wi-Fi standards.
Among the many Task Groups within 802.11 that will be meeting at the show, there’s perhaps no standard more eagerly anticipated than 802.11n. Task Group N (TGn), under the guidance of chair Bruce Kraemer and 802.11 WG’s overall chair Stuart Kerry, is chartered to define “high throughput” speeds for wireless networks. The goal is a speed of at least 100 Megabits per second (Mbps) of actual throughput. This is not just the data-rate, which could be substantially higher. For instance, in 802.11g where 54Mbps is the data-rate, the actual throughput is about half that number.
Despite being a plenary meeting of the 802.11 WG, TGn is only just entering the proposal phase — but no official 11n proposals will be heard at the meeting this week, according to Kraemer.
“[This] meeting will have some technical presentations by various members to potentially introduce topics and get feedback,” says Kraemer, “but there’s no formal proposals until September. Proposals have to be posted by the IEEE server by August 13 so members can have time to review them.”
Already TGn has a list of respondents that includes 22 complete proposals — those that address all requirements, including changes to the the Media Access Control (MAC) and Physical (PHY) layers of 802.11 — and 39 partial proposals.
After September, all partial proposals will have to be rolled into complete proposals as TGn members pair up into groups. Such consolidation may take place even before September.
Kraemer refused to speculate on any IEEE expectations for 802.11n without going through the proposal process first, but for many of the TGn members — which are identified as individual technical experts, though most are there under the auspices of the companies they work for — the path to take toward high throughput may already be picked.
In fact, the first shot in the battle for 802.11n took place officially today, and it showcases the struggles that can take place when the highly democratic process of the IEEE is potentially subverted by corporations who want to see their technologies become standards.
This morning, Agere Systems of Allentown, Penn., a maker of 802.11 silicon, said it’s working to create a 500Mbps data-rate that would drive the 802.11n standard. This technology would use multiple-input multiple-output (MIMO) techniques and 40MHz channel width to get its speed, but dummying back down to the 20MHz channels used today for backwards compatibility. Using a 2×2 MIMO antenna (with two receivers on one end and two transmitters on the other), it would deliver a 250Mbps data-rate. With an optional 4×4 MIMO, the data-rate would approach 500Mbps.
“We want to make sure the coverage is robust enough for the whole home,” says Mary Cramer, Agere’s strategic marketing manager and the company’s representative to TGn. “Starting with 250Mbps gives a better chance to get 100Mb through the whole home. Some of the opposing proposals only squeak that through in a lab.”
She clarifies that this isn’t channel bonding, a technique used in chips by Atheros that combine use of two channels to increase throughput. “It’s two channels next to each other, but you fill the space in between them with data,” says Cramer.
The key is the 40MHz on every device, she says. The other major competing proposal (see below) would stick with 20MHz to ensure backwards compatibility. Some countries, like Japan, don’t allow use of signals for WLANs outside of the 20MHz channel, so TGn Sync [not spelled “TG nSynch” as previously reported] will have to dummy back to it for worldwide support anyway.
Cramer says, however, “You can’t cripple the standard for one country.”
The movement Agere has started here carries the moniker of TGn Sync. Supporters include Intel, Sony, Nokia, Philips and Atheros. In fact, in a story in EE Times that broke the initial TGn Sync announcement, Atheros was pegged as the leader of the group, but Atheros spokesman Colin Macnab says that the “most important thing is that it [TGn Sync] has a good grouping of people, 10-plus players…. we all win with a standard.”
World Wide Spectrum Efficiency
While Agere is making noise, Airgo of Palo Alto, Calif., is still playing things relatively quiet. The company is leading the charge on the other publicized 11n proposal, through a consortium of companies under the banner of World Wide Spectrum Eficiency (WWiSE). Airgo did talk to Wi-Fi Planet about some of its plans.
Airgo is a MIMO pioneer — company CEO Greg Raleigh published papers at Stanford on the subject in 1996, concurrent with work done by Bell Labs. Airgo argues that its MIMO-OFDM architecture found that the problem of multi-path signals could be used to increase throughput. The technique was further improved upon by VK Jones, who is the chief scientist and executive director of Communication System Engineering at Airgo.
Jones says that MIMO-OFDM is the “key ingredient across all the [11n] proposals, but there’s variations on the specifics, the changes to the MAC, to the channels… our product sticks to the 20MHz channels, [used in] the existing 802.11a/b/g standards.”
He says that going to 40MHz, which TGn Sync proposes, would increase throughput, but at the expense of interference and reducing overall system capacity.
Airgo didn’t want to talk about too many details of the WWiSE proposal. The competition says that WWiSE plans to use 4×4 spacial multiplexing MIMO with 20MHz channels.
Agere’s Cramer says of this: “[Airgo has MIMO] products with 20MHz now, they want to continue that. Not everyone is focused on having the best solution. They’re more focused on revenue.”
Airgo’s not alone in WWiSE, however. The EE Times article mentioned previously says that Broadcom, Conexant, Mitsubishi, Motorola, STMicroelectronics, and Texas Instruments are all on board.
Can’t We All Just Get Along?
Outside of the use of MIMO — which most consider a foregone conclusion for 802.11n — it appears that there’s one other area that everyone involved with TGn can agree on: no one wants to end up like ultrawideband. The groups arguing UWB proposals as the 802.15.3a standard became so contentious that they ended up splitting into competing groups.
“Previous discussions and debates took place on competitive technology because the approaches were significantly different. There’s no expectation that there will be significant differences going forward for 11n. That’s not to say that there won’t be differences of opinion… [but] we don’t anticipate massively different proposals,” says TGn chair Kraemer.
However, Agere’s Cramer believes what the WWiSE group proposal is a marketing ploy that doesn’t meet the requirements by the IEEE for RAND-Z (reasonable and non-discriminatory-zero licensing) access to the essential patents needed for the standard. It’s an issue that had an impact on the 802.15.3a/UWB talks, as well.
“If you don’t own the IP [intellectual property], how can you give it away for free,” Cramer asks. “They can’t stand up and say it’s a RAND-Z proposal.”
Airgo’s claims on what it developed for MIMO versus what Bell Labs developed are covered in papers published in 1996, but the specifics would have to be worked out in the IEEE meetings. Cramer says there’s companies that aren’t part of the IEEE that own MIMO patents, and that could lead to lawsuits.
So much for the non-partisan IEEE’s abilities to get past the big company rhetoric and machinations.
Still, Cramer — who spends about 50% of her working time on just 802.11 WG business — says “There’s no huge technical arguments outside of what you’d see from healthy engineers.”
Carl Temme, vice president of marketing at Airgo, says that with 61 TGn proposals already slated for September’s meeting, it’s not a two horse race. He says there’s much more to 11n’s future than just what TGn Sync and WWiSE have to offer.
Ultimately, there will be compromises as those with partial proposals are accepted as part of WWiSE or TGn Sync — or maybe get together to create another consortium, however unlikely that may be. For now, the first battle for 802.11n is barely joined. Come September, we’ll see how things really shake out.