Receive new posts as email.
This site operates as an independent editorial operation. Advertising, sponsorships, and other non-editorial materials represent the opinions and messages of their respective origins, and not of the site operator or JiWire, Inc.
Entire site and all contents except otherwise noted © Copyright 2001-2006 by Glenn Fleishman. Some images ©2006 Jupiterimages Corporation. All rights reserved. Please contact us for reprint rights. Linking is, of course, free and encouraged.
Tim Higgins posted a long review of NetGear’s RangeMax 240 which uses Airgo’s 3rd generation MIMO chips: He found that despite the promise by Airgo for this chipset to back off from using the equivalent of two Wi-Fi channels (40 MHz) when it detected other Wi-Fi networks on active channels nearby, current firmware fails at this task. He has the charts and graphs to show it. He also notes that Airgo and NetGear are aware of this problem.
He did find that the devices can deliver over 100 Mbps of real throughput, which is a first for any Wi-Fi-like device, and more significant still in a consumer-level product. (The NetGear has just 10/100 Ethernet switching, so it can’t deliver more than about 94 Mbps in Tim’s testing to a single port.)
If there are any other 802.11b/g networks in the vicinity, it makes no sense to install this generation of Airgo-based equipment until firmware problems are resolved. eWeek reached the same conclusion (less exhaustively but just as completely) two weeks ago.
Even when they are resolved, the Adaptive Channel Expansion (ACE) algorithm is somewhat different than what I heard from Airgo in Sept. 2005. Tim writes, “I should also note that Airgo told me that when it is working, it will take 5 to 10 minutes (!) to tune away from a neighboring WLAN that is detected after the RM240 completes its initial power-up sequence—if the RM240 sees “lots of continuous traffic” in the neighbor.”
In Sept., Airgo’s director of product marketing, Dave Borison, said that ACE didn’t create negative effects on neighborhing channels, and that it performs a frame-by-frame check on adjacent channels—but apparently, this just applies to b/g clients on the Airgo-based device’s network. Clearly, the current firmware in Tim Higgins testing doesn’t live up to that promise yet for adjacent networks.
802.11n finally moves forward: The IEEE task group on high-throughput wireless local area networking has confirmed the joint proposal group draft which itself came out of the Enhanced Wireless Consortium. Now 802.11n will move forward relatively rapidly to ratification, even though that formal process of finalizing details could take until 2007. That won’t delay shipping products at this point.
Broadcom meanwhile announced that what it’s dubbed its Intensi-fi chips are now available in sampling and incorporated in reference designs for manufacturers and support all mandatory draft 802.11n specifications. The chips will also support any changes in the spec through ratification via software updates. The chips will support over 300 Mbps of throughput.
The vote was 184 to 0 with four absentions, which is a quite remarkable outcome over years of jockeying, vitriol, and drama—and a fine contrast to the disbanding vote today in 15.3a for UWB.
The Joint Proposal (JP) team in the IEEE 802.11n high-throughput wireless task group has accepted the Enhanced Wireless Consortium (EWC) proposal: This is the penultimate step to the EWC proposal, originally developed outside the main standards process by the four overall largest chipmakers, and which eventually included most parties in 802.11n. Task Group N will produce a standard that uses multiple antennas and a host of other strategies to produce at least 100 Mbps of raw throughput, but more likely 200 Mbps. Net throughput, or real data transferred, will be much higher than the current 50% of raw data passed across.
The JP at the task group was composed of the members of two previously at-loggerheads proposals led (in the least nuanced way I can put this) by Intel on the one side and Airgo, leading MIMO chipmaker, on the other. The EWC proposal had many points of similarity with work in the JP. The acceptance of the EWC proposal by the JP sets the stage for supermajority voting of 75 percent that will allow the JP to become the draft upon which final ratification of the standard is based.
The vote passed the JP group 40 to 0, with two members not casting ballots. The JP includes all major players in the industry.
After last week’s CES, I predicted that the Enhanced Wireless Consortium had the votes to win: The EWC, founded by four major chipmakers, has swept most of the industry and most IEEE 802.11n voting members into its camp. The EWC draft will be introduced (I believe by way of a joint proposal group that was already in progress) next week in Hawaii, this article reports. With several chipmakers already having EWC-based prototypes for testing or in sampling, EWC seems a fait accompli.
This article is the first I’ve seen that makes the Airgo side of the equation seem less unbalanced. The company obviously fought hard for some of its most critical ideas to be included, and I see less unhappiness about the result in statements by their CEO. I had no doubt that Airgo can adapt and produce chips that conform to whatever proposal is adopted, but these statements lead me to think we will see no schism and little delay in moving toward industry-standard high-throughput devices.
(Updated) At CES, Atheros and Broadcom showed working products based on early 802.11n chips: The demonstrations were off the show floor in private suites. These chips—along with Marvell, which had a chip to show but no working product—use the Enhanced Wireless Consortium (EWC) proposal as the basis of the silicon.
Atheros, Broadcom, Intel, and Marvell formed this private, originally secret group to cut through a roadblock in the 802.11n task group, which has a goal of higher throughput for future Wi-Fi standards. The EWC was criticized for working outside the IEEE process, but their proposal now has dozens of members of the 802.11n task group signed on.
Marvell’s chip isn’t in sampling yet—meaning it’s not available for manufacturers to start building products around—but a representative at the booth said it would be out any day. I was not invited to Atheros and Broadcom’s demonstrations—if I had been, I would have been under non-disclosure, too—but I expect they are in a close to similar state. The companies could produce the chips in quantities in a few months, meaning that 802.11n-like consumer products could be out as early as May or June.
The sense I got from being on the floor at CES and talking to a number of Wi-Fi equipment makes is that the EWC proposal will easily capture the majority necessary to move to a vote to accept a draft, and that it then has the 75-percent supermajority votes, too. Airgo is the only major wireless data chipmaker that hasn’t signed on to the EWC.
I’m never quite sure how the IEEE establishes rounds of votes, but if both the votes are held next week, the EWC version will win the day and move forward. The IEEE’s Web site puts finalization as early 2007 based on certain assumption.
Buffalo will ship 240 Mbps MIMO gateway, PC Card in February: It uses Airgo’s third-generation chips. The gateway will have a street price of $149; the adapter, $99.
NetGear previously announced its MIMO equipment in Nov.: They’re showing their Airgo-based gear at CES, too, but a shipping date doesn’t appear to be available. They were originally expecting to ship by Christmas.
Meanwhile, Atheros talks about the next-next-generation MIMO: The company is demonstrating a 3x3 array which will provide 300 Mbps of raw bandwidth, or about 25% more than Airgo’s gear. But we don’t know about the net throughput: Airgo says 100 Mbps for their system. (For more on what 3x3 means, see this Wireless Net DesignLine article that spells out why 3x3 might be the better route for 802.11n without dramatically increasing complexity or cost for chipmakers and gateway manufacturers.)
Airgo says 1.5 million devices using their chipsets were produced last year: That’s a nice number, as some companies list “chips” not “chipsets,” and can overcount devices. They expect 5 million “True MIMO” devices to be installed (not just produced) using their technology in 2006.
Airgo’s latest generation of multiple-in, multiple-out technology features up to 240 Mbps of raw throughput by using two spatially separated datastreams that can dynamically expand on a packet-by-packet basis from 20 megahertz (one Wi-Fi channel) to 40 MHz when the airspace is unused.