It has been over two and half years since the Institute of Electrical and Electronics Engineers (IEEE) formed a task group to develop and implement a new 802.11 standard for wireless local area networks (WLAN). The proposed standard is known as 802.11n and is designed to eventually offer speeds up to 600 Mbps (burst mode, quad spatial streaming in the 5GHz band) with average data transfer rates around 200 Mbps and ranges extending up to 200 Feet indoors. The current shipping Draft N products advertise speeds up to 300 Mbps (burst mode, dual channel streaming) with data transfer rates reaching 130 Mbps and indoor distances up to 150 feet. The current 802.11g standard offers speeds up to 54 Mbps with sustained transfer rates around 24 Mbps operating at distances up to 100 feet indoors.

The proposed 802.11n amendment has had a colorful and storied history. This proposed standard does not have the publicity of the continuing HD-DVD/Blu-ray DVD drama. However, it does have similarities in the way two camps consisting of rival competitors have bickered over a next generation standard that should have been ratified by now. During the early creation phase of this project there were two camps jockeying to define 802.11n. These camps included the WWiSE (World-Wide Spectrum Efficiency) whose primary team members consisted of Airgo, Broadcom, and Texas Instruments along with TGNSync that was led by Intel, Atheros, and Marvell.

After months of bickering, these two groups along with the newly formed MITMOT (Mac and Mimo Technologies for More Throughput) team consisting of Motorola and Mitsubishi merged into a single group known as the Joint Proposal Team. This was a short lived alliance as the team fell apart with Intel, Atheros, Marvell, and Broadcom splintering off to create the Enhanced Wireless Consortium (EWC).

The EWC did what the other members and teams could not as they were able to get the IEEE 802.11n task group to approve their specification as the proposed 802.11n standard in January of 2006. In March of 2006, the IEEE committee sent the proposal known internally as 802.11n, Draft 1.0, to its first letter ballot. The Draft 1.0 proposal failed miserably as it could not even gain a simple majority vote in favor of the proposal much less the required 75 percent majority required for ratification.

Over 12,000 comments were received from various IEEE members citing standards issues, operating frequency conflicts, and an overwhelming list of deficiencies (concerns) that include everything from power saving guidelines for handheld devices to proper streaming techniques for audio and video in media servers. In May of 2006, the IEEE 802.11 committee decided not to forward the Draft 1.0 proposal for a full sponsor vote and instead sent the 802.11n task group back to the drawing board to devise a new standard that would meet the concerns of the IEEE membership.

However, the primary members of the EWC that backed the 802.11n, Draft 1.0, specification started shipping "Draft N" product into the market shortly after the proposal was sent to its first letter ballot. This is a decision that could potentially haunt manufacturers and customers alike depending upon what the next draft proposes in the way of compliance to two main issues. While the current suppliers of Draft N products would like you to think that firmware and driver updates could potentially ensure their product is compatible with the upcoming standard there is no guarantee the products will work with the new standard properly, much less each other.

Out of the 12,000 comments about half of them have been resolved at this point through editorial content changes in the initial draft with a significant amount of the remaining comments being duplicates. However, there are a couple of issues that are at the crux of the current controversy. The two main issues with the 802.11n, Draft 1.0, proposal that are being remedied at this time include proper Clear Channel Assessment (CCA) and power-save multi-polling standards for handheld devices. The most critical issue currently is how to resolve the "bad neighbor" effect of the current Draft N hardware.

In essence, only three of the eleven 20MHz channels in the current 2.4GHz band where 802.11b, 802.11g, and proposed 802.11n standards operate do not overlap in most instances. These channels are currently 1, 6, and 11. The 802.11n Draft 1.0 proposal currently stipulates these 20MHz channels in the same frequency band while allowing for the optional 40MHz channel in the same band. This 40MHz channel capability is due to 802.11n defaulting to the use of two 20MHz channels of bandwidth via a throughput enhancing design known as channel bonding that appeared in the original MIMO (Multiple In, Multiple Out) products. The 40MHz signal can potentially overlap all of the 20MHz channels in the 2.4GHz band which allows for the disruption of any devices operating on the 20MHz channels. This creates the "bad neighbor" effect as you have the Draft N devices stepping on the legacy product signals, and this has been one of the major issues with the current release of products.

In order to solve this issue, the current recommendation is to still permit the use of the 40MHz channels but to "ensure" a method of detecting legacy networks operating on the same channel and then having the 802.11n device fall back, even temporarily, to a single or different 20MHz channel in order to avoid a collision. Current Draft N products do this to a certain degree; ensuring this rule occurs is the preferred solution but it will render most of the current Draft N devices useless with the new standard. This change will require additional hardware capability or a silicon change to properly monitor and then switch channels on the fly to guarantee proper CCA is being performed. Also, 40MHz operation will have to be optional as this channel is not available for use in Japan and sections of Europe at this time.

We currently expect the differences to be resolved in time for a second draft to be completed this November with a letter ballot going out in January of 2007. If this is approved we can expect final ratification in late 2007 or early 2008 although we very well could see a Draft 3.0 in the spring if real progress is not made this winter. However, the products released that meet an approved Draft 2.0/3.0 specification are almost certain to be fully compliant with the final 802.11n amendment. Let's see how our first batch of Draft N products performs at this time.

Feature Set and Options: Draft N Routers
POST A COMMENT

22 Comments

View All Comments

  • shoRunner - Wednesday, August 30, 2006 - link

    As hinted at in the article the overall reliability of these draft-n routers is terrible. Having setup 60+ wireless networks in the past few months using many different kinds of routers including these draft-n routers, they have performed very badly some models requiring daily powercycling and constant firmware updates. If you are looking for a reliable fast wireless network the netgear 240 pre-n router is definately the better buy. Reply
  • Myrandex - Wednesday, August 30, 2006 - link

    The Dlink DGL-1000 router has gigabit ethernet and is freaking amazing. Not to mention I enjoy the blue LEDs on the frong, and performance is nice.
    Jason
    Reply
  • blckgrffn - Wednesday, August 30, 2006 - link

    If were are going to sustain 300 megabit throughput on our wireless devices, why isn't the wired backend gigabit?

    Seriously, early adopters of this stuff are also likely to have gigabit networking equipment, as that has been shipping in volume for the at least the last three years or so, and really became affordable as far as switches go last year. My $30 D-link gigabit switch has been working just fine...

    Nat
    Reply
  • bobsmith1492 - Wednesday, August 30, 2006 - link

    On page 2, the feature chart states the three routers have 2.4 GHz bandwidth... I believe that is actually their operating frequency. Reply
  • erwos - Wednesday, August 30, 2006 - link

    The spectrum nuking issue is a real concern to me. I live in an apartment building, so I'm already getting crowded by random wireless phones and microwaves all around me. I _shudder_ to think what will happen when some of these "draft 802.11n" devices become more common. I wish I could claim this kind of callousness was because of 802.11n, but I know it's not true - the original channel bonding schemes for 802.11b/g were infamous for this kind of thing.

    I'm trying my best to be a good citizen and turn that sort of stuff off, but I fear I may to have move to 802.11a, and the less-troubled 5ghz band, soon.

    -Erwos
    Reply
  • DigitalFreak - Wednesday, August 30, 2006 - link

    I moved to 802.11a a long time ago, after more and more b/g APs started showing up. Been running great every since.

    I really have to wonder why the IEEE didn't use the 5ghz frequency for 11n. I know 11a has a shorter range than 11b/g, but I would think it would be easier to overcome that problem than it is to get past the major spectrum issues in the 2.4Ghz range.
    Reply
  • yyrkoon - Wednesday, August 30, 2006 - link

    Well, atleast not here in the US I mean. Reply
  • yyrkoon - Wednesday, August 30, 2006 - link

    2.4GHZ isnt regulated, 5.8GHZ may be, I'm not sure. Reply
  • Lonyo - Wednesday, August 30, 2006 - link

    802.11n might be useful in the home eventually as broadband gets faster (30mbps+ connections), but for real high speed networking, it seems wired is still the only option.
    Can't say it's so suprising, but at least wireless is getting more useful in terms of matching increasing broadband speeds (although with existing MIMO, .11n isn't quite so useful yet, until it can exceed MIMO).
    Reply
  • LoneWolf15 - Wednesday, August 30, 2006 - link

    Are you kidding? There are far more reasons for high-speed wireless in the home than just broadband. Streaming media servers (having all my movies, music, etc. on a server that can be streamed to an HTPC or appliance) are a prime example of a good use of high-speed WiFi, especially for those of us that don't wish to deal with rewiring CAT-5 in our houses.

    As for .11n, it isn't useful yet because there isn't a standard, and yet vendors are trying to capitalize on a need by releasing hastily-designed pre-standard product. It's not robust, and it requires proprietary hardware. When the IEEE finally figures this out (IMO, it should have been some time ago, they've had enough time, though dealing with bickering vendors is an issue) and issues a true standard, things should work out better, much like when V.90 finally was ratified.
    Reply

Log in

Don't have an account? Sign up now