The enterprise Wi-Fi industry has been in limbo this year. Everyone has known - more or less - what Cisco will do with Airespace. But everyone's been waiting to hear it. Trapeze tried to steal Cisco's thunder by integrating Cisco APs, and Aruba tried to change the subject, by producing products that Cisco hasn't thought of.

Other than that, the other vendors have been tweaking their OEM partnerships. Cisco will be building WLANs into its wired networks. Other enterprise WLAN vendors will have to do the same thing with other partners.

Integration like we expected
Cisco's announcement of its wireless LAN plan had few surprises. The logical thing to do with the Airespace technology was to put it in Cisco's wired switches (it was predicted long since), and the logical thing to do with the Airespace brand was to subsume it under the Cisco and Aironet brands.

But in the detail, we find a company that is concerned to manage those transitions very carefully.

"It's one of the first acquisitions where we bought a competitor," says wireless technology marketing manager Andy Oldfield. But it's not as simple as that. Cisco had a huge dominance of the market with its Aironet access points, but Airespace's product did something different.

Cisco bought Airespace because its centralised wireless LAN technology worked better than Cisco's own distributed wireless LAN technology, based on "fat" Aironet access points. Building wireless LANs with Airespace was also a lot cheaper than doing it with Aironet.

Why the Airespace brand will go
To take the brand question first. Airespace had the leading brand in centralised WLANs, but those centralised WLANs were a very small part of the total Wi-Fi market. In the overall market, Cisco's dominance is huge. Most customers won't care about the difference between distributed and centralised system. They've heard of Aironet, they haven't - by and large - heard of Airespace.

So the Airespace brand is going. Which is a pain for those of us who like to distinguish between the old, fat Aironet technology and the new, thin Airespace APs. Forgive us if we still use the term Airespace...

What about product plans?
Cisco had already launched a Wireless LAN service module (WLSM - pronounced "Wilsom") for the Catalyst 6500. It cost around $28,000, including software, to get started and it managed up to 150 old-style fat Aironet APs - in what Cisco called a "distributed" solution, not a "centralised" one.

Wilsom never set the world on fire, but Cisco is not withdrawing it. So the new Cat 6500 Wireless Services Module (WiSM, pronounced "Whizzum") has to be designed to do the Airespace job, but at not too keen a price - Whizzum has to be gentle with Wilsom.

It manages 300 access points, and starts at $46,000, with the ability to cluster and manage far more.

There's no doubt the WiSM could kill off WLSM. All new features will be implemented on the new architecture - for instance, Cisco has launched the 2700 wireless location box, based on Airespace development.

So greenfield sites with Cat 6500 switches will go straight for WiSM and thin APs. But even sites that already have some existing, fat, Aironet access points can go for WiSM, using a software upgrade that lets old-style Aironet APs work with the Airespace technology.

If WiSM is that good, why would anyone want Wilsom? "Some people will prefer a distributed solution," says Oldfield.

Cost comparison
Cisco is very emphatic that there are cost-savings to be had from shoving everything into one fault-tolerant switch. Oldfield has a slide that puts the five year cost of ownership at $190,000 for a WiSM-based switch, and $460,000 for a "non-unified controller" - a wireless switch outside of the main Cat 6500 switch.

There's no way to verify this flattering estimate, because Oldfield won't identify the "non-unified" controller: "We looked at a bunch of options from the industry" - presumably Trapeze or Aruba. It looks as if Cisco has simply doubled the cost of a maintenance contract for a non-integrated solution, and doubled the amount of unplanned downtime - and added $100,000 of planned downtime. "There's no planned downtime with a Cat 6500, because everything is hot swappable," says Oldfield.

Wireless in the branch
Cisco also launched a module for its ISR branch router, the do-everything box that has sold 750,000 since Cisco launched it last year. "The ISR is the fastest ramping product in Cisco's history," says Oldfield. He reckons it now has 80 percent market share in branch offices, and half the ISRs out there have a vacant slot ready for a wireless controller

The wireless LAN controller module (WLCM), which should obviously become known as "Welcome", manages six thin APs, which don't need to be configured individually. It's intended for IT staff remotely managing technology at hundreds of remote sites - and adds wireless to the portfolio they manage. It costs $2300.

And what about mesh?
The mesh announcement is a parallel item, and an entry into a new market.

The product has already met with scorn from Tropos, which also updated its product line, claiming support for carrier-grade voice.

In mesh, Cisco will have to work hard against the existing players, in product and marketing terms. Already, Cisco has disputed Tropos' claims about Cisco's products. While Tropos reckons providers will need 50 to 100 mesh APs to cover a square mile, Cisco puts the figure at 20 or 30.

What next

  • Expect similar announcements from other network vendors using Trapeze or Aruba as partners.
  • Expect efforts to build other things - like Aruba's Mobile Edge - into a distinct category that goes beyond what Cisco has.
  • Expect a rash of price comparisons that show what we already know - there's a premium for the Cisco badge
  • And expect case studies - lots of case studies - of enterprise WLANs.