This year, the most important thing has been the formation of partnerships between startups and established networking vendors. Airespace has Nortel, Alcatel and NEC, Trapeze has a relationship with 3Com and Aruba has announced a partnership with HP. What has the impact of these partnerships been?
Well, it gives the startups a little bit of credibility in that now they have at least a handshake relationship with somebody with a recognized name. It also gives them another sales arm that's out there pushing the equipment. And now they have support organizations. So, if Nortel sells and supports Airespace, Airespace still gets the sale. [The product] is not rebranded, so Airespace gets the mindshare of the customer, which may help when upgrades are necessary. And Airespace doesn't have to worry about providing support. What are the benefits of these partnerships?
Building these systems is hard. Nortel found that out and they gave up on it. Proxim, of all people, gave up on it. If you're a vendor, it gives you headaches. Cisco found this out with its Wireless LAN Solution Engine (WLSE). Speaking of Cisco, in your report you wrote that the company had made "some much-needed upgrades" to the WLSE. What improvements do you think Cisco has made?
Let's take location services. Their first [WLSE] upgrade went from not having any (location services)
Also, the older version couldn't import floor plans, or wouldn't let you do any site planning with them. Now the current version can do that. Cisco has come a long way with the WLSE. I still think it's just OK overall, but given what it was a year ago and what it is now, it's a dramatic improvement.
In some respects, Cisco is finally catching up to where Airespace was two years ago. They're in good enough shape so that an all-Cisco shop won't bolt for the door. But if you have strict performance requirements, strict location, security and planning requirements, there are better solutions. What is your opinion regarding "smart" vs. "dumb" networks? Do you think the industry is on the path to dumb networks with intelligent switches at the edge?
I believe the intelligence needs to be at the edge and in the end-point device, not in the network itself. The network edge is the place where you need to look at the infrastructure that's on each of these portable devices or desktop devices and find out if it's current. It's much easier to take pot shots at a burglar when he's outside your house than to fight a rear-guard action once he's inside. So if I come in from two weeks on the road and I've been plugged into public networks in the hotel, at a Starbucks, etc., you have no idea what's on there. You essentially need to disinfect this thing before it comes on to your network, on the edge. Do you think power over Ethernet will become a given on all wireless switches over the coming year?
Power over Ethernet has to become a no-brainer. And it's not just for wireless. It's for voice over IP too, unless you want to have legions of little power bricks stuck in people's outlets in their cubes. Will Wi-Fi be a significant driver of the market in this coming year?
This is the year we've heard from our clients that they're actually deploying Wi-Fi, so I would say it's going to be a driver for the market. People are expanding their Wi-Fi networks because deployment and security issues have been fixed. For the most part, the systems do what they're supposed to do and the security issues have been resolved. Every laptop that comes out now has some kind of wireless connection in it, so your end-points are all coming out that way. And I think that enterprises are at the point where if they can make the cost justification and the business justification, they'll role it out.