Wireless LANs: Back to the future, sort of

A start up with a 'revisionist' approach to WLANS

It has been interesting, over the 16 years I've worked in the wireless field, to watch the barriers to wireless fall across all applications. I remember the days, for example, of the ARDIS and RAM Mobile Data networks, when we were lucky to get 2.4Kbit/sec. in a small number of locations around the country.

Today's near-megabit wireless wide-area network (WWAN) throughput was discussed as a theoretical possibility. Indeed, it seemed likely to me, given appropriate advances in radio architecture and very large-scale integration (VLSI), but everyone was truly happy just to have text-based e-mail in a five-pound package -- a palmtop computer and a wireless modem "brick." Wireless LANs were terrific for the warehouse, but were too slow and too expensive for general use. Regardless, mobile computers back then were hardly mobile by modern standards, so demand was low anyway.

But all of the key elements for the long-term success of wireless have matured and converged to the point when the WLAN can be the default access for almost anyone with a mobile computer or similar mobile device. But I must confess that I am still a little surprised at the continuing debate over WLAN architecture.

Early WLANs, based on what we today call a traditional or robust architecture (OK, the politically incorrect term "fat" is also used a lot here) clearly couldn't scale. No one would be able to manage such a network without a lot of centralized capability. And, regardless, someone got the bright idea to move common intelligence out of the access point and into a centralized switch. (The credit goes to Symbol Technologies Inc., now part of Motorola Inc.) This model further evolved into today's thin access point/centralized-controller configurations, which dominate enterprise-class WLAN thinking. It's practically a given that this is the right way to go, with all of the industry leaders producing products that fit this architecture.

So, along comes start-up Aerohive Networks Inc. with a revisionist approach of sorts. Aerohive has invented something it calls the Cooperative Control Wireless LAN Architecture, or CCWA. The company provides a centralized management appliance, which I view as essential in any enterprise WLAN network. But the control element of the network, which has lived in the controller for the past few years, goes back into the access point, which Aerohive calls a HiveAP. And along with distributed control comes fully distributed data, allowing the HiveAP to route data directly to a desired destination without sending it first through a controller.

Aerohive thus makes a key claim -- that any possible bottlenecks related to congestion at the controller are eliminated and that the distributed control technique it uses accomplishes all of the functions of a controller but more efficiently. And there's more: Each HiveAP can also be part of a wireless mesh, for both configuration flexibility and for redundancy. Should a wired connection to a HiveAP fail, the access point will rejoin the network as a mesh node. And, in theory, this approach should have economic advantages in that no controller is required, and venues with small numbers of access points won't require the small controller frequently required in modern controller-based architectures. All WLAN vendors price and discount aggressively, but eliminating the controller should result in less cost, no matter what.

And I must confess that I no longer know what a "thin" access point is. In theory, the thinness of modern access points should be leading to significantly lower prices, but we haven't seen this yet. And even with the pricing of enterprise-class WLANs today being subject to massive competition and correspondingly significant discounts, Aerohive is most certainly worth a look. The company has rekindled the WLAN architecture debate, and I'm looking forward to testing its products as soon as I can -- and to hear what their competitors have to say.

And, if nothing else, the fact that Aerohive is a venture-backed start-up is a powerful indication of the opportunity and vibrancy that will keep WLANs innovative, fascinating and valuable for many years to come.

Craig J. Mathias is a principal at Farpoint Group, an advisory firm specializing in wireless networking and mobile computing. He can be reached at craig@farpointgroup.com

Copyright © 2007 IDG Communications, Inc.

7 inconvenient truths about the hybrid work trend
Shop Tech Products at Amazon