Arguably, there has been no hotter technology over the past decade than wireless LAN. The drivers for this have been the sheer explosion of wirelessly enabled devices combined with the continued evolution of the 802.11 standard. In fact, ever since the long awaited 802.11n standard was finally ratifie, wireless LAN adoption has been through the roof. Additionally, companies are aggressively trying to expand the footprint where wireless LAN is the primary networking technology. Historically, companies would deploy a network where wired was the primary network and then wireless was used to augment that and connect conference rooms and other shared areas. Today a few companies I have talked to, and this is more the exception than the norm, are attempting to use wireless as the primary, and only use cable where necessary. This can have huge cost implications when upgrading a network from just the savings on cables alone. This trend has primarily been limited to campus environments like schools and hospitals, but the industry is trending that way.
I do wonder though, is there another evolutionary step needed in wireless LAN before we can have reliable, ubiquitous connectivity (and I’d like to emphasize the word reliable)? Once we move to an all wireless network, the wireless now needs to support real time applications like VoIP and video, high bandwidth applications like VDI, and mission critical applications like patient telemetry systems in hospitals, on line testing at schools and point of sale applications in retail environments--all of which put demands on the wireless network that haven’t been there before. Additionally, once a company decides to not upgrade the wired network, it won’t be there to be used as a backup in case there's a problem with wireless.
As popular as microcell has been, and microcell is the majority of wireless today, it does have some limitations that could cause many critical applications to not perform consistently. The main issue with microcell, and many people don't understand this, is that it's a shared technology. The wired equivalent of this is a hub, not a switch. There’s not a company out there of significant size that would consider using hubs for its corporate network, but that’s what the user experience in a microcell environment would be like in a large scale, multimedia environment.
The biggest issue that the shared nature of microcell has is that it delivers a highly inconsistent user experience, particularly in high volume areas. Go to any location where the user density changes widely during the day--a trade show for example, which all of us go to several times a year. In the mornings, prior to the keynote you can be working on the wireless network and everything is fine. The keynote starts, everyone jumps on the network and the performance goes to crap. This isn't a signal strength issue, everyone can see the wireless LAN on their laptop configuration application, but the quality is poor because the network is saturated. In fact, in many cases you’re lucky to even get on the network. This happens in hotels, coffee shops, company cafeterias, etc. To get around this, I've given up on public wireless LAN totally. If I can get on to a dedicated wireless LAN network for press and analysts I’ll generally use it, otherwise I defer to my Verizon EVDO card (which also has its problems because of saturation of its cellular network, but that’s another blog). Corporate workers won’t stand for a network that provides lightning fast access at some points in the day and then cripplingly poor access at other times. This is the nature of a shared network and the primary reason wired hubs have disappeared from the corporate workplace despite their low cost.
So if microcell can't scale, what's the answer? I think the virtualized wireless from Meru actually has a great chance of gaining ground. Anyone that knows me has heard me say that significant share change only happens at points of market transition. Cisco gained in voice when it transitioned to VoIP, Microsoft became the dominant OS vendor when we migrated to PCs and I think Meru has the opportunity to gain share as organizations transition to ubiquitous wireless.
Meru's "virtualized wireless" is the equivalent of a network switch, not a hub. It allows the network manger to dedicate bandwidth out to each user, meaning if Michael Scott sits in his office and watches YouTube all day it doesn’t hamper Dwight Schrute’s ability to sell paper because his on-line ordering system is performing properly.
I have to admit that when Meru first briefed me on what they were doing, I was highly skeptical, but after talking with a few of their customers that have deployed Meru in large scale environments, I can say that it does work and does replicate the performance of a wired switch. Wireless mobility with wired switch reliability--this can be used to create an enterprise that wants to go all wireless.
In fairness, there are ways of building a large microcell network that does improve the performance, by overbuilding capacity--but it does have its limits. Similarly, one could have used hubs to build a wired network by overbuilding capacity and giving each worker his or her own hub, but there are limits as well. A shared network--whether it is a cable head end, wired hub, cellular network or microcell based wireless LAN--will all perform poorly when heavily saturated.
As far as I know, Meru is the only vendor that currently offers a "switch"-like wireless LAN solution, so they’ve got a head start. Obviously, it takes more than just technology to create a market leader, but the drive to ubiquitous wireless gives Meru a better than punchers chance at gaining ground on the big boys.
