Toward the Wireless Office: Page 2

(Page 2 of 2)

Overcoming infrastructure

While most office buildings today are built with Cat-5 wiring already installed and most large companies have long since pulled all or most of the Ethernet cabling they need, in greenfield scenarios or where you’re extending a network, wireless “may” be cheaper, Mathias says.

The vendors say it always is, or at least can be. The cost inherent in wired networks, they point out, is not the cabling itself or other infrastructure, it’s the labor involved.

“The cost of wiring up a new building versus putting in a pervasive wireless network is substantial,” says Meru’s Troyer.

Finally, there are applications that are only possible with wireless. RFID tracking that rides on top of a Wi-Fi network can be used not just to track assets, but also people. A hospital in Thailand, for example, is using RFID tracking on its pervasive Wi-Fi network to locate the nearest doctor in an emergency.

Locationing technology could also be used to refine facility security systems, sending out alerts when non-authorized personnel enter a restricted area, for example. Or it could be used for network security to ensure visitors to a facility only have access to the network in the room or area they’ve been assigned.

None of these incentives would count for anything, of course, if the technology had not improved to the point that it was feasible in the first place to think of using it as a primary network medium.

Safety first

The emergence of WPA (wired protected access) and bullet-proof WPA2 encryption systems as replacements for the compromised WEP (wired equivalent protocol) of earlier Wi-Fi generations, has “solved the security problem,” Mathias says.

Wireless security is now as good as wired networks “or probably better from a physical connection perspective,” Troyer argues. “After all, anybody can plug into a wired port and get access to your network, but without the encryption mechanism for WPA2, there is no access [on a wireless network].”

The emergence of central wireless controller products from several vendors in the last three or four years has also made it easier to manage networks and client devices on them. This was a key requirement from the perspective of IT managers who were accustomed to similar manageability on wired networks. And network management is probably even more important on wireless networks.

“Wi-Fi is a shared medium,” Troyer notes. “Unless you’re able to control or coordinate the way clients have, and applications get, access to that shared medium, you’re always going to have unreliability, or the perception of unreliability. And as a result companies won’t use [Wi-Fi] pervasively.”

Enhancements to quality of service (QoS) mechanisms with the ratification of 802.11e helped make wireless viable as a medium for latency-sensitive applications, such as voice and video.

Mathias studied one Meru all-wireless office implementation with 123 VoWi-Fi phones. The wireless network delivered “very good toll quality” connections on all of them, he says. 

He calls the emergence of cellular/Wi-Fi dual-mode phones “a key development.” Many employees now carry cell phones around the office, which they can use to receive internal calls but only at a cost--cellular minutes--and only if in-building reception is good enough.

Dual-mode phones would let them be fully mobile within the office without having to incur cellular charges, and without having to carry two cordless phones, one VoWi-Fi, one cellular. Plus, there are the mobility-enhancing benefits of being able to hand off between cellular and Wi-Fi networks.

But, the single most important advance paving the way for the wireless office is 802.11n, with its ability to deliver bandwidth and throughput comparable or superior to fast Ethernet.

The need for speed

“11n changes the game,” Troyer says. Tennefoss agrees. “[It] was really the power boost that wireless needed to compete with copper,” he says.

The new Wi-Fi standard still isn’t as fast as Gigabit Ethernet, the standard for enterprise wired networks. But 11n delivers enough bandwidth--over 100 Mbps--that “almost any application is viable,” Troyer says.

Mathias adds, “It could be argued that you never really have enough [bandwidth on a network]. And there is no network without a degree of latency, and no network that never has capacity issues at least some of the time.” Nevertheless, 11n is sufficient for most office applications.

Nobody, however, is seriously suggesting the wireless office means no wires anywhere.

Even Mathias says, “I would never recommend going entirely wireless. If you’re talking about a server farm or interconnection of stationery devices, wire is a very good strategy. I never thought a wireless help desk or contact center made much sense either.”

More and more vendors, though, are touting the benefits of wireless mesh for distribution rather than running Ethernet cabling to APs. It delivers additional cost benefits by avoiding those cable runs. And in environments where cabling is impossible--Motorola’s Hajela cites the company’s all-wireless deployment at the 12th century Kilkenny Castle in Ireland--it can be the only way.

Troyer says, “Companies today are using Meru at the access layer for all wireless applications, but over time most are planning to migrate to using it at the distribution layer as well.”

Proceed with caution

Improvements in Wi-Fi don’t mean that implementing an all-wireless, or nearly all-wireless, strategy is a slam dunk, however. Network planning becomes much more critical. And vendors warn of other challenges--which of course, their products and architectures are uniquely able to overcome.

Tennefoss notes that first-generation 11n products, unlike Aruba’s, draw too much power to be able to use power over Ethernet (PoE), which could cause problems in some deployments. And older-design wireless network controllers may not be able to handle the traffic volumes inherent in an all-wireless approach.

Troyer argues that network controllers that let client devices decide which access point they’ll connect through--most, besides Meru’s--will make it difficult to do the kind of load balancing needed in a high-volume wireless network.

Nor do the lures of an all-wireless enterprise or the undoubted technological improvements mean the Fortune 1000 will embrace the concept en masse. There are still hold-backs--much of it just inertia, Mathias says.

IT professionals have lots of questions, he says. They’re still unsure about security and manageability.

“All those questions have answers, but they’re not well established yet.”

Another hold-back, Troyer says, is that cellular carriers have not promoted the idea of dual-mode cellular/Wi-Fi phone service--perhaps out of concern for lost cellular minutes.

That is changing, however. T-Mobile launched its Wi-Fi-based Hotspot at Home service in June. And Meru has worked with one regional carrier in the Northwest that is introducing an enterprise dual-mode service.

The wireless office: it may be an idea whose time has come--or it might take a few more years to mature. But, we’re guessing the inevitable convergence of wireless in the office and wireless outside the office means total wirelessness will become a reality for many companies sooner rather than later.

This article was first published on Wi-Fi Planet.


Page 2 of 2

Previous Page
1 2
 





Comment and Contribute

 


(Maximum characters: 1200). You have characters left.