802.11n Places Increased Demands on Wireless Networks

Tech Analysis: Wireless administrators should start evaluating their infrastructure to gauge whether their systems can handle the new technology.

As the first enterprise-class 802.11n wireless products trickle into the marketplace, wireless administrators should take a hard look at their existing implementations to assess how the systems will cope with the new technology.

In April, Meru Networks announced the first enterprise-class wireless products that support the 802.11n draft standard. Meru unveiled its AP300 family of access points, which includes units with single 802.11n radios and others with dual radio configurations that can simultaneously advertise both 802.11n and legacy 802.11a or 802.11b/g networks.

Merus 802.11n products are currently in beta but should be generally available in the third quarter.

Merus single-channel architecture gives the company a leg up in 802.11n adoption, as it can offer a full 40MHz channel in the 2.4GHz band. Other solutions need separate channels for each access point and therefore cannot provide 40MHz channels in the 2.4GHz band. However, we expect to see numerous announcements from competing wireless vendors in the coming months that will parrot the latest 802.11n draft standards recommendations limiting 802.11n to 20MHz channels in the 2.4GHz band by default, leaving 40MHz usage for the 5GHz band instead.

Although more 40MHz channels are possible in the 5GHz band, administrators will still need to evaluate the merits of the wider channel in this band as well. We expect that most companies managing dense WLANs (wireless LANs) will opt to forgo wide channels altogether, especially when first introducing 802.11n to the network.

With these products and announcements looming, we advise wireless administrators to start evaluating their WLANs to gauge whether they can stand up to the increased demands the new technology will place on the infrastructure. But administrators should make sure to conduct assessments on both the wireless and the wired sides of the network.

802.11n promises dramatically increased throughput over the air, as weve tested individual access points able to push more than 130M bps of traffic. Companies looking to squeeze every last drop of performance out of an 802.11n wireless network will therefore need to look at the costs of deploying a Gigabit Ethernet switch infrastructure behind the access points—a potentially significant cost increase that most companies have likely not yet undertaken.

More critically, however, administrators with centrally controlled wireless deployments need to look at whether back-end controllers can be upgraded to support 802.11n and can handle the increased traffic load that will result from the new technology. Although most wireless controllers now support Layer 3 connectivity to access points, these controllers may still be in the data plane—providing encryption or VLAN (virtual LAN) services, for example—and present the possibility of a bottleneck.

Meru recognized this shortcoming in its existing products. So, along with announcing the AP300 line, the company unveiled a new tiered architecture that separates the data plane from the management plane. This allows customers to maintain a master controller for management while deploying new controllers to handle data services.

On the wireless side of the network, administrators will need to take a hard look at channel management and usage. They not only will need to decide when and where to use 20MHz versus 40MHz channels but also will need to evaluate how interference levels will change around the network and how the new technology will interplay—or overlap—with existing neighboring deployments. And, of course, administrators will need to reconsider the role of the 5GHz band to help resolve some of these concerns.

Administrators should also consider the impact of the new technology on access point density. With better antenna diversity and MIMO (multiple input, multiple output) technology, 802.11n offers much greater range than was possible with previous wireless products, so cell sizes could certainly be larger and require fewer access points.

However, 802.11n adoption on the client side will not happen in lock step with uptake on the infrastructure side. With Intels "Santa Rosa" chip set imminent, 802.11n laptop adapters will soon be more abundant, but 802.11 b/g adapters will have a home in the network for years to come.

For instance, we have yet to see any mobile devices or voice-over-Wi-Fi phones that support the new draft standard. 802.11n-enhanced WLANs will need to be deployed to ensure that legacy client products will get coverage equivalent to what they had in previous networks, which means administrators must plan to use a deployment pattern similar to that of their existing 802.11b/g network or layer 802.11n cells over the existing WLAN. ´

Senior Technical Analyst Andrew Garcia can be reached at andrew_garcia@ziffdavis.com.


Check out eWEEK.coms for the latest news, reviews and analysis on mobile and wireless computing.