802.11ac: Back on the Enterprise Radar

Comment | Print | RSS

• Login to Rate

In that article, I described several shortcomings of the newest WiFi protocol and why it wasn't a particularly useful upgrade for companies that support a lot of wireless users. Much of the reasoning was due to the fact that, to achieve much faster speeds than what 802.11n can provide, you have to sacrifice user capacity to achieve higher throughput rates. With the continued proliferation of BYOD, user capacity is increasing, which negates any possibility of configuring a network for more speed.

The primary problem behind the capacity vs. speed issue is how 802.11ac achieves faster speeds than 802.11n. Speed gains are found primarily through channel bonding techniques. Channel bonding links multiple wireless channels together, so that a single wireless communication leverages two, four, and even eight times the amount of 5GHz frequencies as a standard 20MHz 802.11n transmission. But bonding channels together limits the number of wireless devices that can talk at one time without overlapping frequencies. If there are very few WiFi users, channel bonding works great. But you quickly run into capacity issues as the number of simultaneous users increases.

In the US, the current 5GHz spectrum allocated for WiFi consists of 22 sets of 20MHz channels. Quite frankly, it's simply not enough space to create 160MHz or even 80MHz bonded channels on wireless networks that have a moderate or high number of simultaneous users. Trying to get ahead of the obvious problem, the FCC has announced its intention to allocate more frequency space in the 5GHz range, raising the number of possible 20MHz channels to 35. This will help alleviate much of the overlapping claustrophobia and will give wireless admins the opportunity to bond channels -- even in heavily used wireless networks.

Even with this news, you might not want to run out and buy new 802.11ac hardware just yet. For one thing, the adoption of 802.11ac end devices is likely to be slow. For another thing, the FCC announced only that it is looking to allocate more bandwidth. This hasn't been done yet. Who knows if and when this might occur?

But the point is that 802.11ac cannot be completely written off in the enterprise the way we once thought. If you need to put in new installations or upgrade an aging system in the next 12 to 24 months, the cost differences between implementing 802.11n and an 802.11ac wireless network are certainly worth a look.