Dynamic Frequency Selection Part 2: Challenges With 802.11ac

DFS is critical for protecting mission-critical systems, but meeting DFS requirements in 5GHz wireless, especially 802.11ac, is difficult.

Jennifer Jabbusch

October 15, 2013

2 Min Read
Network Computing logo

In my last article, I discussed Dynamic Frequency Selection (DFS) and its critical role in protecting mission-critical systems sensitive to RF interference from Wi-Fi and similar devices. In the second part of this three-part series, I examine the unique challenges we face in trying to accommodate the DFS requirements in 5GHz wireless -- especially 802.11ac with its wide-bandwidth channels that can be four to eight times wider than previous 5GHz Wi-Fi.

DFS requirements demand 5GHz Wi-Fi (such as 802.11ac) be designed with out-of-the-box features to identify radar systems and then take action to avoid using channels that interfere with any identified radar systems.

It sounds easy, right? Well, it’s not. The first challenge is accurately identifying the radar. Remember, the radar patterns are just Layer 1 RF noise; not a data packet with a header detailing who sent it. You could compare this pattern matching to how an ornithologist identifies bird calls. It takes a trained ear, focused attention, and an environment with a tolerable ambient noise level.

Similarly, our 5GHz Wi-Fi systems have to know the patterns of radar RF, have enough listening (no-transmit) air time to be able to hear the radar, and the RF noise level must be low enough that the 802.11 wireless device can clearly identify the RF as being radar. Compounding these challenges is the fact that radar signatures change as new technology is introduced, and military radar behaviors are unique but details are classified, making it even more difficult to pinpoint those RF patterns.

[Read about the basics of the radio frequencies used in wireless networks and the phenomenon of attenuation in "Wireless For Beginners: RF and Waves."]

The next challenge for Wi-Fi is the avoidance requirement of DFS, in which access points have a limited time to cease all transmissions in the interfering channel and move clients to a new channel. DFS allots 10 seconds for the entire process, but only with an aggregate of 260 milliseconds worth of control data, which is required for the APs to issue instructions to clients for ceasing communications and moving to another specified channel.

If you’ve worked with wireless for any length of time, you’ll already understand the challenges this presents with slower data rates, slower clients or poor wireless connections. Clients can get dropped or lost, and even in the best circumstances, the client has to disassociate and re-associate with the AP, meaning session-based applications (such as VoIP) are dropped. After the non-occupancy period and a re-scan period, the AP may choose to move clients back to the original channel, repeating the disassociation yet again.

The complexity of operating 802.11ac wireless within the DFS-designated bands is why most wireless vendors have recommended customers deploy their 5GHz wireless without using the DFS channels. In my next post, I will discuss the complications of restricting 5GHz channels in 802.11ac.

About the Author(s)

Jennifer Jabbusch

null

Jennifer Minella is VP of Engineering and consulting CISO at Carolina Advanced Digital, and an author, speaker and consultant for infrastructure security for government, education and Fortune 100 and 500 corporations.

SUBSCRIBE TO OUR NEWSLETTER
Stay informed! Sign up to get expert advice and insight delivered direct to your inbox
More Insights