Channel Planning with 40 MHz ChannelsBy CWNP On 09/13/2007 - 4 Comments
Let's not even consider 40 MHz channels in the 2.4 GHz ISM band. There can only be two non-overlapping HT MIMO channels in the 2.4 GHz band. Considering the mess of protection mechanisms, competing 20 MHz channels, and the dozens of interfering non-WiFi devices in that band, 802.11n in 2.4 GHz is a lost cause from day 1. Instead, let's consider the only reasonable (IMHO) enterprise-class 802.11n deployment methodology: 5 GHz.
40 MHz channels require dual adjacent 20 MHz channels for operation. For example, channels 36 and 40 might be used together as a 40 MHz channel or perhaps channels 44 and 48. When site surveying, we are doing channel planning for between 8 (UNII-1/UNII-2) and 23 (UNII-1 (4), UNII-2 (4), UNII-2E (11), and UNII-3 (4)) useable channels. I'm leaving out the channel 165 5 GHz ISM channel as a consideration here. It's hoped that most 802.11n equipment (especially enterprise-class equipment) will support all four of these UNII bands. With 23 (or 24) available individual channels, 40 MHz channel planning should be a reasonable task.
The fun will come when there are existing 20 MHz channels in use by neighboring devices out of your control. Planning 40 MHz deployments in the presence of 20 MHz channels that are auto-selected by a neighbor's WLAN controller's RF Management feature will be bonus fun. First, their 20 MHz systems may cause your system to change to HT_MF mode, which will cause use of protection mechanisms. Second, their channels may be configured such that you (or your system's automated features) won't be able to find an interference-free channel pair. In fact, to design a good 40 MHz channel reuse pattern, you'll need no fewer than three 40 MHz channels (each consisting of a pair of 20 MHz channels) per WLAN infrastructure.
Can you imagine what the mess might look like if you have three WLAN controllers in a multi-tenant building, each using 5 GHz 40 MHz deployments with automatic RF management features enabled? At a very minimum, where three U.S.-based organizations worked together on their design (which will never, ever happen), there would be 18 channels in use - three pair per organization. That's not considering the individual 20 MHz channels that might also be in use here and there. Yikes.
This makes me think that companies championing the single-channel architecture MAY have a reasonable advantage in deploying somewhat-interference-free networks in 5 GHz (40 MHz) 802.11n networks, and even more of an advantage in the 2.4 GHz band (since two non-overlapping 40 MHz channels do not have to be used). An 802.11n feature that SEEMS to give more of an advantage to the single-channel architecture than it does to the traditional channel-reuse architecture is Transmit Beamforming (TxBF). TxBF cuts down on the bulk of co-channel interference by directionally focusing transmissions. This is good for traditional architectures, but REALLY good for the single-channel architecture. There's also the possible drawback that TxBF might cause clients to be sticky.
Organizations who already have a significant investment in 802.11a/g networks should consider the following before an upgrade:
1) As many clients as possible should be upgraded to 802.11n to get additional throughput benefits
2) Some 802.11n APs may need more power than is provided by 802.3af PoE
3) A new site survey may have to be performed
4) New cable pulls may be required
In summary, get ready. Job security is on its way, and it's called 802.11n.
Blog Disclaimer: The opinions expressed within these blog posts are solely the author’s and do not reflect the opinions and beliefs of the Certitrek, CWNP or its affiliates.