I’ve recently had a number of interesting conversations with end-users and technology suppliers alike about 802.11ac channel plans. Granted, 802.11ac runs in the relatively-underutilized 5 GHz. bands, but 80 MHz. channels may still be difficult to manage in some cases (to say nothing of the complexities inherent in 160-MHz. channels). So, what are the practical issues and best practices with respect to applying these larger channels in production environments? And we’ve had a couple of questions about why .11ac is 5-GHz. only, so could you comment on that as well?
The easiest part of that question to answer is why 802.11ac is 5 GHz only. That's been a common question, and I first answered it as a blog response to Tom Carpenter at last fall's Wireless Field Day: http://blogs.aerohive.com/blog/the-wi.... Briefly, you need lots of spectrum to get high speeds, and the 2.4 GHz band is so crowded and has such a high noise floor that it can't take full advantage of the features of 11ac.
An important note about channel width in 11ac is that you don't have to pick a single channel width for all clients. 11ac lets you switch the channel width on a per-frame basis, so you can transmit at 20 MHz bandwidth to client 1, 40 MHz bandwidth to client 2, and 80 MHz to client 3 (provided the whole channel is free at the width that you want to use). Devices can even negotiate the channel width. One of the ways that this helps is that if you space out the channels, you can "burst" to the highest capacity available at the time. Here's a part of a figure from my 11ac book. When the entire 80 MHz in the figure is idle, a transmitter can use the whole 80 MHz. However, if part of the channel is busy, you can fall back to a narrower bandwidth.
Someone asked earlier if 2.4 couldn't take advantage of at least some of the technologies and benefits inherent in .11ac. After all, 2.4 isn't necessarily crowded everywhere nor all the time. Is the IEEE essentially abandoning 2.4 for the future? And what happens when 5 becomes crowded?
We currently have five 80 MHz channels available. If the rules proposed by the FCC at the end of Julius Genachowski's chairmanship go through, we'll get four more. More importantly, we'll go from one 160 MHz channel to four. (I should point out that the 802.11ac spec allows you to build a "160 MHz" channel either by taking 160 MHz of continuous spectrum or taking two 80 MHz channels and treating them together.)