Wider Channels?

K7DXS,   SNR becomes significant with the longer distant links we generally  have on an AREDN mesh.  Because the transmitter has to split the power across the 40Mhz channel width (it becomes 2 x 20Mhz channels in 802.11n), the power is cut in half -3dB on each channel.    We're not seeing 20Mhz links that are able to achieve the max 144Mbps MCS15 max rate in 802.11n, unless these are short links, like less than a ~mile, depending on particular device and antenna gain.   I've tested 10Mhz compared to 20Mhz channels and find that better throughput is achieved in 10Mhz for the long distant links in the area--not even using 20Mhz.

Also what happens is the total noise doubles when doubling the channel width, while the transmit power stays the same.    We end up with overall less SNR to double the channel width.  

802.11ac, which can do up to 160Mhz, does not offer increased benefit to use even greater channel widths for the same reasons--but does offer some other benefits.    In the closed confirms of one's home, there's sufficient SNR to get extremely high throughput with the additional bandwidth all the way up to 160Mhz channels.   

40Mhz has been considered, there's a ticket in the system for this.  We had concluded that we'd not really get any benefit to spend our time on this verses working on something else.

Joe AE6XE