On Sun, 8 Apr 2018 00:32:05 -0000 (UTC) Dan Purgert <d...@djph.net> wrote:
> David Wright wrote: > > On Sat 07 Apr 2018 at 20:17:56 (-0000), Dan Purgert wrote: > >> David Wright wrote: > >> > On Fri 06 Apr 2018 at 16:26:47 (-0000), Dan Purgert wrote: > >> >> > >> >> It's a nuance in the semantics of what it means to "repeat" wifi. > >> >> Suffice to say, in order to "repeat" wifi, you have one radio splitting > >> >> its time between pretending to be an AP for a client device, and > >> >> pretending to be a client device to the upstream AP. > >> > > >> > Then I'm not sure why you wrote "Good deal". I'd be wanting the > >> > wireless connectivity described above as not needed, though obviously > >> > on a separate band/channel. Were you implying that that would kill > >> > throughput for everything too? > >> > >> If he's using the buffalo device to "repeat" the wifi signal (which he > >> isn't), then yes the throughput would tank. > > > > OK, I'll just assume you don't know. Anybody else actually doing this > > (separate band (like 2/5 GHz) or channel (like channel 1/6/11) for the > > backhaul (inter-router) link)? > > If you have a device repeating a WiFi signal, it *will* use the same > channel as the upstream AP. It *cannot* use a different channel. > > In the event you have a dual-band AP, and the following conditions are > true > > - 5GHz uplink > - 2.4 GHz for clients > > Then you are not "repeating" the WiFi signal to the downstream client > devices (and the throughput losses I mentioned would not come into > play). There are also apparently some units (even consumer grade ones), that have two diferent radios both on (different) 5 GHz bands, so one could use one for client access and one for uplink (although I have no experience with this): https://www.linksys.com/us/r/resource-center/basics/multiple-wifi-bands-difference/ Celejar
