On Sat, Aug 16, 2008 at 5:21 PM, Lee Sharp <leesharp at hal dash pc dot org> wrote:
> I have never had a problem with a user calling in to connect in the lobby
> to a different AP than from the room. :)
> We have a standard deployment. We start the IP addressing at x.x.x.20 and
> go up. Then we NAT 10120 to .20 and so on. The SSIDs are "AON @
> (HotelBrand) 1" and so on. That way if a customer say they are having
> trouble at "AON @ HolidayExp 5" I know to look at 192.168.52.24, or the
> firewall IP port 10124 to find them. There is no question which is easier.
Yes indeed, the only "problem" is people could see a lot of AP's appearing
instead of one when you use the same ssid.
> As to the range... What AP did you test? Some are junk. Also, did you
> use any kind of directional antenna? Note that an 18db Omni can give much
> grater range at the expense of vertical range, and may help if the floors
> really are a barrier. But I can't think of a single hotel we have (and we
> have some old crappy hotels) where we don't get 2 or three floors per AP.
Well, because you were not the first one telling the ap's should do for two
floors, I decided to drive over to that location today and do the tests
again with the AP's we are going to use. The wall mounts are already in
place so setting up an accurate testing environment was pretty easy now. I
use openwrt driven linksys wrt54gl's (setup as plain AP, so with disabled
dhcp, etc) power set to 100mW. I chose one of the four shafts and placed an
AP on ground floor, second floor and fourth floor. Did some testing on all
floors and noticed pretty good signals in network stumbler.
Also floors 1 and 3 were served with a weaker, though decent signal. I could
almost walk to the next shaft (were the next AP will be) before actually
losing the signal.
So I start with 3 AP's per shaft and see what it brings us. All cabling is
there already so putting an extra AP if someone complains is only five
Strange, my initial testing gave me a very poor signal on the floor above.
The only difference was that I did the testing with a 100mW Senao AP placed
in a room (needed power) and not in a shaft in the corridor.
Thanks for your input,