Here at the Mono Summit we had not-so-great Internet access. One of the things we tried to remedy this was to plug in our own routers. This was itself an adventure as the network did not have any DHCP and wasn’t very friendly to us. We used etherape, wireshark, and a switch to figure out the settings on the not-so-reliable but very attractive Cisco Aironet 1130AG and put these settings on a Linksys WRT54G. Aaron Bockover got that hack working eventually. This alone wouldn’t improve the situation much so we added a couple more routers to the network. Ideally you would use a simple access point to do this so that everyone would be on the same subnet and have fewer NAT routers to go through.
Real APs are hard to come by these days but you can use a wireless router as an AP by plugging in one of it’s LAN (not WAN) ports into a LAN port of your router. There are a few more things you must do for this to work properly:
Plug one of the APs LAN (not WAN) ports into a LAN port on the router (may require a crossover cable)
Disable the DHCP server on the AP
Change the internal IP address of the AP to something different than the router (but on the same subnet)
Without these you will have conflicts that will prevent anyone from getting out to the Internet. If you were at the conference and wondered why we didn’t have Internet most of Thursday (and some of Wednesday) this is why.
In theory it should be possible to use the same SSID for each of the routers and clients will roam between them but we didn’t want to push our luck. Also I don’t know if roaming would distribute the load evenly whereas we hoped that folks would choose an AP more or less randomly. This is not the same thing as WDS.