Basically, the extender has to be smart enough not to try to connect to itself or any of its downlinks (so it needs to have some idea about the topology) for this to work.Īs the wireless infrastructure set up in this way is a pain in the arm to manage - especially when you are changing the PSKs of your wireless networks ( it looks like you are using PSKs, so you must be changing them frequently, right?), you seriously should be looking into vendor-specific WDS systems where you would not need to touch every single device in the right order to get your keys changed without interrupting connectivity for your management data path. Whether or not your extenders would have any issues with setting the extended network's SSID to the value of the uplink's SSID would be implementation-dependent.

The address conflicts reported by the host may be arising from either a true address conflict (something is claiming the IP address the host wants to use and responding to ARP who-has requests) or a network loop where the ARP request from the host would be reflected back to it, answered and the answer reflected back as well - which would look like another host is claiming the address, while in fact it would just be a looped back transmission. Obviously, round-trip times would increase for the connected clients as each hop would add to latency. Where this is not possible and if your requirements permit it, you can do a radio-to-radio forwarding by the means of 'extenders'.Īs each extender creates an own collision domain, you could daisy-chain them indefinitely - pretty much as you could do with wired switches.

Under performance and reliability considerations, an access point should be connected upstream by a wired link. The difference to a 'classical' access point configuration would be that their upstream link is wireless - similar (but not necessarily identical) to the Universal Repeater Mode. Extenders typically are nothing but Wireless Access Points - with two radios if they are any good.

Setting up wifi extenders