But before we forge ahead into the examples, there’s some prep work you’ll need to do to give yourself the best chance of having your WDS setup work the first time. Here are three mandatory and two optional steps to take before putting your APs in bridging mode :
Check that your wireless client can associate and pass data through each AP
Assign a static IP address to each AP
This is good practice in general when dealing with the gear that runs your network. But it’s especially important for APs in WDS networks, since you’ll know where to look for each AP and have one less variable to consider when trying to debug a broken connection. Make sure that you assign the static IPs outside the range of your LAN’s DHCP server or you’ll risk getting a duplicate IP leased at some point... usually when it’s least convenient !
Set all APs to the same (clear) channel
Since all APs in a WDS network need to communicate with each other wirelessly, they need to be on the same channel. For 802.11b and g WLANs, I recommend you use Channel 1, 6, or 11. Whichever channel you choose, make sure it’s not in use by neighboring WLANs, or at least not one right close by. See our When Wireless LANs Collide ! ProblemSolver if you have trouble getting a clear channel.
[Optional] Set each AP to a different SSID
WDS APs know each other by MAC address and could care less what their SSID is set to. On the other hand, wireless clients associate by SSID. Technically, each AP in a WDS network is part of the same Extended Service Set (ESS) and should therefore have the same SSID.
But the roaming algorithms incorporated into most wireless clients don’t "aggressively" roam and tend to stay associated with an AP long after they should, resulting in poor performance. This can be especially frustrating when you’ve gone to the trouble and expense of adding repeaters to your WLAN, and your notebook refuses to use them !
By assigning different SSIDs to your WDS APs, you’ll first have the advantage of being able to see each one, even if you’re using WinXP’s built-in "Zero Config" utility, which doesn’t show multiple APs with the same SSID. You’ll also be able to easily force your client to connect to the closest AP without having to remember its MAC address.
[Optional] Assign a static IP to your wireless clients
I’ve found that it sometimes takes awhile to lease a new IP after associating with an AP. Assigning static IP information to your wireless clients (don’t forget to include gateway and DNS info) gives you one less thing to go wrong when switching association among APs. It also works around the problem that some products have (or at least had !) with properly passing DHCP messages to bridged clients.
In addition to the above, you also need to carefully consider placement of your WDS APs. As with any other wireless LAN equipment, the speed of a WDS link depends primarily on signal strength. Since each WDS "hop" already cuts available throughput approximately in half, you don’t want to further reduce your link speed by spacing your WDS APs too far apart.
You’ll need to experiment to get an combination of range and performance that’s acceptable to you, but don’t expect good link speed if you try to place your repeater near the limit of your current wireless range ! A good compromise is to place your repeater in an area where the link speed (as indicated by your client utility) is 5.5Mbps or better for 11b equipment and 24Mbps or better for 11a or 11g gear, i.e. about half the max transmit rate.
With the preliminaries out of the way, we just need to gather the MAC address information we need and we’ll be on our way !