WDS with Dynamic and Static Links

What it is

WDS (Wireless Distribution System) lets access points connect to each other over the air and pass Layer 2 frames between them.

A normal access point talks to stations and uses a cable for the rest of the network. With WDS, two access points can also talk to each other, so a wired network can be extended without cable between them.

Each WDS link appears as an extra interface that you can add to a bridge, just like an Ethernet port.

How it works

A WDS link is a point to point connection between two access points, and it is either created automatically or defined by hand.

Why WDS is needed

Standard 802.11 frames between a station and an access point carry three addresses. That is not enough to pass a frame from a device behind one station with its own source address. WDS uses frames with four addresses, so the original source and destination are kept, and the link is transparent at Layer 2.

Because of this, bridging over a WDS link keeps the real MAC addresses of the devices on both sides.

The WDS mode setting

The setting that turns WDS on is wds-mode on the wireless interface:

/interface wireless set wlan1 mode=ap-bridge wds-mode=dynamic wds-default-bridge=bridge1

Dynamic WDS

With dynamic WDS, the access point accepts a WDS partner that it finds. When the partner is found, a new interface is created automatically, and it is marked as dynamic in the interface list. The link goes away when the partner goes away.

The wds-default-bridge setting names the bridge to which these interfaces are added at once. This saves manual work, and it keeps the network working when links come and go.

Dynamic WDS is quick to set up. It suits links between devices that you control. It accepts partners automatically, so it relies on the security profile to keep out strangers.

Static WDS

With static WDS, you create the WDS interface yourself and name the partner by MAC address:

/interface wireless wds add name=wds1 master-interface=wlan1 wds-address=00:11:22:33:44:55
/interface bridge port add bridge=bridge1 interface=wds1

• master-interface is the wireless interface that carries the link.
• wds-address is the MAC address of the partner's wireless interface.

The same entry must exist on the other side, naming the first device. You add the interface to a bridge yourself.

Conditions for a WDS link

For a WDS link to come up, these must match on both devices:

• The same SSID, unless wds-ignore-ssid is set to yes.
• The same band and frequency.
• Compatible security, because the link uses the security profile of the master interface.
• A wireless mode that allows WDS, such as ap-bridge.

There is also a wds-slave mode for the end that should follow an access point. It acts like an access point, but it scans for an access point with the same SSID and links to it, following its frequency.

Checking the result

/interface wireless wds print
/interface wireless registration-table print
/interface bridge port print

The first command lists the WDS interfaces, and a flag shows whether each is dynamic. The registration table shows the WDS partner as a connected peer. The bridge port list shows whether the WDS interface is in the bridge.

A word on capacity

All WDS links and all clients on the interface share the same radio and channel. A WDS link between two access points that also serve clients uses the same airtime as the clients. For a busy backhaul, use a separate radio for the link.