Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

Normally, when setting up a network of multiple Wi-Fi APs with a single ESSID, one would set the APs to be simple link-layer bridges between wireless and wired Ethernet.

Is this setup possible with the APs acting as routers instead of simple bridges?

share|improve this question

1 Answer 1

up vote 0 down vote accepted

The 802.11 standard does not require the "Distribution System" (DS)—that is, the backhaul connections between the APs—of an ESS to be an Ethernet, nor does it require the APs of an ESS to just bridge between wireless clients and the DS.

So, yes, you can set it up this way, but there are some things to keep in mind if you want to do this:

  • Routers block most service discovery protocols, so you'll often discover that devices on one AP can't see clients on another AP, since the APs effectively blocking the discovery protocols.
  • Most so-called "Wi-Fi routers" are not plain IP forwarding routers but NAT (technically what the IETF calls NAPT, or what Cisco calls PAT) gateways, and NAPT gateways by their nature end up blocking any connections from being originated from their "public" side to their "private" side.
  • Routers (and NAPT gateways) require different IP subnets on different interfaces. If you're using a bunch of the same brand of consumer Wi-Fi routers, they're all going to default to using the same subnet on the "private" side. If you connect the WAN port on one to the LAN port on another, you'll end up with the same subnet on both sides of the downstream device. You can avoid this by changing the subnet one of them uses, or by not daisy-chaining them in this way.
  • Some wireless clients might not refresh their DHCP leases when roaming from one AP to the other, so you might end up with IP address conflicts. Let's say you have two routers A and B, and both are using the same 192.168.2.x/24 subnet on their individual WLANs. If you also have two clients, X and Y, and X associates to A and gets 192.168.2.2 on A's network, and Y associates to B and gets 192.168.2.2 on B's network, and then let's say X moves away from A and toward B, and eventually roams to B. If X is well-behaved, its wireless card would have notified its host network stack of the roam event, and the host network stack's DHCP client would temporarily stop using the 192.168.2.2 address until it can reconfirm that the lease is valid on this potentially different network. But if X is poorly behaved, then X and Y will both be on B and both think they own 192.168.2.2.
share|improve this answer
    
If I have two routers A & B. Each of them creates 192.168.2.x subnet in its own wifi zone. Is there any error in this setup (sorry, I haven't tried yet). –  hongquan Sep 28 '12 at 2:48
    
@hongquan Is the WAN port of one plugged into the LAN port of the other? If not, then there's no real problem, although you'll have to hope your clients are smart about refreshing their DHCP leases any time they roam from one AP to the other. I'll add a note about that to my Answer. Also, regardless of how you have them connected, don't expect clients of A to be able to connect to clients of B or vice-versa. –  Spiff Sep 28 '12 at 2:53
    
Maybe not. My FON2100 router only has 1 Ethernet port. I think it is LAN port. Those routers are wired to my house LAN network via that Ethernet port. Thank you and sorry that I cannot vote your answer up because I do not have enough reputation to do. –  hongquan Sep 28 '12 at 3:09
    
@hongquan If you feel my answer correctly answers your question, you can click the checkmark outline to accept the answer. –  Spiff Sep 28 '12 at 3:24

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.