I am having difficulty in understanding the difference between the routing and path selection.

I am working on IEEE 802.11s-based mesh networks and reading about the default routing protocol HWMP. In its wikipedia page, it is written that they use "path selection" instead of "routing" term because it uses MAC addresses, but not IP addresses. It does not make sense it is called path selection just because of the use of MAC addresses.

Then, I checked the routing page in wikipedia. They define routing as "the process of selecting a path for traffic in a network". From the perspectives of these definitions, they look the same.

Is it possible for anyone to explain the difference briefly and put an end to my confusion?

2 Answers 2


Routing is considered a layer-3 function, using layer-3 (IPv4, IPv6, IPX, AppleTalk, etc.) addresses. Routing involves a path selection based on the layer-3 destination address. A router uses a routing table to determine the path used to forward the layer-3 packets, and it drops any packets for which it has no path.

Bridging/switching is considered a layer-2 function, using the layer-2 (MAC, DLCI, VPI/VCI, etc.) addresses. Bridging involves a path selection. A bridge/switch will use a table, e.g. MAC address table, to determine where to forward the layer-2 frames. For example, an ethernet switch will maintain a MAC address table that shows on which interface the destination MAC address was last seen. If the MAC address is not in the table, it will flood the frame to all interfaces.

Really, the wording comes down to the network layer. Bridging/switch happens on OSI layer-2, and routing happens on OSI layer-3. Traffic is sent from host-to-host at layer-2, and network to network at layer-3.

  • Can we say that the thing that happens within the same network is actually path selection while the thing that happens between the networks is routing? I drew this conclusion from your last paragraph. Apr 20, 2018 at 6:03
  • Well, in either case, a device (router, switch, etc.) will need to determine to which interface it should send a datagram (frame or packet), and that is the path election. Each device will independently decide in which direction to send the traffic. None actually predetermine the entire path, just where to send it next.
    – Ron Maupin
    Apr 20, 2018 at 12:34

Routing is configured in layer3 devices, packet with destination ip address checks for matching route entry in other words. Routing is the process of selecting a path for traffic in a network.

Path selections

Static route

If two static routes are configured towards same destination

Path selection will depends upon lowest AD value.

For RiP

Path selection toward destination depends on hops count


Path selection toward destination depends on cost cost is inversely proportional to bandwidth

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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