Forum

Hi,

Why is the concept of connecting different BSSs to form a ESS necessary? Suppose the config is

STA1--->AP1 -----DS------ AP2<---STA2

If STA1 wants to communicate with STA2, why can't the packet delivery work as if BSS1 and BSS2 are different broadcast domains? What is the advantage of having this concept of ESS that makes STA2 in a different BSS appear to be in the same BSS at the LLC layer?

By taking approach 1 (no ESS) then we do need this 4 address field idea, since the eventual destination and actual source will be in the IP header.

Arvind

You are right that packet delivery can work correctly based on regular layer 2 mechanisms, regardless of whether an ESS is formed or not. My interpretation has been that the primary purpose of an ESS is to facilitate roaming between APs. I bet Criss can do a better job of answering, as he is the master of the standards.

Hi Arvind:

IEEE 802.11 documentation does not tell us why the designers included the concept of pooling the basic service sets (BSS) of multiple access points (AP) into a single extended service set (ESS) making the BSS boundaries invisible to attached logical link control (LLC) entities. But they did.

Avoiding layer three technologies such as IP in forwarding decisions creates a system that is cheaper and faster, although typically constrained to a local area (LAN). Allowing two BSS's to interoperate extends the reach of the technology making it more useful.

In short ESS gives us higher performance at lower cost.

I hope this helps. Thanks. /criss