6.2 WLAN, Bridging, Peer-to-Peer and Mesh

CWNA-109 objective 2.7 asks you to pick a use case, not to recite every MAC address field. A later chapter treats stations, BSSID, SSID, distribution system, and distribution system media as protocol machinery. This section is the Harborline selection guide: when is a Basic Service Set (BSS) enough, when do you extend it into an Extended Service Set (ESS), when do you bridge LAN segments, when is peer-to-peer Independent Basic Service Set (IBSS) acceptable, and when does mesh earn the extra moving parts.

Harborline Terminals still sits on that coastal pier. The gatehouse is tiny. The warehouse is huge. Fiber across the slip needs a marine permit. Container stacks kill Ethernet runs to light poles. Two laptops on a vessel sometimes need a file copy when shore Wi-Fi is gone. Those four sentences are four different 802.11 jobs. Mixing them on one omnidirectional indoor AP is how tickets multiply.

Wireless LAN: BSS and ESS

A BSS in infrastructure mode is one AP radio cell plus the stations associated to it. One network name in that cell, traffic that can reach the wired LAN through that AP, coverage that matches one room or one small building. Harborline's gatehouse is a BSS problem: a handful of PCs, a badge printer, and a camera. One indoor AP, correct channel and power, done. If you add a second AP in the same closet "for backup" with a different SSID and no roaming design, you did not build high availability—you built a second WLAN that users must pick by hand.

An ESS is multiple BSSs that cooperate so a user still sees one network name while moving. The APs share an SSID (and usually a roaming design, VLANs, and a wired or wireless distribution path). Harborline's warehouse is an ESS problem: forklift tablets must not drop when they pass from receiving to shipping. Twelve APs with twelve unique SSIDs are twelve WLANs, not an ESS. Twelve APs with one SSID, planned channels, and a path to the same LAN are an ESS.

Keep the protocol depth for later. You do not need four-address frames to choose ESS versus BSS. You need the intent. BSS = one cell. ESS = a campus of cells that should feel like one WLAN. Roaming quality, fast secure roaming, and controller architectures are later design topics. Objective 2.7 only needs you to recognize that walking users across many APs is an ESS use case, and one AP for one office is a BSS use case.

Harborline should not "solve" warehouse coverage by cranking one BSS to outdoor EIRP. That is a spectrum-rules failure from the previous section, not an ESS. Coverage holes want more cells, not one screaming omni.

Wireless bridging

A wireless bridge connects LAN segments. The payload is Ethernet between switches, not "employees with laptops" as the primary community. Two layouts dominate teaching:

• Point-to-point (PtP): two radios, two buildings, one hop. Directional antennas, outdoor SKUs, a dedicated channel. Harborline's HQ intermediate distribution frame to the cold-storage switch across a 180-meter slip is PtP. The marine permit for fiber is six months out; the bridge is this quarter's path for WMS scanners already on that far LAN.
• Point-to-multipoint (PtMP): one root sector, several remote LANs. Harborline's yard office feeding three small outbuildings (customs booth, maintenance shed, reefer-plug hut) is PtMP. Each remote is still a LAN, not a roaming client.

Bridging is not "an AP with a long-range SSID." Walking clients may be forbidden on the bridge radio or parked on a different radio. The hop often uses a dedicated 5 GHz channel. UNII-3 is a frequent outdoor pick because classic DFS surprises are fewer than on UNII-2/2e—spectrum choice from objective 2.6 applied to a use case. If Harborline's goal is "connect two networks," choose bridging. If the goal is "people walk around with phones," choose BSS/ESS or mesh access. Collapsing both jobs onto one omni AP usually produces a noisy client cell and a fragile backhaul.

PtP versus PtMP is a geometry choice. Two closets: PtP. One closet and a handful of small remote LANs that can share a sector: PtMP. Three long independent hops with different azimuths may still be cleaner as three PtP pairs if you have the spectrum and the mounting points. Bridging does not replace an ESS inside each building. Cold storage still needs its own indoor BSS or ESS after the frames arrive on copper.

Wireless peer-to-peer (IBSS)

An IBSS is ad hoc 802.11: stations talk peer-to-peer with no access point. There is no AP to attach the cell to Harborline's switched LAN. Useful when two laptops on a vessel need a file copy and the shore WLAN is out of range. Painful as an enterprise architecture: no central roaming, no easy wired handoff, weak manageability, and security is easy to get wrong.

Harborline does not run the warehouse as IBSS. They might allow a technician IBSS for a ten-minute packet capture when an AP is dark, or a temporary laptop-to-laptop share on a ship. Exam items that describe "no access point, two stations, temporary" want IBSS. Items that describe "corporate SSID, DHCP from the LAN, roaming" want BSS/ESS.

Do not rename a misconfigured infrastructure BSS as "ad hoc" because two phones can see each other. If an AP is present and stations associate to it, you are in infrastructure BSS/ESS territory. IBSS means the service set has no AP.

Wireless mesh

Mesh extends coverage when Ethernet backhaul is missing. Nodes forward frames over a wireless backhaul. Two roles you must name:

• Mesh point (MP): participates in the mesh and forwards backhaul traffic. It may not offer a client BSS.
• Mesh access point (MAP): an MP that also serves access clients in a BSS.

Backhaul versus access is the other split. Backhaul is node-to-node, often a dedicated radio, SSID, or channel (frequently 5 GHz). Access is the SSID scanners and phones join (2.4 GHz, 5 GHz, or both). Collapsing both onto one 20 MHz 2.4 GHz channel is how Harborline's yard melts: every hop burns airtime twice, and a scanner frame may ride the same contention as the backhaul that is trying to deliver it.

A root node still needs a path to the wired LAN. Mesh is not magic Ethernet. It is a use case for awkward geometry: Harborline's container stacks, moving cranes, and cameras on light poles. If a mid-yard MP fails, a well-designed mesh can shift paths. If the only root dies, the graph has nowhere to send LAN-bound frames. Plan roots the way you plan uplinks.

Do not confuse mesh with bridging. A PtP bridge is two LAN islands and a dedicated hop. Mesh is a graph of MPs/MAPs that can heal when a node fails, plus optional client access on MAPs. A MAP in the yard is doing two jobs: forward toward the root, and serve Harborline scanners. An MP on a light pole with no SSID is doing one job: backhaul relay.

Harborline decision matrix

Walk the pier with that table in your pocket.

The gatehouse stays a BSS. Adding mesh "because wireless is cool" adds failure modes without covering more users.

The warehouse stays an ESS on indoor APs with Ethernet and Power over Ethernet. Mesh inside a building where cable is already in the truss is usually the wrong spend. If one aisle is a coverage hole, add an AP on copper—not a MAP that backhauls across the same 5 GHz you needed for clients.

The slip stays a PtP bridge. Do not publish the warehouse SSID on the bridge radios and hope forklifts roam across water. Forklifts belong to the ESS on each side; the bridge only joins the two Ethernet islands.

The outbuildings can be PtMP if each hut is a small LAN behind a switch. If a hut is really "two scanners and no switch," a MAP or a tiny BSS with some other backhaul may be simpler than pretending those two devices are a LAN segment.

The yard poles are mesh. Put MAPs where scanners walk. Put MPs on poles that only need to relay. Keep backhaul on a dedicated radio or channel—often outdoor 5 GHz with directional or sector antennas—and keep access as its own BSS. Mix DFS-heavy UNII-2e into that backhaul only if you accept a minute of silence after a radar event; many Harborline templates park backhaul on UNII-3 for that reason.

The vessel laptops may use IBSS when the ship is off the ESS. The moment those laptops can see a Harborline AP, they should join the BSS/ESS like every other station. Ad hoc is a spare tire, not the fleet.

Contrasts the exam likes to blur

ESS versus mesh. Both can cover a large area with many radios. ESS assumes wired (or otherwise provisioned) backhaul into each AP. Mesh assumes wireless backhaul between MPs/MAPs. If Harborline can pull copper, ESS is the simpler WLAN. If they cannot, mesh is the WLAN that still reaches the pole.

Bridge versus MAP. Both use wireless to extend beyond a single room. A bridge's community is LAN segments. A MAP's community is associated clients plus the mesh graph. If the far side is a switch closet, bridge. If the far side is a walking scanner, MAP (or an ESS AP with copper).

IBSS versus BSS. Both are service sets. IBSS has no AP. BSS in this use-case chapter means infrastructure—an AP is the center of the cell. "Peer-to-peer" in objective 2.7 means IBSS, not two phones associated to the same Harborline AP.

PtP versus PtMP. Both are bridging. Count the LAN endpoints. Two: PtP. One root plus several remotes: PtMP.

Harborline's production picture, said in one sentence: ESS indoors, PtP across the slip, PtMP only where remote closets share a sector, mesh in the yard with MAP versus MP split on purpose, IBSS only in the break-glass kit. That sentence is objective 2.7. Leave BSSID math, four-address frames, and distribution-system media for the service-set chapter that follows.