VoWiFi Handsets Dropping Calls Between Floors
A 480-bed hospital runs a single WPA2-Enterprise WLAN with 210 APs managed by three controllers. Nurses carry VoWiFi handsets and roam between floors. The nursing director reports roughly 25 dropped calls per shift, mostly when staff walk from a patient tower into the adjacent stairwell and elevator lobby. Handset logs show the device stays associated with its original AP until RSSI falls to about −80 dBm, then takes 8–12 seconds to reassociate and re-authenticate. A few handsets sit on 2.4 GHz for entire shifts even when 5 GHz is available at −55 dBm. APs transmit at full power on 2.4 GHz, and the same APs carry an open guest SSID. The infrastructure supports 802.11k, 802.11v and 802.11r, but none are enabled. Change control requires evidence before configuration changes, and clinical downtime is not permitted.
1. Diagnose the drop mechanism and match each 802.11 feature to the part of the problem it fixes.
Read the worked answer
Two faults compound. First, the handset is sticky: it holds a link down to −80 dBm instead of roaming near −67 to −70 dBm, so voice frames are already failing before the roam begins. Second, when the roam finally happens, 8–12 seconds means a full reauthentication, not just a reassociation - that is far longer than a voice handoff budget.
Feature mapping:
- 802.11k neighbor reports cut the scan phase. The handset asks the infrastructure for a neighbor list instead of scanning all channels, reducing the discovery time that dominates the front of the roam.
- 802.11v BSS Transition Management lets the infrastructure suggest a better target AP, which addresses the sticky-client behavior by prompting a roam before −80 dBm.
- 802.11r Fast BSS Transition addresses reauthentication. With FT, the handset derives keys ahead of time (over-the-air or over-the-DS) so the new AP does not need a full EAP exchange. Confirm the handsets actually support FT; if they do not, enable PMK caching/OKC as a partial substitute and plan handset replacement.
Also fix band behavior: reduce 2.4 GHz power so handsets do not cling to 2.4 GHz, and verify the voice WLAN is 5 GHz-first. None of these fixes a coverage hole in the stairwell - check that first.
2. Change control demands evidence. What data would you collect, and what is a safe pilot plan?
Read the worked answer
Collect evidence that isolates each stage of the roam, not just complaints:
- Client-side logs from a sample of handsets: RSSI at roam decision, time from last good voice frame to first frame on the new AP, and authentication type used on reassociation.
- Over-the-air capture on the roaming channel showing the authentication and association exchange and whether a full EAP conversation occurs each time.
- Infrastructure counters: controller roam statistics, 802.11 association/authentication failure counts per AP, and channel utilization.
- Beacon and probe inspection confirming whether the WLAN currently advertises mobility domain and FT capability, and whether k/v capability bits appear.
- RF survey of the stairwell and elevator lobby to test the hypothesis that the drop zone is also a coverage hole.
Pilot plan: enable 802.11k and 802.11v on the voice WLAN only, on one floor of one tower, outside shift change. Measure roam time with the same handset sample. Then enable FT with a mobility domain, validating key derivation and that non-FT handsets still associate. Keep a documented rollback per step, and schedule changes so a controller failure cannot coincide with a clinical peak.
3. The stairwell and elevator lobby report is separate from the roaming settings. What design actions apply there, and what is the risk of simply raising AP power?
Read the worked answer
Stairwells and elevator lobbies are typically metal-and-concrete enclosures with doors that close, so the roam target may be weak or absent exactly where the handset decides to move. Design actions: add coverage at the transition zones (an AP or antenna serving the stairwell landing and the lobby on each floor, or at minimum every other floor), and place APs so that the overlap between cells occurs in the corridor before the door, not inside the stairwell. Ensure the voice WLAN uses 20 MHz channels for more reuse and lower delay spread.
Raising AP power is the wrong lever: higher 2.4 GHz power makes cells larger and stickier, increases co-channel interference across floors, and creates more clients at the cell edge holding on too long. It also raises the number of clients per AP in a high-density clinical environment. If anything, lower 2.4 GHz power and tune 5 GHz coverage to the actual roaming paths, then verify with a walk test that the handset roams in the corridor at around −67 dBm with a sub-100 ms interruption.