CWTS study materials
CWTS study resources

VoWiFi Calls Drop in the Stairwell

CWTS · Put your knowledge to work in context.

All case studies

VoWiFi Calls Drop in the Stairwell

St. Alder Medical Center runs 120 APs across three floors and 500 VoWiFi handsets for nurses. The voice vendor's design standard is -67 dBm RSSI, 20% cell overlap, and roaming under 50 ms. Handsets support 802.11k, 802.11v, and 802.11r. Calls drop consistently when nurses use the stairwell between floors 2 and 3, and occasionally in the corridor outside ICU. Handset logs show a full 802.1X re-authentication taking 1.4-2.1 s at the moment of the drop, and the device associates to a distant AP at -74 dBm before it roams. The controller shows 802.11r enabled on floor 1 and floor 3 APs but disabled on floor 2, where a legacy handset pilot turned it off two years ago. Band steering pushes handsets to 5 GHz; 2.4 GHz radios run at 20 dBm and 5 GHz at 14 dBm. The stairwell has one floor-2 AP mounted 3 m from the door and no AP inside the stairwell.

1. Diagnose the stairwell drops. Explain why the handset associates at -74 dBm and why re-authentication takes 1.4-2.1 s.

Read the worked answer

The stairwell is a coverage hole and a roaming boundary at the same time. The floor-2 AP is 3 m outside the door, so inside the stairwell the handset sees a weak signal from behind a wall and a fire door; the nearest strong AP is on another floor. Because 802.11r is disabled on floor 2, a roam from a floor-2 AP to a floor-3 AP cannot use fast transition, so the handset falls back to full 802.1X/EAP-TLS authentication, which takes 1.4-2.1 s, long enough to drop a voice call. The -74 dBm association is a sticky-client symptom: without 802.11k neighbor reports or 802.11v BSS transition management, the handset does not know a better AP exists and stays on the floor-2 AP until its own roam threshold is crossed. 5 GHz at 14 dBm versus 2.4 GHz at 20 dBm widens the mismatch even with band steering.

2. List the configuration changes you would make on the controller and APs, and the coverage changes in the stairwell.

Read the worked answer

Enable 802.11r on every AP with the same mobility domain and FT-EAP, and verify the handsets actually negotiate FT rather than falling back to OKC. Enable 802.11k and 802.11v so clients get neighbor reports and BSS transition requests. Fix the stairwell coverage: add an AP inside the stairwell or on a landing, or move the floor-2 AP closer to the door and use a directional antenna; aim for -67 dBm on the landings and 20% overlap between floors. Rebalance power: 5 GHz at 14 dBm is 6 dB below 2.4 GHz, so raise 5 GHz or lower 2.4 GHz to make 5 GHz the preferred band, and check that the floor-2 5 GHz cell overlaps the floor-3 cell vertically. Do not simply raise all power, which enlarges cells and creates more sticky clients.

3. How would you validate the fix and what acceptance metrics would you report?

Read the worked answer

Validate with a walk test on the stairs and the ICU corridor using the actual handsets, logging RSSI, SNR, roaming type, authentication time, and packet loss. Acceptance: RSSI at or above -67 dBm on every landing and corridor, SNR at least 25 dB, 802.11r FT used for at least 95% of roams, roam completion under 50 ms, packet loss under 1%, and ten consecutive test calls with no drop. Collect controller statistics before and after, and run a 24-hour sample of voice quality/MOS from the handset management system. Repeat the test during a shift change when the stairwell is crowded, because body absorption and client density change the RF. Document floor plans, AP changes, and channel/power settings so the result can be reproduced at the next audit.