I85PREP
⌘ K
ENWLSI 300-430 study materials
ENWLSI 300-430 study resources

Voice and Coverage Judgement in a High-Rack Distribution Center

ENWLSI 300-430 · Put your knowledge to work in context.

All case studies

Voice and Coverage Judgement in a High-Rack Distribution Center

Northline Logistics runs a 180 m by 90 m distribution center served by a C9800-40 on 17.9.4 with 48 Catalyst 9130AXI access points mounted in the ceiling grid at 12 m. Steel racking is 10 m tall and product is stored to the top, so only 2 m of clearance remains above the racks; aisles are 3 m wide. Clients are picker voice handsets on a 5 GHz-only WPA2-PSK WLAN (mix of 7925 and 8821 models), forklift tablets on the corporate WPA2-Enterprise WLAN, and a guest WLAN using web authentication. During shift change the 2.4 GHz utilization reaches 78 percent and the corporate WLAN peaks at 1,100 clients. Pickers report dropped calls and 300 to 800 ms of one-way voice latency specifically between aisles 12 and 20; RRM shows about -73 dBm at 1 m above the floor in those aisles while the AP radios themselves report -52 dBm. Clients commonly stay associated to an AP until -82 dBm. No new cabling can be pulled, and the APs must remain in the existing ceiling grid positions.

1. Explain why the ceiling APs deliver acceptable RSSI at the AP but poor signal at floor level in aisles 12 to 20, then state the design changes you would make under the no-new-cabling constraint and the RF targets you would hold the installer to.

Read the worked answer

Why it fails

  • The AP is only 2 m above the top of 10 m racking. In a 3 m-wide aisle bounded by full-height steel and dense product, the path from ceiling AP to handheld is a diffraction over the rack edge plus multipath down a lossy canyon, not free space. Budget 15 to 25 dB of excess loss versus the clean line-of-sight number RRM is implying.
  • The internal antenna pattern of a 9130AXI is designed for ceiling coverage below the AP; with stock to the rack top there is no useful main lobe inside the aisle.
  • RRM neighbor data is computed from AP-to-AP path loss, so RRM sees a healthy cell even where client coverage is poor. Do not size this site from RRM alone.

Fixes within the constraint

  • Replace selected 9130AXI with 9130AXE (external antenna ports) and side-mount them on rack ends or on existing cable tray at 3 to 4 m, feeding 90-degree patch antennas aimed down the aisle.
  • Where side-mounting is impossible, drop the AP from the ceiling grid on the existing cable to a 4 to 5 m pole at the aisle head.
  • Keep ceiling-mounted 9130AXI for open floor and staging areas only.

Design targets

  • -67 dBm or better at 1 m above the floor in every aisle, SNR of at least 25 dB, and at least two APs above -72 dBm for redundancy.
  • 20 MHz channels in 5 GHz; 40 MHz only if a second, data-only WLAN needs it.
  • 15 to 20 percent cell overlap, and per-AP client counts capped near 30 for voice.
  • Prove it with a post-install survey on the actual handset, not a laptop NIC.

2. Handsets hold a session until about -82 dBm. Which C9800 features would you enable to shorten roam decisions and cut mid-call drops, and what compatibility issue must you resolve first given both 7925 and 8821 handsets?

Read the worked answer

Why the handset waits The roaming decision belongs to the client. A Cisco handset with CCX will scan and roam on its own thresholds, but in a canyon environment the neighbor list it builds is poor, so it hangs on until the link is nearly unusable. The infrastructure's job is to give it better information and to remove the slow-rate safety net.

Configuration on the C9800

  • Disable 1, 2, 5.5 and 11 Mbps on the 5 GHz voice WLAN and set 12 and 24 Mbps as mandatory, 36 and above as supported. Forcing higher rates makes the client roam earlier because the usable cell edge shrinks.
  • Enable 802.11k neighbor list so the handset scans known neighbors rather than all channels.
  • Enable 802.11v BSS Transition Management on the WLAN; it lets the controller steer a client that is far from its AP.
  • If available in your release, use the per-WLAN client roaming (optimized roaming) check with a conservative threshold, for example -75 dBm over a 6 s interval. Too aggressive a value causes ping-pong and worse voice quality than no feature at all.

The compatibility issue

  • The 7925 series is a CCKM-era device; the 8821 is the generation that does 802.11r/k/v. Fast transition (802.11r) and CCKM are generally not both usable on the same WLAN on AireOS-style configuration, so confirm the behavior on your release in the lab before enabling FT.
  • If they conflict, split the voice SSID into a CCKM WLAN for 7925 and an FT WLAN for 8821, or accelerate handset replacement. Do not silently break half the fleet.
  • Also enable call admission control (ACM) for voice so a busy AP rejects new calls instead of degrading existing ones.

3. 2.4 GHz sits at 78 percent utilization at shift change and the corporate WLAN carries 1,100 clients. Describe how you would reduce 2.4 GHz pressure and how you would verify the change actually helped rather than just moving the problem.

Read the worked answer

Reduce 2.4 GHz pressure

  • Enable band select on the corporate WLAN: suppress 2.4 GHz probe responses for about 200 ms, cycle count 2 to 3, age-out around 20 s, acceptable client expiry around 300 s. This pushes dual-band clients up without touching the single-band forklift tablets.
  • Consider disabling 2.4 GHz radios entirely on APs in dense open areas, keeping 2.4 GHz only where single-band devices genuinely operate.
  • Trim 2.4 GHz to 1, 6 and 11 only, disable 1, 2, 5.5 and 11 Mbps data rates so beacons and management frames do not burn airtime at legacy rates.
  • Turn on airtime fairness so one slow client cannot consume the cell.
  • Verify the corporate WLAN is not running 802.11b-compatible settings for a device population that is actually 802.11n/ac/ax.

Verification that means something

  • Baseline before/after: per-radio channel utilization, retry rate, and the count of clients per band from the controller, plus a spectrum snapshot in the worst aisle.
  • Watch the voice WLAN specifically. Band steering that pushes handsets onto an already-congested 5 GHz channel makes calls worse, so confirm voice channel utilization and call quality metrics, not just total client counts.
  • Confirm the single-band population did not lose connectivity: count failed associations per SSID and per band.
  • Re-check roaming: if more clients now sit on 5 GHz, revisit channel width and the optimized-roaming threshold, because the first change can invalidate the second.