Wi-Fi airtime and channel reuse: a capacity acceptance worksheet

A coverage survey can show that every room receives a strong signal while the network still performs poorly at occupied hours. Wi-Fi clients and access points contend for shared airtime. Slow clients, retries, wide channels, neighbouring APs on the same channel and non-Wi-Fi interference can consume that airtime before the wired uplink is close to full.

1. Convert the business requirement into a load model

Build the test from real zones and user activity. A reception area with messaging and browsing differs from a training room with video calls, a hotel floor with streaming, a warehouse with scanners or a control room with mobile video. Record peak concurrent clients, device types, band support, application mix, target latency, throughput and acceptable packet loss.

2. Understand channel utilization

Channel utilization is the percentage of observation time during which a radio considers the channel unavailable. It can include transmissions from the local AP, its clients, other Wi-Fi networks and energy that triggers clear-channel assessment. Cisco distinguishes this from RF duty cycle and explains that protocol timers and carrier-sense conditions contribute to overall channel unavailability.

The per-client figure is a planning indicator, not a guaranteed scheduler allocation. Clients send unequal traffic, have different data rates and consume management overhead. Use it to expose impossible density assumptions and then validate with application tests.

3. Distinguish own traffic, neighbouring Wi-Fi and interference

A high utilization number is a symptom. Diagnose its source before changing channels or adding APs.

• High own transmit or receive airtime: the cell is genuinely busy or slow clients are consuming excessive time.
• High Wi-Fi utilization from neighbours: co-channel cells or external networks are sharing the same contention domain.
• High non-Wi-Fi energy: investigate interference using spectrum analysis where available.
• High retries: look for low SNR, collisions, hidden nodes, interference, rate mismatch or excessive cell size.
• Low throughput with low utilization: check client capability, authentication, power saving, QoS, wired path and application server.

Save controller radio counters and a spectrum view during quiet and busy periods. An average across a whole day can hide short capacity failures during shift changes, meetings or evening hotel usage.

4. Use channel width as a capacity decision

Wider channels can raise the data rate of one compatible client, but consume more spectrum and reduce the number of non-overlapping reuse opportunities. HPE Aruba's current RF guidance recommends narrower channels in high-density environments because more cells can operate without sharing the same channel. Cisco similarly warns that 80 MHz bonding can reduce overall performance where many APs compete for limited spectrum.

This simple ratio does not replace a regulatory-domain channel plan. DFS availability, local rules, equipment support and existing interferers affect the usable list. It does show why doubling channel width can sharply increase co-channel reuse in a dense floor.

5. Audit co-channel reuse

Two nearby APs on the same channel usually defer to each other's transmissions if they can hear them. This is contention, not necessarily packet corruption, but the cells share airtime. A distant client may hear both APs differently, creating hidden-node conditions and retries.

At each test zone, record:

• Serving AP and radio.
• Channel and channel width.
• Neighbouring APs heard on the same channel.
• Neighbour signal strength and whether it belongs to the managed network.
• Client retry rate and data rate.
• Channel utilization at the serving and neighbouring APs.

Look for repeating channels in adjacent cells, transmit power that makes cells unnecessarily large, and 2.4 GHz radios enabled more densely than the spectrum can support. Aruba notes that some 2.4 GHz radios may need to be disabled in Wi-Fi 6 campus designs to reduce co-channel interference, provided coverage remains adequate.

6. Measure retry cost, not only retry percentage

A retransmitted frame consumes additional airtime and may fall back to a lower rate. A short frame retry and a large frame sent repeatedly at a low modulation rate do not have the same impact. Use controller counters together with throughput, latency and channel occupancy.

Do not set a universal pass value without vendor definitions. Some platforms count retry attempts, some report retried frames, and aggregation changes interpretation. Establish the metric source, take a quiet baseline and compare the same counters under the planned load.

7. Test with representative clients

An enterprise AP can support modern Wi-Fi 6 features while a large portion of clients cannot. Include at least one common phone, laptop, voice device, scanner or IoT type from the project. Record band, channel width, spatial streams, negotiated data rate, roaming features and driver version.

1. Measure quiet-period RSSI, SNR, latency, retries and throughput.
2. Add clients until the expected concurrent count is reached.
3. Run the planned application mix rather than only parallel speed tests.
4. Measure busy-period channel utilization and per-radio client count.
5. Walk key paths while voice or video sessions remain active.
6. Repeat at the busiest channel-reuse boundary.
7. Check performance after one neighbouring AP is taken out of service.

8. Include latency and loss under load

Bulk throughput can remain impressive while voice, transactions and remote-desktop sessions become unstable. Measure round-trip latency, jitter, packet loss, DNS and authentication response, and application completion time during the same load interval.

Use a wired reference to separate wireless delay from WAN or server delay. Record whether QoS markings are preserved through the WLAN, switch and WAN path, but do not assume QoS can create airtime that has already been exhausted.

9. Test RF automation and DFS recovery

If the network uses automatic channel, power or width assignment, record the settled plan and observe whether it remains stable through the acceptance window. HPE Aruba recommends using RF automation to manage channel plans and avoid radar-affected channels, while persistent DFS events may justify targeted channel exclusion.

Do not trigger radar simulation without authorized tools and procedures. Instead, review historical DFS events and, where a controlled channel change is supported, measure client interruption, reassociation and plan convergence. Confirm that excluded channels are scoped only where necessary.

10. Acceptance checklist

1. Peak client concurrency and application mix are documented by zone.
2. Representative client capabilities and drivers are recorded.
3. Coverage, capacity and application targets are separate.
4. Radio channel, width and power settings are exported.
5. Quiet and busy channel-utilization baselines are captured.
6. Own traffic, neighbouring Wi-Fi and non-Wi-Fi energy are distinguished.
7. Co-channel neighbours are mapped at critical zones.
8. Channel width matches density and reuse requirements.
9. 2.4 GHz radio density is deliberately planned.
10. Retry definitions and baseline counters are recorded.
11. Multi-client load uses representative applications.
12. Latency, jitter, loss and transaction success are measured.
13. Goodput is compared with negotiated data rate.
14. Roaming is checked while the planned load is active.
15. One-AP failure behavior is tested in a critical zone.
16. RF automation stability and DFS history are reviewed.
17. AP switch port, PoE, VLAN and uplink health are verified.
18. Controller exports, survey files, results and deviations are handed over.

11. Handover package

Include the predictive and post-installation survey, floor plans, AP IDs and mounting details, RF profile, channel and power plan, controller version, client inventory, quiet and busy counters, spectrum captures, co-channel map, test traffic profile, latency and throughput results, retry measurements, DFS history, switch-port mapping, deviations and retest evidence. Keep raw survey and controller exports, not only screenshots.

Primary technical references

• Cisco: Enterprise wireless best practices for Apple devices
• Cisco: Wireless RF reference guide
• HPE Aruba Networking: RF planning and channel optimization
• HPE Aruba Networking: Campus RF design
• Extreme Networks: Wi-Fi capacity analytics