Designing Wi-Fi for a 1,000-seat university lecture hall

A business school auditorium with 1,000 seats, used for large lectures that are also taught live to students attending remotely. The ageing Wi-Fi could not meet the demand of a full room, and professors were reporting performance problems that affected both audiences at once.

The room fails when it is full. Coverage in an empty auditorium tells you nothing about what 1,000 devices do to it, and a stall during a live lecture interrupts the teaching itself.

• High capacity was required for a full room, not an average day.
• Real-time classroom applications needed predictable performance, not best effort.
• The existing infrastructure was aging (802.11n).
• Mounting options in a tiered auditorium are limited, so the answer could not simply be more access points overhead.

We started from what has to work in the room, then modelled it, measured it, and validated it after deployment.

• Collected the real user and application requirements from teaching and IT staff, and wrote them down as numbers.
• Built a predictive Wi-Fi model of the auditorium against its actual construction and seating, taking into account capacity requirements.
• Took on-site measurements to correct the model against measured reality rather than a drawing.
• Performed post-deployment validation in the room during a large lecture, signed by a named CWNE.

400 Mbps+

effective throughput at full occupancy

40%

of clients connected on 6 GHz

0

complaints from professors and students

30x Wi-Fi 7 access points with directional antennas, 6 GHz in service, and placement driven by the measurements instead of a ceiling grid. The hall now holds its performance at full occupancy with no problem.

• Directional antennas keep each cell on its own section of seating, which holds interference down in a single large volume and help with load balancing the clients between the access points.
• The 6 GHz band carries a meaningful share of the load, leaving 5 GHz less congested while providing better throughput.
• The university holds the validation measurements, so any future complaint starts from data.