Middle School 1:1 Rollout: RF Congestion or Bandwidth?
Walnut Grove Middle School (grades 6-8; 720 students, 88 certified staff) began a 1:1 Chromebook rollout in January. The building has 26 dual-band access points, but the only student SSID is broadcast on 2.4 GHz, with channels 1, 6, and 11 repeated in every wing. The district's sole WAN link is a 50 Mbps fiber circuit shared with the adjacent high school. Since the rollout, teachers report the LMS taking 20-40 seconds to load and unit quizzes timing out; the wireless dashboard shows airtime utilization above 80% in three science classrooms during periods 2-4 and retry rates near 30%. The high school reports no performance change. The principal wants to order a bandwidth upgrade before the spring Illinois Assessment of Readiness window opens in nine weeks. The business office has $22,000 left in the fiscal year, and the technology director asks you for a recommendation.
1. Before any purchase, what is the most probable cause of the slowdown, and what evidence would confirm or rule it out?
Read the worked answer
The high school is unaffected while sharing the same 50 Mbps circuit, which argues strongly against WAN saturation as the primary cause. The middle school symptoms point to the radio layer: a 2.4 GHz-only SSID, three non-overlapping channels reused across every wing, airtime above 80%, and roughly 30% retries. Retransmissions consume airtime, so latency climbs even when total throughput is low.
Evidence to collect before spending:
- Per-AP associated client counts and the band capability of the Chromebooks (most are 5 GHz capable, so if they are associating on 2.4 GHz the SSID configuration is pinning them there).
- Retry percentage, SNR, and channel utilization per AP, broken out by hour.
- Router interface utilization for both buildings during periods 2-4. If the middle school peaks well under 50 Mbps, the circuit is not the bottleneck.
- A passive survey in the three science rooms showing co-channel overlap and RSSI at the student seats.
- Firmware versions, band-steering settings, and channel width.
If WAN utilization is moderate while retries are high on specific APs, the finding is RF contention and client density, not insufficient bandwidth.
2. Sequence the remediation you would fund with the $22,000, explaining what comes first and what you would defer.
Read the worked answer
Order matters more than the total spend.
- Configuration changes first (near zero cost, reversible, days not weeks): enable band steering and a 5 GHz-preferred SSID, drop 2.4 GHz channels to 20 MHz, disable legacy 802.11b rates, set a minimum basic rate, update AP firmware, and rebalance transmit power so cells shrink instead of bleeding across the building. Re-measure immediately.
- A professional RF survey plus targeted AP additions or replacements (roughly $8,000-$12,000 installed) only where the survey shows coverage holes or more than about 30-40 active clients per radio. Do not blanket-add APs; over-dense deployment creates its own co-channel problem.
- Assessment-readiness measures: a dedicated testing SSID or QoS marking, allowlisting the assessment domains, caching, and suspending automatic Chromebook updates during the window. Cheap, and it protects the nine-week deadline.
- Defer the WAN upgrade. It is the largest line item and the least supported by the data. If router graphs later show sustained utilization above roughly 70% at peak, pursue it as an E-Rate Category 1 request in the next cycle with evidence attached.
Hold about 10% as contingency and record baseline metrics so the effect of each step is measurable.
3. Write the three-to-four sentence message to the principal explaining why the bandwidth order should wait.
Read the worked answer
The high school shares the same 50 Mbps circuit and has not slowed down, which tells us the circuit is not the bottleneck - all 720 Chromebooks are competing for three crowded radio channels in the 2.4 GHz band. For roughly a third of the cost of a bandwidth upgrade we can reconfigure the wireless network this month and add access points only where a survey shows real gaps, which should cut load times before the assessment window. If router data after those changes still shows the circuit maxed out, I will bring you an E-Rate Category 1 request for the next cycle with the evidence to justify it. I will report baseline and post-change numbers at the end of week three so you can see whether the fix worked.