Warehouse 2.4 GHz Roaming and Channel Plan
An 18,000-square-foot distribution center has 14 ceiling-mounted APs, mostly 9 m high above steel racking. Handheld 2.4 GHz barcode scanners and 5 GHz laptops must roam across aisles. The 2.4 GHz plan uses channels 1, 6, and 11, but the installer also assigned channels 3 and 9 to four APs to spread load. All APs transmit at maximum power. Scanners report random disconnects and retransmits near aisle intersections; laptops on 5 GHz are mostly stable. A spectrum sweep shows strong co-channel energy and several non-Wi-Fi interferers in the 2.4 GHz band. The business cannot replace scanners this quarter, and downtime during inventory is unacceptable. You must improve roaming and reliability without adding APs. The scanners are older and support only 2.4 GHz; no firmware update is available.
1. What is the most likely root cause of the scanner instability, and what channel and transmit-power changes would you make first? Explain the reasoning in terms of RF and roaming.
Read the worked answer
The channel plan is the first problem. In 2.4 GHz, only channels 1, 6, and 11 are non-overlapping; channels 3 and 9 overlap with neighbors and create adjacent-channel interference. Maximum power compounds this by making each AP cell large, so scanners stay associated to a distant AP until the signal is very low and then roam late. That fits the failures at aisle intersections.
First, remove 3 and 9 and set 20 MHz channels using only 1/6/11, then survey and assign channels so same-channel APs are physically separated and not both strong at the same scanner location. Reduce 2.4 GHz EIRP in 3 dB steps on the APs with the largest overlap, while checking scanner coverage at rack level. Target roughly -67 dBm or better and SNR of at least 25 dB at required scan points; do not lower every AP blindly because high racking creates nulls. Enable 802.11k/v only if the scanners support it.
2. A colleague proposes raising 2.4 GHz power to 30 dBm so scanners see a stronger signal everywhere. Why may that worsen the problem?
Read the worked answer
More power does not fix contention. A 30 dBm cell overlaps more neighboring cells, raises the noise floor, and increases co-channel interference. Clients also see a usable signal farther away, so they stick to the old AP and roam even later; the roam may then be abrupt. In a metal warehouse, strong reflections increase multipath and retries. It can also exceed regulatory EIRP after antenna gain. The goal is controlled cell edges and good SNR, not the strongest possible RSSI everywhere. Raising power would likely make scanner disconnects less frequent at one spot but more widespread under load.
3. After the change, which three measurements would you collect, and what would make you accept or reject the result?
Read the worked answer
Collect (1) a scanner walk test with retry/disconnect counts and roam points per aisle, (2) RSSI and SNR at scanner height along the same path, and (3) per-radio channel utilization and noise floor during a busy period. Accept only if the walk test shows no disconnect and retry rate stays below about 10%, required scan points remain at -67 dBm or better with SNR >=25 dB, and 2.4 GHz utilization stays below roughly 50% at peak. Reject if improvement depends on maximum power, if coverage holes appear at rack faces, or if retries fall at one intersection but rise elsewhere. Recheck 5 GHz laptop coverage because power changes can alter band-steering behavior.