Test one wired device, check the plan, then compare results at different times. Change radio settings only after you have a repeatable baseline.
Run a wired test first
Connect a computer to a LAN port and turn off its Wi-Fi so the test uses Ethernet. Pause other traffic where practical. Check the negotiated Ethernet link speed on the computer: a slow port, adapter or damaged cable can limit the result before the cellular connection is fully used.
Use the same test service and server for repeated comparisons. Record three runs, including upload and latency, and keep the middle result as well as the range. Avoid running two speed tests at once. Large tests can use a substantial part of a metered data allowance.
If wired tests are consistently faster than Wi-Fi tests from the same computer, investigate local Wi-Fi placement, interference and client capability. Changing the cellular antenna will not fix that local bottleneck.
Check the SIM plan and operator status
Look for a monthly high-speed allowance, a speed cap, roaming restrictions or different terms for router use. Check for a reported outage. A connected SIM can still be subject to a lower service speed.
Use the APN supplied by your operator for that SIM. An APN copied from an unrelated online guide is not a speed upgrade. Ask the operator to confirm the correct configuration if the plan and router settings do not agree.
Compare quiet and busy periods
Repeat the wired test in the morning and during your normal busy period without moving the router. Log the active bands and signal readings as well as the speed.
Similar radio readings with much lower evening speeds suggest congestion or traffic management, but do not prove it. Share the timed results with the operator. A larger antenna cannot create capacity on an overloaded network.
Check signal quality, not just strength
A strong received signal can still contain substantial interference. If available, compare RSRQ and SINR while trying a different router position. Do not assume that one extra signal bar means a faster connection.
A phone is useful as a second reference, but it is not a like-for-like router test. Its modem, antennas, supported band combinations and plan may differ. Even nearby devices can use different network resources.
Would locking a band help?
Only test this if the exact model and firmware support it. Save the automatic settings first and stay connected locally so you can restore them. Compare automatic selection with the supported alternative under similar conditions; include upload, latency and connection stability.
Restricting bands can remove useful carrier-aggregation combinations. Carrier aggregation allows compatible devices and networks to use multiple carriers together. A restriction that improves one test can reduce speed or coverage later. Return to automatic selection if the benefit is not repeatable.
Do not apply a band list from another country or operator. Do not treat a band lock as a cell lock. A separate, model-specific procedure is needed for either setting.
Send useful evidence with a support request
Include the router model, firmware version, operator, approximate location, time of the problem and wired test results. Add signal readings and active bands if available. Say whether the problem affects every site or only one application.
If your organization allows it, compare the affected application with and without its VPN. Do not disable a required security service. A problem limited to one VPN or destination needs a different investigation from a slow connection everywhere.
- Do not post SIM identifiers, passwords or full diagnostic logs publicly.
- Note recent changes to antennas, cabling, firmware or the data plan.
- Describe the practical symptom: broken calls, slow uploads or long page loads.