Put the symptom in context
A performance symptom rarely explains its own cause. Poor measured service may relate to coverage, interference, congestion, an operational incident or the test conditions. Treating the first available dataset as the whole answer risks sending an engineering team down the wrong path.
In the Metricell Platform, an optimisation investigation starts with an area, site or cell and a relevant time window. The configured RAN Optimisation agent can correlate the connected RF, performance, configuration, ticket and measurement sources. Its role is to support the engineer’s investigation, including identifying evidence that is missing or contradictory.
Follow the data through the investigation
Site and cell configuration establish the technology, bands and network arrangement. OSS performance records provide the available trends in utilisation, availability or service indicators. Incident and planned-work records add operational context. Predicted coverage and crowd or field measurements provide different views of service in the area.
Those sources are not interchangeable. A prediction describes expected service under its model assumptions. A crowd dataset represents observations over its reporting period. A ticket may describe current work but still need checking against the time of the reported problem. The agent’s configured method should retain these distinctions.
The platform material illustrates an area investigation where supporting evidence is presented beside the narrative. An unavailable source remains an explicit limitation. The example demonstrates the investigation pattern; it is not a claim that a particular live network has the same condition today.
Keep the intervention accountable
The useful output is a likely explanation, supporting evidence and proposed next steps at the right technical depth. The engineer can drill into source records, refine the area or ask for another check. Where the evidence is insufficient, collecting additional measurements may be the right recommendation.
The investigation supports radio network planning and optimisation by helping engineers assess the evidence, identify potential improvements and decide what action to take. The agent does not have permission to change radio parameters.
After an authorised intervention, compare suitable pre- and post-change evidence. Keep the period, load conditions, technology and test method sufficiently comparable to avoid attributing every improvement to the action.
Evaluate the workflow in your environment
A practical evaluation can track time spent locating evidence, repeated requests between teams, engineer corrections to findings and completion of post-change checks. These are measures to establish with the operator, not promised results.
Ask Metricell for a RAN investigation demonstration using an agreed scenario. Start with the operational question, then examine the sources, the method and the decision the engineer is able to make.
Enquiries: marketing@metricell.com
.png)
