How Moldtelecom and Kenmei turned a fresh 5G NSA deployment and a triple-band LTE network into measured performance gains, in seven weeks.
The situation
Moldtelecomhad new spectrum on air. A triple-band LTE deployment was complete and 5G NSA was live on C-band.
Both were working. Neither was yet delivering what the spectrum made possible.
This is the normal state of a network right after a deployment. Carrier aggregation is configured but not tuned. Voice parameters carry defaults that no longer match the layer plan. The neighbour plan lags behind the new carriers, so mobility suffers in exactly the places where the new capacity was supposed to help.
None of it shows up as an outage. It shows up as a network that cost more than it returns.
The approach
Moldtelecomasked Kenmei to validate and optimize the network across four areas: LTE multi-band and carrier aggregation, 5G NSA, initial VoLTE tuning, and the cells holding performance back. The timeline was fixed at seven weeks.
The engagementran remotely, on Moldtelecom's own configuration and performance data. Three things shaped how it worked:
• Joint review before implementation. Everyrecommendation was assessed with Moldtelecom's engineers, who know the networkand its history, before anything was changed.
• Implementation in batches, measured after each one.Changes went in as two batches, each followed by a measurement window, so theeffect of each was visible rather than assumed.
Along the way,one problem the data flagged turned out not to be a radio problem at all. Theevidence pointed outside the radio domain entirely, and went to the team thatowned it with the analysis attached. Finding out early what a problem is notsaves as much time as fixing it.
Analysis, not dashboards.
Kenmei's optimization algorithms read configuration, performance counters, topology and traces together, and returned ranked recommendations.
Joint review before implementation
Every recommendation was assessed with Moldtelecom's engineers, who know the network and its history, before anything was changed.
Implementation in batches, measured after each one.
Changes went in as two batches, each followed by a measurement window, so the effect of each was visible rather than assumed.
Along the way, one problem the data flagged turned out not to be a radio problem at all.
The evidence pointed outside the radio domain entirely, and went to the team that owned it with the analysis attached.
Finding out early what a problem is not saves as much time as fixing it.
What changed
Measured bycomparing the two weeks before implementation with the two weeks after thenetwork had stabilized.
What changed
Measured bycomparing the two weeks before implementation with the two weeks after thenetwork had stabilized.
Handover success between sites and frequencies improved to a stable 99,9%.*
Lowest-performing cell for inter-site handover went from 14% to 91%.
Data packet delay on 4G: −35%
Data packet delay on 5G: −25%
In Their Words
“Kenmei delivered a high-quality, detailed proposal and professional execution, shared valuable knowledge and worked with us as one integrated team. Their expertise, transparency, and collaborative approach set a new standard for how we approach network optimization.”
