MRO

MRO exists because handover parameters are a balancing act and the right value drifts over time. Set the trigger too aggressive and the UE hands over too early — it bounces into a neighbour that quickly gets worse. Too conservative and it hands over too late, hanging onto a dying serving cell until the radio link fails. And if two cells keep tossing a UE back and forth, that's ping-pong. All three hurt, and all three show up in your drop and handover-failure counters.

MRO watches the symptoms — radio link failures, the timing of reconnections, handovers that fail just after completing — and infers which problem is happening, then adjusts the relevant offsets and timers (think the handover offset and Time to Trigger for the affected relation). The key idea is it diagnoses by pattern: an RLF a second or two after a successful handover smells like too-early; an RLF before any handover attempt smells like too-late. Getting that classification right is most of the work, which is why MRO tuning is conservative and incremental rather than dramatic.