Mobility Load Balancing (MLB)

The core mechanism of MLB is a closed-loop control process that reacts to measured cell load imbalances. When a serving cell's composite available capacity (CAC) exceeds a defined threshold, the MLB function initiates a load transfer.

• Trigger: Monitors metrics like PRB utilization, RRC connected users, or hardware load indicators.
• Action: Modifies the Cell Individual Offset (CIO) for a specific neighbor relation, effectively shrinking the overloaded cell's footprint and expanding the target cell's coverage area.
• Constraint: Adjustments must respect a defined handover boundary margin to prevent coverage holes and must not cause a ping-pong effect where a UE is repeatedly handed back and forth.

MLB achieves load transfer by dynamically tuning the parameters that govern the A3 event, the primary LTE/5G NR handover trigger. The key parameter is the Cell Individual Offset (CIO).

• A3 Event: Triggered when a neighbor cell's RSRP is offset better than the serving cell's RSRP.
• CIO Tuning: A positive CIO makes a neighbor cell appear artificially stronger, encouraging UEs at the cell edge to handover earlier. A negative CIO delays the handover.
• Granularity: CIO adjustments can be applied per neighbor relation, allowing for highly targeted traffic steering between specific cell pairs rather than a uniform coverage shrink.

MLB is distinct from simple coverage-based mobility. It is a load-aware process that prioritizes traffic distribution over pure signal strength optimization.

• Objective: Equalize the composite load between cells, not just maximize individual UE throughput.
• Metric Hierarchy: The algorithm considers a multi-dimensional load metric, often combining PRB utilization, PDCCH CCE utilization, and backhaul capacity, not just radio conditions.
• Inter-Frequency MLB: In multi-carrier deployments, MLB can trigger inter-frequency handovers to move UEs to a less loaded carrier layer, a capability critical for 5G NR deployments using carrier aggregation.

MLB operates in a delicate balance with Mobility Robustness Optimization (MRO), another critical SON function. Uncoordinated CIO adjustments can cause network instability.

• Conflict: An aggressive MLB-triggered CIO change to offload traffic can inadvertently cause a Too Early Handover (TEH) or a handover to an unsuitable cell, which MRO would then detect and try to correct.
• SON Coordination: A logical coordinator entity is required to arbitrate between MLB's load distribution goals and MRO's mobility reliability goals.
• Joint Optimization: Advanced implementations use a joint utility function that weighs the cost of a handover failure against the benefit of load equalization, finding a Pareto-optimal parameter setting.

The MLB algorithm can be deployed in two distinct architectural models, each with trade-offs in latency and global awareness.

• Distributed MLB: Each eNB/gNB runs its own MLB algorithm, exchanging load information with neighbors via the X2/Xn interface using the Resource Status Reporting procedure. This offers low latency but a limited, local view of the network state.
• Centralized MLB: An MLB function runs on a centralized controller, such as an O-RAN Non-RT RIC, ingesting global cell load data. This enables globally optimal load distribution but operates on a slower control loop (>1 second).
• Hybrid Approach: A common deployment uses a centralized function to set policy and target load ranges, while a distributed function executes fast, sub-second CIO adjustments.

The effectiveness of an MLB implementation is measured by a specific set of KPIs that track both load distribution and mobility integrity.

• Load Distribution Index: A Jain's fairness index applied to cell load, where a value closer to 1.0 indicates perfectly equal load distribution across the cluster.
• Call Drop Rate (CDR): Must be monitored to ensure that aggressive load balancing is not causing an increase in dropped connections due to failed handovers.
• Handover Success Rate (HOSR): The percentage of successful handover executions, which should remain above a strict threshold (e.g., 99.9%) during MLB operations.
• UE Throughput at Cell Edge: Monitored to ensure that UEs forced to handover to a less-loaded cell are not experiencing a significant degradation in their individual data rate.

The specific forecasting of Physical Resource Block (PRB) usage, the fundamental unit of time-frequency resource allocation in LTE and 5G NR networks. A PRB consists of 12 subcarriers over one slot duration. When PRB utilization on a cell approaches 100%, new users experience blocking or degraded throughput. MLB uses PRB utilization as its primary load metric because it directly reflects the cell's capacity headroom.

• Downlink PRB usage: Typically the bottleneck due to asymmetric traffic
• Uplink PRB usage: Critical for video conferencing and IoT telemetry
• Prediction horizon: 1-10 seconds for proactive MLB decisions

A per-neighbor-pair configuration parameter that biases handover measurement reports. The CIO is added to the serving cell's measurement of a specific neighbor. A positive CIO makes the neighbor appear stronger, triggering earlier handover; a negative CIO delays handover. MLB algorithms continuously adjust CIO values across cell clusters to redistribute traffic. The adjustment range is typically limited to ±24 dB to prevent coverage holes or excessive interference.

• A3 Event: Triggered when neighbor becomes an offset better than serving cell
• CIO range: -24 dB to +24 dB in 0.5 dB steps
• Per-QCI adjustment: Different offsets for different QoS Class Identifiers

A companion SON function that detects and corrects handover problems including Too-Late Handovers (TLH), Too-Early Handovers (TEH), and Handover to Wrong Cell (HWC). MRO and MLB must coordinate because aggressive MLB CIO adjustments can inadvertently cause handover failures. A conflict resolution mechanism ensures that MRO's stability requirements are not violated by MLB's load-balancing objectives.

• RLF Reports: Radio Link Failure reports used by MRO for problem detection
• Coordination: MRO can veto MLB CIO changes that risk stability
• Ping-pong detection: Identifies UEs repeatedly handed back and forth