Self-Healing Masking Rate (SHMR)

GMI-TEL-004TelecommunicationsPilot-ReadyPattern O-10: Deferral Compounding

Definition

Percentage of autonomous self-healing events where the same or substantially similar fault condition recurs within a defined window on the same network element, indicating symptom suppression rather than resolution.

The Measurement Gap

Decision Supported: Is network self-healing genuinely resolving faults, or is it creating a recurring suppression cycle that degrades infrastructure while every NOC metric says the network is healthy?

Measurement Gap:

Self-healing platforms report "events resolved", but resolved and fixed are not the same thing. A protection switch that reroutes traffic around a degrading fiber is a successful self-healing event in every dashboard, but the fiber is still degrading. The problem is that self-healing success metrics count the automation's response, not the outcome's durability. Without recurrence tracking, nobody knows how many "resolved" events are actually the same fault cycling through automated remediation repeatedly. The network looks green while its resilience margin gets consumed one self-healing event at a time. SHMR makes the masking rate visible so operations can distinguish genuine healing from automated symptom suppression (TMForum, 2023; Ericsson Network Analytics, 2024).

Formula

Formula (Plain Language): Count all self-healing events. Count how many were followed by the same fault signature on the same network element within N days. Divide.

Notation:

SHMR = Self_healing_events_with_recurrence / Total_self_healing_events × 100

Variable Definitions:

  • Self_healing_events_with_recurrence = autonomous healing events where the same fault alarm class fires again on the same network element within N days (count, dimensionless)
  • Total_self_healing_events = all autonomous self-healing events in the measurement period (count, dimensionless)
  • N = recurrence window (proposed starting point: 7 days for transport network elements, 3 days for RAN elements; adjust based on the self-healing platform's typical remediation cycle time)

Units:

Percentage. 0% means every self-healing action was durable: no recurrence. 100% would mean every healing event recurred.

Sign Convention:

Lower is better. Rising SHMR over time indicates accumulating unresolved root causes behind the self-healing automation.

Edge Cases:

  • Intermittent faults (e.g., temperature-dependent fiber degradation that recurs daily): legitimate recurrence that is not masking in the traditional sense. Segment SHMR by fault type and flag known intermittent categories separately from true masking.
  • Self-healing action type matters: traffic reroute versus equipment restart have different expected recurrence profiles. Compute SHMR per action category to avoid mixing fundamentally different remediation types.
  • Low event volume on individual elements: aggregate to cluster or ring level for statistical meaning. Minimum sample: 100 self-healing events per measurement period across the measured population.

Proposed formula, pending validation pilots.

At a Glance

Unit
Percentage
Trend Direction
Lower
Reporting Frequency
Monthly
Category
Self-Healing Reliability and Resilience
← Redundancy Budget Depletion Index (RBDI)SLA Adjustment Proportionality Score (SAPS) →

Assess Your Organization

The AI-GPS Assessment evaluates your AI governance maturity and recommends KPIs like this one for your implementation roadmap.

Take the Free AI-GPS Assessment