Definition
Ratio of delivered to committed service quality across subscriber tiers during spectrum contention, measuring whether AI allocation honors the pricing hierarchy customers paid for.
The Measurement Gap
Measurement Gap and Decision
What SAEI Measures: SAEI calculates the ratio of delivered service quality to committed service quality across subscriber tiers during contention periods. It reveals whether dynamic spectrum sharing (DSS) honors tier commitments or systematically favors the dominant traffic type at the expense of premium subscribers.
Decision Supported:
Is DSS delivering on the promise of equitable service-tier treatment, or are premium subscribers being silently degraded during peak periods? Without SAEI, operators monitor average network quality and average subscriber satisfaction without examining whether specific tiers are systematically underserved when capacity is constrained.
Why This Is Unmeasured:
Network quality reporting is typically aggregated across all subscribers or segmented by technology (4G vs. 5G) rather than by subscriber tier. DSS promises to share spectrum dynamically while honoring service commitments, but no standard metric verifies this during contention, which is the only time it matters. During off-peak hours, every tier gets adequate service; the question is what happens when demand exceeds supply. Measuring SAEI requires joining per-tier quality metrics with spectrum utilization data and subscriber tier assignments, a three-way join that crosses RAN engineering, BSS, and product management boundaries (GSMA, "5G Spectrum Sharing: Implementation Lessons," 2024; 3GPP TS 28.554, "End-to-End KPI for 5G Networks," 2023).
Formula
Plain Language: During periods of spectrum contention (when total demand exceeds allocated capacity), compare the service quality each subscriber tier actually received to the quality their tier promises. A ratio of 1.0 means the commitment was met. The portfolio score uses the worst-served tier to prevent averaging from hiding systematic starvation.
Notation:
SAEI_t = Delivered_quality_t / Committed_quality_t [during contention periods only]
SAEI_portfolio = min(SAEI_t) across all tiers t
Variable Definitions:
- Delivered_quality_t = measured quality for tier t during contention periods, using throughput at the 10th percentile, latency at the 90th percentile, or session completion rate (units match the quality metric selected)
- Committed_quality_t = the quality target for tier t per the service definition in the product catalog (same units as delivered quality)
- SAEI_t = per-tier equity ratio; 1.0 = commitment met, less than 1.0 = commitment violated during contention
- SAEI_portfolio = the minimum SAEI_t across all tiers: the worst-served tier determines the portfolio score
Units: Dimensionless ratio per tier. 1.0 = commitment met. Less than 1.0 = commitment violated. Portfolio score uses the worst tier to prevent "average" from masking systematic starvation of a specific service class.
Sign Convention:
Higher is better, with 1.0 as the target. Values above 1.0 mean the tier received better quality than committed.
Edge Cases:
- Contention detection: define a utilization threshold (recommended starting point: 80% of allocated capacity) that distinguishes normal from contended periods. If the threshold is too low, SAEI includes non-contention periods and dilutes the signal.
- Mixed contention: 4G-5G contention is different from within-5G slice contention. Compute SAEI separately for each contention type.
- Minimum sample: 100 contention-period measurement intervals per tier per quarter.
Proposed formula, pending validation pilots.