Key Performance Indicators

Description

Key Performance Indicators (KPIs) in 3GPP are standardized, measurable values that demonstrate how effectively a network, network slice, or service is achieving key operational and business objectives. They are derived from measurements collected by various network elements (NEs), management systems, and probes. The architecture for KPI management involves a hierarchical data collection and processing chain: Performance Measurement (PM) data is generated by NEs (e.g., gNB, AMF, UPF) as raw counters, events, and gauges. This data is then collected by Network Management (NM) or Element Management (EM) systems via standardized interfaces (e.g., Itf-N). The management systems then apply defined formulas—often specified in 3GPP documents—to aggregate and compute the KPIs from the raw PM data.

How KPIs work is a continuous cycle of measurement, collection, computation, and analysis. For example, a fundamental Radio Access Network (RAN) KPI like the LTE Radio Resource Control (RRC) Setup Success Rate is computed by an EM system. It collects the raw counter "AttConnEstabAtt" (number of RRC connection establishment attempts) and "AttConnEstabSucc" (number of successful attempts) from the eNodeB over a defined measurement period (e.g., 15 minutes). The KPI is then calculated as (Successful Attempts / Total Attempts) * 100%. This computed KPI value is then stored, visualized, and compared against target thresholds to assess performance. KPIs span all network domains: Accessibility (e.g., call setup success rate), Retainability (e.g., drop call rate), Integrity (e.g., throughput, latency), Availability (e.g., node uptime), and Mobility (e.g., handover success rate).

Key components in the KPI ecosystem include the Performance Measurement Jobs configured in network elements, the Management Data Analytics (MDA) functions that can process KPIs for automation, and the standardized definitions and formulas provided in 3GPP Technical Specifications (TS) and Technical Reports (TR). Their role is multifaceted: for engineers, they are diagnostic tools for root cause analysis; for planners, they guide capacity expansion; for operators, they ensure Service Level Agreement (SLA) compliance, especially for network slicing where slice-specific KPIs are crucial. The description and calculation methods for hundreds of KPIs are detailed across numerous 3GPP specs, including service requirements (22-series), management (28-series, 32-series), and performance aspects (38.8xx series for NR).

Purpose & Motivation

KPIs exist to provide an objective, quantitative basis for assessing network performance and service quality. The fundamental problem they solve is the need to move from subjective, anecdotal assessments of network health to data-driven decision making. In the early days of mobile networks, operators lacked standardized metrics, making benchmarking and interoperability difficult. KPIs provide a common language for defining what "good performance" means for aspects like voice call quality, data speed, and network reliability.

The creation and continuous evolution of KPIs in 3GPP are motivated by several factors: the introduction of new technologies (e.g., 5G NR, network slicing), new services (e.g., Ultra-Reliable Low Latency Communication - URLLC), and the need for automated network management. Each new release introduces capabilities that require new KPIs to measure them effectively. For instance, 5G introduced KPIs for beam management, network slice performance, and energy efficiency. They address the limitations of previous approaches by providing increasingly granular, service-aware, and real-time capable metrics. This allows operators to not only monitor the network's technical performance but also to correlate it with the end-user's Quality of Experience (QoE), enabling proactive optimization and efficient resource utilization.

Key Features

• Standardized definitions and calculation formulas across vendors
• Span all network domains: RAN, Core, Transport, and End-to-End
• Support for traditional network metrics and new 5G/service-aware metrics
• Enable performance benchmarking, SLA monitoring, and root cause analysis
• Foundation for closed-loop automation and Self-Organizing Networks (SON)
• Critical for Network Slice performance assurance and management

Evolution Across Releases

Defining Specifications