O1 Interface

The O1 interface provides a comprehensive implementation of the FCAPS model (Fault, Configuration, Accounting, Performance, Security) standardized by ITU-T. It serves as the primary conduit for:

• Fault Management: Streaming real-time alarms and event notifications from O-RU, O-DU, and O-CU to the SMO
• Configuration Management: NETCONF/YANG-based provisioning of network function parameters
• Performance Management: Bulk PM data collection using streaming telemetry protocols
• Security Management: Certificate lifecycle and access control policy distribution

The O1 interface uses NETCONF as its primary configuration protocol with YANG as the data modeling language. This enables:

• Vendor-agnostic provisioning: Standardized YANG modules define configuration schemas for all O-RAN network functions
• Transaction-based changes: Atomic configuration commits prevent partial or inconsistent network states
• Model-driven telemetry: YANG-defined subscription models allow the SMO to dynamically subscribe to specific performance counters
• Backward compatibility: Versioned YANG modules ensure smooth upgrades across multi-vendor environments

The VNF Event Stream (VES) protocol, standardized by ONAP and adopted by O-RAN, is the primary mechanism for real-time fault and performance event delivery over the O1 interface. Key characteristics include:

• JSON-structured events: Standardized common event format with domain-specific extensions for RAN
• High-frequency telemetry: Supports sub-second streaming of performance measurements for closed-loop automation
• Domain-specific stanzas: Dedicated fields for heartbeat, fault, measurement, and state change notifications
• Kafka-based transport: Leverages distributed messaging for scalable, reliable event delivery to the SMO data lake

The O1 interface mandates mutual TLS (mTLS) for all management traffic between the SMO and network functions. Security provisions include:

• Certificate-based authentication: Both SMO and network functions must present valid X.509 certificates
• Encrypted transport: All FCAPS data, including configuration payloads and performance telemetry, is encrypted in transit
• Role-based access control: NETCONF access control models restrict which SMO users can modify specific YANG subtrees
• Audit logging: All configuration changes are logged with timestamps and user attribution for compliance

For bulk performance data collection, the O1 interface supports file-based PM streaming as an alternative to real-time VES events. This mechanism is optimized for:

• High-volume counters: 15-minute and 5-minute granularity PM bins from hundreds of cells
• SFTP/FTP secure transfer: Encrypted file push from network functions to SMO collectors
• XML and JSON formats: Standardized 3GPP TS 32.435 performance data structures
• Scheduled collection: Configurable collection intervals to balance granularity with transport overhead
• Bulk processing efficiency: Enables the SMO to batch-process historical PM data for AI/ML training pipelines

Within the SMO framework, the O1 interface terminates at the Data Collection and Distribution Framework. This termination point provides:

• Protocol mediation: Translates between NETCONF, VES, and RESTCONF for internal SMO consumers
• Data filtering and routing: Distributes filtered event streams to registered rApps and analytics engines
• Schema validation: Validates incoming YANG data against O-RAN standard models before processing
• High-availability design: Clustered termination points ensure no single point of failure in the management plane

The SMO is the central management entity that terminates the O1 interface. It provides unified orchestration, administration, and lifecycle management of O-RAN network functions. The SMO hosts the Non-RT RIC and integrates with existing OSS/BSS systems. Through the O1 interface, the SMO pushes configuration changes, collects performance data, and manages faults across all managed elements.

The O1 interface relies on NETCONF (Network Configuration Protocol) and RESTCONF as its transport protocols, with YANG (Yet Another Next Generation) as the data modeling language. YANG models define the hierarchical structure of configuration and operational data for O-RAN network functions. The O-RAN Alliance standardizes YANG modules for the O1 interface to ensure multi-vendor interoperability, covering areas like RU, DU, and CU configuration.

VNF Event Stream is a critical data collection protocol operating over the O1 interface. It enables network functions to publish real-time performance, fault, and event data to the SMO. VES uses a RESTful API to push JSON-formatted event batches, supporting domains like heartbeat, fault, measurement, and stateChange. This streaming telemetry is essential for the AI/ML analytics engines within the Non-RT RIC.

The O-CU is a logical node hosting the RRC, SDAP, and PDCP protocols. It is a managed element directly connected to the SMO via the O1 interface. Through O1, the SMO performs lifecycle management, configuration of cell parameters, and collection of PM (Performance Measurement) counters. The O-CU also terminates the E2 interface for near-real-time control by the Near-RT RIC.

The O-DU hosts the RLC and MAC layers and is managed via the O1 interface. The SMO uses O1 to configure cell-level parameters, manage software upgrades, and retrieve real-time KPIs related to scheduling and HARQ performance. The O-DU's tight latency requirements make its fault management via O1 critical for maintaining network stability and user throughput.

The O1 interface extends to the O-Cloud, the cloud infrastructure hosting virtualized network functions. It manages the lifecycle of O-Cloud Resources and receives notifications about infrastructure faults or capacity changes. This allows the SMO to correlate application-layer issues from the O-CU/O-DU with underlying hardware or virtualization layer problems, enabling true root-cause analysis.