VoNR uses the 5G Core (5GC) and NG-RAN for transport, and the same IMS as VoLTE for call control. The AMF handles mobility, the SMF manages QoS flows, and the UPF forwards user-plane packets. Voice QoS uses 5QI 1 (replacing LTE’s QCI 1). SIP and RTP operate identically to VoLTE.
The 5G Core (5GC) uses a service-based architecture (SBA) where network functions communicate via HTTP/2 APIs rather than point-to-point interfaces. The key functions for VoNR are:
| Function | Full name | VoNR role |
|---|---|---|
| AMF | Access and Mobility Management Function | Handles UE registration, mobility, and N1/N2 interface to gNB. Replaces the 4G MME. |
| SMF | Session Management Function | Manages PDU sessions and QoS flows. Establishes the 5QI 1 GBR flow for VoNR voice. Replaces 4G MME/PGW session management. |
| UPF | User Plane Function | Forwards user-plane packets including RTP voice media. Replaces 4G SGW/PGW user plane. Can be distributed close to the edge for low latency. |
| PCF | Policy Control Function | Provides policy and QoS rules for sessions. Equivalent of 4G PCRF. Authorises 5QI 1 flow for VoNR calls. |
| UDM | Unified Data Management | Subscriber data repository combining 4G HSS and PCRF functions. Provides authentication and subscriber profiles. |
| AUSF | Authentication Server Function | Handles authentication of UEs registering on the 5G network. |
VoLTE uses EPS bearers with QCI identifiers. A dedicated GBR bearer with QCI 1 is established per voice call, with explicit resource reservation on the LTE radio.
VoNR replaces this with QoS flows identified by 5QI (5G QoS Indicator). The voice flow uses 5QI 1, which has the same parameters as LTE QCI 1: GBR resource type, priority level 2, 100ms packet delay budget, 10^-2 packet error rate. The conceptual model is the same; the implementation layer changes.
One of VoNR’s genuinely new capabilities is network slicing. The 5G SA architecture allows an operator to partition the network into logical slices, each with dedicated resources and QoS policies. A VoNR deployment can use a dedicated voice slice that guarantees bandwidth, latency, and packet loss targets independently of general data traffic. This is particularly relevant for enterprise deployments where audio quality SLAs are required.