Home/VoNR vs VoLTE

VoNR vs VoLTE

Summary

VoNR and VoLTE share the same IMS signalling architecture. The differences are in the radio (5G NR vs LTE), the core network (5GC vs EPC), the codec (EVS vs AMR-WB), and new capabilities like network slicing. For most users the experience improvement is primarily audio quality. For network engineers the change is in the core architecture and QoS model.

Full comparison

Feature VoLTE (4G) VoNR (5G SA) Change
Radio access LTE (E-UTRAN, eNodeB) 5G NR (NG-RAN, gNB) New radio layer
Core network EPC (MME/SGW/PGW) 5GC (AMF/SMF/UPF) New core – SA required
Signalling SIP via IMS SIP via IMS No change
Media protocol RTP/UDP RTP/UDP No change
IMS elements P-CSCF, S-CSCF, TAS, HSS Same elements No change
Voice codec (mandatory) AMR-WB (IR.92) EVS (GSMA 5G voice profile) Major quality upgrade
Audio frequency 50 Hz – 7 kHz 50 Hz – 20 kHz (fullband EVS) Near 3x wider
QoS model EPS bearers, QCI QoS flows, 5QI Architecture change
Voice QoS value QCI 1 5QI 1 Same parameters
Packet delay budget 100ms 100ms No change
Network slicing No Yes – dedicated voice slice New capability
UPF placement PGW (centralised) UPF (can be distributed/edge) Lower latency potential
SIM requirement ISIM on UICC ISIM on UICC No change
SRVCC equivalent SRVCC to 2G/3G EPS fallback to VoLTE Falls to VoLTE not 2G/3G
UK availability Live – all 4 operators Limited – 5G SA rollout in progress Deploying now

The codec difference in practice

The most perceptible improvement for end users is audio quality. AMR-WB (VoLTE HD Voice) was already a significant step over 3G narrowband. EVS fullband is another step again – comparable to a high-quality digital audio recording rather than a telephone call.

  • AMR-NB (2G/3G): 300 Hz to 3,400 Hz. Voices sound thin and telephone-like. Background noise creates significant artefacts.
  • AMR-WB (VoLTE HD Voice): 50 Hz to 7,000 Hz. Noticeably clearer. Low-frequency voice presence restored. Less background noise artefact.
  • EVS Super-Wideband: 50 Hz to 14,400 Hz. Music-like quality for voice. Consonants are sharper.
  • EVS Fullband: 20 Hz to 20,000 Hz. Full-range audio capture. The human voice is reproduced as it sounds in person.

EPS Fallback – the VoNR equivalent of SRVCC

VoLTE uses SRVCC to hand off to 2G/3G if LTE coverage is lost mid-call. VoNR uses EPS Fallback to hand off to VoLTE on the LTE bearer if 5G SA coverage is lost. From the user’s perspective the call continues; the network transfers the IMS session from the 5G NR bearer to an LTE VoLTE bearer transparently. This means VoNR deployments are resilient even where 5G SA coverage is incomplete, provided VoLTE coverage exists in the same area.