Volte Service Description and Implementation Guidelines 26 March 2014

GSM Association Non-confidential
Official Document FCM.01 - VoLTE Service Description and Implementation Guidelines
VoLTE Service Description and Implementation Guidelines

Version 1.1
26 March 2014
This is a Non-binding Permanent Reference Document of the GSMA
Security Classification: Non-confidential

Access to and distribution of this document is restricted to the persons permitted by the security classification. This document is confidential to the
Association and is subject to copyright protection. This document is to be used only for the purposes for which it has been supplied and
information contained in it must not be disclosed or in any other way made available, in whole or in part, to persons other than those permitted
under the security classification without the prior written approval of the Association.
Copyright Notice
Copyright © 2014 GSM Association
Disclaimer
The GSM Association (“Association”) makes no representation, warranty or undertaking (express or implied) with respect to and does not accept
any responsibility for, and hereby disclaims liability for the accuracy or completeness or timeliness of the information contained in this document.
The information contained in this document may be subject to change without prior notice.
Antitrust Notice
The information contain herein is in full compliance with the GSM Association’s antitrust compliance policy.
V1.1 Page 1 of 121

GSM Association Non-Confidential
Official Document
Table of Contents
1 Introduction 6
1.1 Overview 6
1.2 Relationship to Standards 6
1.3 Scope 7
1.4 Definition of Terms 7
1.5 Document Cross-References 12
2 VoLTE Architecture 16
2.1 VoLTE Functional Node Description 17
2.1.1 VoLTE UE (User Equipment) 17
2.1.2 Evolved Universal Terrestrial Access Network (E-UTRAN) 17
2.1.3 Evolved Packet Core 17
2.1.4 IMS 18
2.1.5 Additional Network Functionality 19
2.2 VoLTE Interface Description 20
2.2.1 LTE-Uu Interface (UE – eNodeB) 20
2.2.2 S1-MME Interface (UE – MME) 20
2.2.3 S1AP Interface (eNodeB – MME) 20
2.2.4 S1-U Interface (eNodeB – SGW) 20
2.2.5 X2 Interface (eNodeB – eNodeB) 20
2.2.6 S5 Interface (SGW – PGW) 20
2.2.7 S6a Interface (HSS – MME) 20
2.2.8 S9 Interface (H-PCRF – V-PCRF) 20
2.2.9 S10 Interface (MME – MME) 21
2.2.10 S11 Interface (MME – SGW) 21
2.2.11 Gx Interface (PCRF – PGW) 21
2.2.12 Rx Interface (PCRF – P-CSCF) 21
2.2.13 SGi Interface (PGW – P-CSCF) 21
2.2.14 Cx Interface (I/S-CSCF – HSS) 21
2.2.15 Sh Interface (VoLTE AS – HSS) 21
2.2.16 Gm Interface (UE – P-CSCF) 21
2.2.17 Ut Interface (UE – VoLTE AS) 21
2.2.18 Mx Interface (x-CSCF – IBCF) 22
2.2.19 Mw Interface (x-CSCF – x-CSCF) 22
2.2.20 Mg Interface (xCSCF – MGCF) 22
2.2.21 Mi Interface (xCSCF – BGCF) 22
2.2.22 Mj Interface (BGCF – MGCF) 22
2.2.23 ISC Interface (S-CSCF –TAS) 22
2.2.24 Mr Interface (S-CSCF – MRF) 22
2.2.25 Mr’ Interface (TAS – MRF) 22
2.2.26 Cr Interface (TAS – MRF) 22
2.2.27 Mb Interface (media bearer) 22
2.2.28 Ici Interface (IBCF – IBCF) 23
V1.0
Official Document
2.2.29 Izi Interface (TrGW – TrGW) 23

2.3 Related GSMA Permanent Reference Documents 23
3 VoLTE Implementation - Single PMN 25
3.1 General 25
3.2 VoLTE Basic Call Flows 25
3.2.1 VoLTE UE Attachment and IMS Registration 26
3.2.2 VoLTE UE Initiated Detach and IMS Deregistration 32
3.2.3 Basic VoLTE UE to VoLTE UE Call Establishment – Originating Side 35
3.2.4 Basic VoLTE UE to VoLTE UE Call Establishment – Terminating Side 39
3.2.5 Basic VoLTE UE to VoLTE UE Call Clearing - Initiated 43
3.2.6 Basic VoLTE UE to VoLTE UE Call Clearing - Received 45
3.3 VoLTE-CS Interworking 47
3.3.1 Basic VoLTE UE to CS Call Establishment – Originating Side 47
3.3.2 Basic VoLTE UE to CS Call Establishment – Terminating Side 50
3.3.3 Basic VoLTE UE to CS Call Clearing - Initiated 53
3.3.4 Basic VoLTE UE to CS Call Clearing - Received 54
3.4 Supplementary Services 55
3.4.1 General 55
3.5 ENUM/DNS 57
3.5.1 General 57
3.5.2 Number Portability 57
3.5.3 IP Service Routing 57
3.5.4 Number Resolution 57
3.5.5 ENUM 57
3.6 Diameter Signalling 60
3.6.1 General 60
3.6.2 Diameter Agents 60
3.6.3 Diameter Transport 60
3.6.4 Diameter Peer Discovery 60
3.6.5 Diameter Capability Exchange 60
3.6.6 Diameter Routing 61
3.7 Traffic Management and Policy 61
3.7.1 General 61
3.7.2 Policy and Charging Control 61
3.7.3 DiffServ 63
3.7.4 Mapping between QCI and DiffServ 63
3.8 Session Border Controllers 64
3.9 Emergency Call 64
3.10 Lawful Intercept 64
3.11 Security 64
3.11.1 General 64
3.11.2 Security Gateway 65
3.11.3 IMS Media Plane Security 65
3.12 SMS over IP 65
V1.0
Official Document
3.13 Support of Legacy Proprietary CS Services 65

3.14 Complementing VoLTE with 2G/3G Voice 66
3.14.1 SRVCC 66
3.14.2 PS Handover 66
3.14.3 IMS Service Centralization and Continuity 66
3.15 Charging 66
3.16 Codecs 67
3.17 IP Version & Transport 67
3.18 Home eNodeB (HeNB) 67
4 VoLTE Implementation - Interconnect 69
4.1 General 69
4.2 VoLTE Interconnect 69
4.2.1 Basic VoLTE UE to Peer IMS Call Establishment – Originating Side 70
4.2.2 Basic VoLTE UE to Peer IMS Call Establishment – Terminating Side 73
4.2.3 Basic VoLTE UE to Peer IMS Call Teardown - Initiated 76
4.2.4 Basic VoLTE UE to Peer IMS Call Teardown - Received 78
4.3 Bi-lateral Interconnect 80
4.3.1 Physical Configuration of Bi-lateral Interconnect 80
4.3.2 Usage of ENUM/DNS 80
4.3.3 Usage of Session Border Controllers 81
4.4 IPX-Based Interconnect 81
4.4.1 Configuration of IPX-based Interconnect 81
4.4.2 Usage of ENUM/DNS 82
4.4.3 Usage of Session Border Controllers 82
4.5 CS Interconnect 82
4.6 Charging 82
5 VoLTE Implementation - Roaming 83
5.1 General 83
5.2 VoLTE Roaming Basic Call Flows 84
5.2.1 Roaming VoLTE UE Attachment and IMS Registration 84
5.2.2 Roaming VoLTE UE Initiated Detach and IMS Deregistration 90
5.2.3 Roaming VoLTE UE to VoLTE Call Establishment – Originating Side 92
5.2.4 Roaming VoLTE UE to VoLTE UE Call Establishment – Terminating
Side 96
5.2.5 Roaming VoLTE UE to VoLTE UE Call Clearing - Initiated 100
5.2.6 Roaming VoLTE UE to VoLTE Call Clearing - Received 102
5.3 Roaming Architecture for Voice over IMS with Local break-out (RAVEL) 104
5.4 Optimal Media Routing 105
5.5 Diameter Signalling 105
5.6 Traffic Management and Policy 106
5.7 Session Border Controllers 106
5.8 IMS Emergency Call 106
5.9 Lawful Intercept 106
5.10 Security 106
V1.0
Official Document
5.11 Charging 106

6 Implementation Guidelines 107
6.1.1 Open Implementation Issues 107
6.1.2 VoLTE Device Implementation Guidelines 107
6.1.3 LTE/EPC Implementation Guidelines 109
6.1.4 VoLTE IMS Implementation Guidelines 111
6.1.5 Other Guidelines 116
Document Management 121
Document History 121
V1.0
Official Document
1 Introduction
1.1 Overview
Voice over LTE, or VoLTE is a GSMA profile of the standards definition for the delivery of
services currently provided via Circuit Switch networks - mainly voice and SMS - over the
Packet Switched only network of LTE, leveraging the core network IP Multimedia Sub-
System (IMS). When mobile networks deploy LTE radio access technology, conformity to
the VoLTE profile provides operators with assurance of interworking between their LTE
network and the devices that connect to it, as well as providing for the expected user
experience of voice Multi-Media Telephony service and SMS. In combination with Policy
Control, IMS provides for the required QoS appropriate for voice service using LTE radio
access technology, thereby providing the user experience of voice calls that subscribers
expect. Moreover, VoLTE is designed to fully integrate with the existing user experience that
is currently implemented with circuit switched voice devices, and therefore whether the call
is a circuit switched call or a VoLTE call is transparent to the end user (including when
moving in and out of LTE coverage) and is dependent only on which radio access
technology to which the user is attached. At the same time, using new, wideband codecs
can provide higher voice quality and enhance the user experience.
VoLTE is in accordance with 3GPP specifications and additional profiling is defined within
GSMA Permanent Reference Documents.
GSMA PRD IR.92 [54] defines the UNI for IMS voice and SMS. It defines a profile that
identifies a minimum mandatory set of features which are defined in 3GPP specifications
that a wireless device (UE) and network are required to implement in order to guarantee an
interoperable, high quality IMS-based telephony service over LTE.
The NNI for VoLTE is defined in the IMS Roaming & Interworking Guidelines GSMA PRD
IR.65 [51].
VoLTE Roaming is defined in the LTE Roaming Guidelines GSMA PRD IR.88 [53].
This document defines the VoLTE service description and implementation guidelines in
order to provide an end-to-end VoLTE deployment.
Note that in this version of the document, CSFB and SRVCC are not in scope.
1.2 Relationship to Standards

VoLTE is based on publically available and published 3GPP specifications as listed in
Section ‎1.5. 3GPP Release 8, the first release supporting LTE, is taken as a basis for the
VoLTE profile. It should be noted, however that not all the features mandatory in 3GPP
Release 8 are required for compliance with VoLTE.
Conversely, some features required for compliance with VoLTE are based on functionality
defined in 3GPP Release 9 or higher releases.
Unless otherwise stated, the latest version of the referenced specifications for the relevant
3GPP release applies.
Detailed information related to the support of specific 3GPP functionality per release is
defined within GSMA PRD IR.92 [54].
V1.0
Official Document
1.3 Scope
This document is separated into 5 main sections.
 Section 2 VoLTE Architecture:- defines the logical architecture, functional node

description, and interfaces required for VoLTE deployment.
 Section 3 VoLTE Implementation – Single PMN:- defines what is required to deploy
VoLTE within a single MNO's domain; detailing call flows, supplementary services
required, Diameter configuration aspects, traffic management and QoS aspects,
security, etc.
 Section 4 VoLTE Implementation – Interconnect:- defines what is required in
addition to an Intra-PMN deployment, for interconnecting two MNO's VoLTE
deployments; detailing call flows, bi-lateral interconnect, IPX interconnect,
ENUM/DNS, usage of Session Border Controllers, etc.
 Section 5 VoLTE Implementation – Roaming:- defines what is required in addition to
an Intra-PMN deployment, for allowing subscribers to roam between two MNO's
VoLTE deployments; detailing call flows, Local-BreakOut (LBO), Impacts to
Diameter routing, policy and QoS aspects, etc.
 Section 6 Implementation Guidelines:- highlights the issues discovered during
VoLTE Interoperability testing (IOT) and in commercial operator deployments.
1.4 Definition of Terms
Term Description
3GPP 3rd Generation Partnership Project
A-SBC Access Session Border Controller
ACR Anonymous Call Rejection
AMBR Aggregate Maximum Bit Rate
AMR Adaptive Multi-Rate
AMR-WB Adaptive Multi-Rate Wideband
API Application Programming Interface
APN Access Point Name
ARP Allocation and Retention Priority
AS Application Server
AUTN Authentication Token
AVP Attribute Value Pair
BGCF Border Gateway Control Function
BICC Bearer Independent Call Control
Customised Application for Mobile network
CAMEL
Enhanced Logic
CDIV Communication Diversion
CDR Charging Data Record
CN Core Network
CONF Conferencing
V1.0
Official Document
CS Circuit Switched
CSCF Call Server Control Function
CSFB Circuit Switched Fall Back
CW Call Waiting
DEA Diameter Edge Agent
DiffServ Differentiated Services
DL DownLink
DNS Domain Name System
DPI Deep Packet Inspection
DRA Diameter Relay Agent
DRX Discontinuous Reception
DSCP DiffServ Code Point
ECGI E-UTRAN Cell Global Identifier
e2ae end to access edge
e2e end to end
eKSI E-UTRAN Key Set Identifier
ENUM E.164 Number Mapping
EPC Evolved Packet Core
EPS Evolved Packet System
ERAB E-UTRAN Radio Access Bearer
ESM EPS Session Management
eSRVCC Enhanced Single Radio Voice Call Continuity
European Telecommunications Standards
ETSI
Institute
E-UTRAN Evolved Universal Terrestrial Access Network
FDD Frequency Division Duplex
GAA Generic Authentication Architecture
GBA Generic Bootstrapping Architecture
GBR Guaranteed Bit Rate
GPRS General Packet Radio Service
GRX GPRS Roaming eXchange
GSM Global System for Mobile communications
GTP GPRS Tunneling Protocol
HLR Home Location Register
HPMN Home Public Mobile Network
HSPA High Speed Packet Access
HSS Home Subscriber Server
HTTP Hypertext Transfer Protocol
I-CSCF Interrogating Call Session Control Function
V1.0
Official Document
I-SBC Interconnect Session Border Controller

IBCF Interconnection Border Control Function
icid IM CN subsystem charging identifier
ICS IMS Centralised Services
ICSI IMS Communication Service Identifier
IETF Internet Engineering Task Force
iFC Initial Filter Criteria
IM IP Multimedia
IM-GW IP Media Gateway
IMEI International Mobile Equipment Identity
IMS IP Multimedia Subsystem
IMS-AKA IMS Authentication and Key Agreement
IMS-AGW IMS Access Gateway
IMS-ALG IMS Application Level Gateway
IMSI International Mobile Subscriber Identity
IOT Interoperability Testing
IP Internet Protocol
IP-CAN IP-Connectivity Access Network
IPsec IP Security
IPX IP Packet Exchange
ISIM IM Services Identity Module
ISUP ISDN User Part
LBO Local Breakout
LTE Long Term Evolution
MAC Medium Access Control
MBR Maximum Bit Rate
MCC Mobile Country Code
ME Mobile Equipment
MGCF Media Gateway Control Function
MME Mobility Management Entity
MMS Multimedia Messaging Service
MMTel Multimedia Telephony
MNC Mobile Network Code
MNO Mobile Network Operator
MRF Media Resource Function
MSISDN Mobile Subscriber ISDN Number
MSRP Message Session Relay Protocol
MTU Maximum Transmission Unit
V1.0
Official Document
MWI Message Waiting Indicator

NAPTR Name Authority Pointer
NAS Non-Access Stratum
NAT Network Address Translation
NNI Network to Network Interface
OIP Originating Identification Presentation
Originating
OIR
Identification Restriction
P-CSCF Proxy Call Session Control Function
PCC Policy and Charging Control
PCEF Policy and Charging Enforcement Function
PCO Protocol Configuration Options
PCRF Policy Charging and Rules Function
PDN Packet Data Network
PGW Packet Data Network Gateway
PHB Per Hob Behaviour
PLMN Public Land Mobile Network
PMN Public Mobile Network
PS Packet Switched
QCI QoS Class Identifier
QoS Quality of Service
RAN Radio Access Network
RAND RANDom number (used for authentication)
RAT Radio Access Technology
RES user RESponse (used in IMS-AKA)
RLC Radio Link Control
RRC Radio Resource Control
RTCP RTP Control Protocol
RTP Real-time Transport Protocol
S-CSCF Serving Call Session Control Function
SAE System Architecture Evolution
SBC Session Border Controller
SCTP Stream Control Transmission Protocol
SDP Session Description Protocol
SEG Security Gateway
SGW Serving Gateway
SGSN Serving GPRS Support Node
SIGCOMP Signalling Compression
V1.0
Official Document
SIP Session Initiation Protocol

SIP-I SIP with encapsulated ISUP
SMS Short Message Service
SON Self-Organising Networks
SRTP Secure RTP
SRVCC Single Radio Voice Call Continuity
TAS Telephony Application Server
TAI Tracking Area Identity
TCP Transmission Control Protocol
TDD Time Division Duplex
TDM Time Division Multiplexing
TEID Tunnel End Point Identifier
TFT Traffic Flow Template
THP Traffic Handling Priority
TLS Transport Layer Security
TIP Terminating Identification Presentation
TIR Terminating Identification Restriction
TrGW Transition Gateway
TTM Time To Market
UDC User Data Convergence
UDP User Datagram Protocol
UDR User Data Repository
UE User Equipment
UICC Universal Integrated Circuit Card
UL Uplink
ULI User Location Information
UMTS Universal Mobile Telecommunications System
UNI User to Network Interface
URN Uniform Resource Name
URI Uniform Resource Identifier
USIM Universal Subscriber Identity Module
VLR Visitor Location Register
VoHSPA Voice over HSPA
VoLTE Voice over LTE
VPMN Visited Public Mobile Network
XCAP XML Configuration Access Protocol
XML Extensible Markup Language
XRES eXpected user RESponse (used in IMS-AKA)
V1.0
Official Document
1.5 Document Cross-References
Document
Ref Number Title
[1] 3GPP TS 23.002 Network Architecture
[2] 3GPP TS 23.003 Numbering, addressing and identification
General Packet Radio Service (GPRS); Service description; Stage
[3] 3GPP TS 23.060
2
[4] 3GPP TS 23.203 Policy and charging control architecture
[5] 3GPP TS 23.228 IP Multimedia Subsystem (IMS); Stage 2
General Packet Radio Service (GPRS) enhancements for Evolved
[6] 3GPP TS 23.401
Universal Terrestrial Radio Access Network (E-UTRAN) access
Conferencing using the IP Multimedia (IM) Core Network (CN)
[7] 3GPP TS 24.147
subsystem; Stage 3
IMS Multimedia telephony communication service and
[8] 3GPP TS 24.173
supplementary services; Stage 3
IP multimedia call control protocol based on Session Initiation
[9] 3GPP TS 24.229
Protocol (SIP) and Session Description Protocol (SDP); Stage 3
Non-Access-Stratum (NAS) protocol for Evolved Packet System
[10] 3GPP TS 24.301
(EPS); Stage 3
Messaging service using the IP Multimedia (IM) Core Network
[11] 3GPP TS 24.247
(CN) subsystem; Stage 3
[12] 3GPP TS 24.341 Support of SMS over IP networks; Stage 3
Communication Diversion (CDIV) using IP Multimedia (IM) Core
[13] 3GPP TS 24.604
Network (CN) subsystem; Protocol specification
Conference (CONF) using IP Multimedia (IM) Core Network (CN)
[14] 3GPP TS 24.605
subsystem; Protocol specification
Message Waiting Indication (MWI) using IP Multimedia (IM) Core
[15] 3GPP TS 24.606
Originating Identification Presentation (OIP) and Originating
[16] 3GPP TS 24.607 Identification Restriction (OIR) using IP Multimedia (IM) Core
Terminating Identification Presentation (TIP) and Terminating
[17] 3GPP TS 24.608 Identification Restriction (TIR) using IP Multimedia (IM) Core
Communication HOLD (HOLD) using IP Multimedia (IM) Core
[18] 3GPP TS 24.610
Anonymous Communication Rejection (ACR) and Communication
[19] 3GPP TS 24.611 Barring (CB) using IP Multimedia (IM) Core Network (CN)
subsystem; Protocol specification
Communication Waiting (CW) using IP Multimedia (IM) Core
[20] 3GPP TS 24.615
Network (CN) subsystem; Protocol Specification
V1.0
Official Document
Extensible Markup Language (XML) Configuration Access

[21] 3GPP TS 24.623 Protocol (XCAP) over the Ut interface for Manipulating
Supplementary Services
Interworking between the Public Land Mobile Network (PLMN)
[22] 3GPP TS 29.061 supporting packet based services and Packet Data Networks
(PDN)
Interworking between the IP Multimedia (IM) Core Network (CN)
[23] 3GPP TS 29.163
subsystem and Circuit Switched (CS) networks
[24] 3GPP TS 29.165 Inter-IMS Network to Network Interface (NNI)
[25] 3GPP TS 29.212 Policy and Charging Control (PCC); Reference points
[26] 3GPP TS 29.214 Policy and charging control over Rx reference point
Policy and Charging Control (PCC) over S9 reference point; Stage
[27] 3GPP TS 29.215
3
IP Multimedia (IM) Subsystem Cx and Dx Interfaces; Signalling
[28] 3GPP TS 29.228
flows and message contents
Cx and Dx interfaces based on the Diameter protocol; Protocol
[29] 3GPP TS 29.229
details
Evolved Packet System (EPS); Mobility Management Entity
[30] 3GPP TS 29.272 (MME) and Serving GPRS Support Node (SGSN) related
interfaces based on Diameter protocol
3GPP Evolved Packet System (EPS); Evolved General Packet
[31] 3GPP TS 29.274 Radio Service (GPRS) Tunnelling Protocol for Control plane
(GTPv2-C); Stage 3
General Packet Radio System (GPRS) Tunnelling Protocol User
[32] 3GPP TS 29.281
Plane (GTPv1-U)
IP Multimedia (IM) Subsystem Sh interface; Signalling flows and
[33] 3GPP TS 29.328
message contents
[34] 3GPP TS 29.329 Sh interface based on the Diameter protocol; Protocol details
Media Gateway Control Function (MGCF) - IM Media Gateway;
[35] 3GPP TS 29.332
Mn interface
[36] 3GPP TR 29.809 Study on Diameter overload control mechanisms
Characteristics of the IP Multimedia Services Identity Module
[37] 3GPP TS 31.103
(ISIM) application
Telecommunication management; Charging management;
[38] 3GPP TS 32.240
Charging architecture and principles
Telecommunication management; Charging management; IP
[39] 3GPP TS 32.260
Multimedia Subsystem (IMS) charging
Telecommunication management; Charging management;
[40] 3GPP TS 32.298
Charging Data Record (CDR) parameter description
[41] 3GPP TS 33.107 3G security; Lawful interception architecture and functions
[42] 3GPP TS 33.328 IP Multimedia Subsystem (IMS) media plane security
[43] 3GPP TS 33.401 3GPP System Architecture Evolution (SAE); Security architecture
Evolved Universal Terrestrial Radio Access (E-UTRA) and
[44] 3GPP TS 36.300
Evolved Universal Terrestrial Radio Access Network (E-UTRAN);
V1.0
Official Document
Overall description; Stage 2

[45] 3GPP TS 36.413
S1 Application Protocol (S1AP)
[46] 3GPP TS 36.423
X2 Application Protocol (X2AP)
Telecommunications and Internet converged Services and
Protocols for Advanced Networking (TISPAN);
[47] ETSI TS 183 038 PSTN/ISDN Simulation Services;
Extensible Markup Language (XML) Document Management;
Protocol Specification
[48] GSMA PRD IR.34 Inter-Service Provider IP Backbone Guidelines
[49] GSMA PRD IR.58 IMS Profile for Voice over HSPA
[50] GSMA PRD IR.64 IMS Service Centralization and Continuity Guidelines
[51] GSMA PRD IR.65 IMS Roaming and Interworking Guidelines
DNS/ENUM Guidelines for Service Providers & GRX/IPX
[52] GSMA PRD IR.67
Providers
[53] GSMA PRD IR.88 LTE Roaming Guidelines
[54] GSMA PRD IR.92 IMS Profile for Voice and SMS
[55] IETF RFC 768 User Datagram Protocol
[56] IETF RFC 2246 The TLS Protocol Version 1.0
[57] IETF RFC 2617
HTTP Authentication: Basic and Digest Access Authentication
[58] IETF RFC 3550
RTP: A Transport Protocol for Real-Time Applications
[59] IETF RFC 3588
Diameter Base Protocol
[60] IETF RFC 3711 The Secure Real-time Transport Protocol (SRTP)
RTP Payload Format and File Storage Format for the Adaptive
[61] IETF RFC 4867 Multi-Rate (AMR) and Adaptive Multi-Rate Wideband (AMR-WB)
Audio Codecs
[62] IETF RFC 4961
Symmetric RTP / RTP Control Protocol (RTCP)
Private Header (P-Header) Extension to the Session Initiation
[63] IETF RFC 5009
Protocol (SIP) for Authorization of Early Media
Design Considerations for Session Initiation Protocol (SIP)
[64] IETF RFC 6357
Overload Control
Interworking between SIP-I based circuit-switched core network
[65] 3GPP TS 29.235
and other networks
[66] 3GPP TS 23.205 Bearer-independent circuit-switched core network; Stage 2
[67] 3GPP TS 23.231 SIP-I based circuit-switched core network; Stage 2
[68] IETF RFC 3966 The tel URI for Telephone Numbers
[69] IETF RFC 3261 SIP: Session Initiation Protocol
[70] IETF RFC 3312 Integration of Resource Management and Session Initiation
Protocol (SIP)
[71] IETF RFC 3262 Reliability of Provisional Responses in the Session Initiation
Protocol (SIP)
V1.0
Official Document
[72] 3GPP TS 29.238 Interconnection Border Control Function (IBCF) – Transition

Gateway (TrGW) interface, Ix interface; Stage 3.
[73] 3GPP TS 29.334 IMS Application Level Gateway (ALG) – IMS Access Gateway
(IMS-AGW); Iq interface; Stage 3.
[74] GSMA PRD AA.80 IP Packet eXchange Service Agreement
[75] 3GPP TS 23.335 User Data Convergence (UDC); Technical Realization and
Information Flows; Stage 2
[76] 3GPP TS 29.079 Optimal Media Routing within the IP Multimedia System (IMS);
Stage 3
[77] IETF RFC 5031 A Uniform Resource Name (URN) for Emergency and other Well-
Known Services
[78] 3GPP TS 23.167 IP Multimedia Subsystem (IMS) Emergency Sessions
[79] 3GPP TS 29.213 Policy and Charging Control signalling flows and Quality of
Service (QoS) Parameter Mapping.
[80] IETF RFC 4028 Session Timers in the Session Initiation Protocol (SIP)
V1.0
Official Document
2 VoLTE Architecture
The VoLTE logical architecture is based on the 3GPP defined architecture and principles for
VoLTE UE, Long Term Evolution (LTE), Evolved Packet Core network (EPC), and the IMS
Core Network. It consists of the following:-
 VoLTE UE: The VoLTE UE contains functionality to access the LTE RAN and the
EPC to allow mobile broadband connectivity. An embedded IMS stack and VoLTE
IMS application are required to access VoLTE services.
 Radio Access Network. The Evolved Universal Terrestrial Radio Access Network
(E-UTRAN); this is often referred to as Long Term Evolution (LTE). LTE radio
capabilities for FDD LTE only, TDD LTE only, or both FDD and TDD LTE are
applicable for VoLTE.
 Core Network. The Evolved Packet Core (EPC).
 IMS Core Network. The IMS Core Network within the VoLTE architecture provides
the service layer for providing Multimedia Telephony.
The VoLTE logical architecture, including roaming and interconnect, is shown in Figure 1.
IMS Core
Telephony
Sh Application Mr’ BGCF
MRF
Server
I Mj
Mr
Cx S Mi
C MGCF/IMS-
P- CSCF/ MGW
HSS IMS-ALG/ Mw I/S- CSCF Mx IBCF/ TrGW
IMS- AGW
Mg BICC/ ISUP/ SIP-I
S6a Sh
eNodeB S6a Cx Ici/Izi
X2 ENUM
Rx Rx CS Network
SGi
Diameter Agent
VoLTE UE MME H- PCRF S9
(DRA/ DEA)
LTE-Uu
S1-MME Gx
SEG S11
Gx
S1-U
eNodeB SGW S5 PGW
HPLMN
Diameter App ID=0
ENUM / IPX
Service ENUM
Provider ENUM IPX
Server
Network
ENUM
CS Network
Ici/Izi
VPLMN Cx Mg
P- CSCF/
S6a Diameter Agent MGCF/IMS-
Rx IMS- ALG/ Mw I/S- CSCF Mx IBCF/ TrGW
S9 ( DRA/ DEA) MGW
IMS- AGW
I Mi
Sh Mr
S6aCx S Mj
C
MME V- PCRF Rx Sh
Telephony
HSS Application Mr’ MRF BGCF
S1- MME Gx Server
SEG S11
VoLTE UE LTE-Uu Gx
S1-U
SGW S5 PGW SGi
eNodeB
IMS Core
Figure 1: VoLTE Logical Architecture

NOTE: The Gm interface (UE to P-CSCF) is included in the VoLTE architecture although
not shown in the above figure.
NOTE: The Ut interface (UE to TAS) is included in the VoLTE architecture although not
shown in the above figure.
V1.0
Official Document
NOTE: The figure details the logical nodes within the VoLTE architecture; however it is
possible to combine functional nodes into a single physical node implementation (e.g. SGW
and PGW). When this is performed, the relevant interfaces between the logical nodes (e.g.
S5) become internal interfaces and therefore are not exposed in the network.
2.1 VoLTE Functional Node Description
The main functional nodes of the VoLTE architecture are defined by 3GPP and are
described below. Further information can viewed in 3GPP TS 23.002 [1].
2.1.1 VoLTE UE (User Equipment)

The User Equipment that is used to connect to the EPC, in the figure above this is an LTE
capable UE accessing EPC via the LTE-Uu radio interface. Other access technologies may
also be supported by the UE.
2.1.2 Evolved Universal Terrestrial Access Network (E-UTRAN)
2.1.2.1 eNodeB
The EUTRAN consists of a single node, the eNodeB that interfaces with the UE. The
eNodeB hosts the Physical (PHY), Medium Access Control (MAC), Radio Link Control
(RLC), and Packet Data Convergence Protocol (PDCP) layers that include the functionality
of user-plane header-compression and encryption. It also offers Radio Resource Control
(RRC) functionality corresponding to the control plane. It performs many functions including
radio resource management, admission control, scheduling, enforcement of negotiated UL
QoS, cell information broadcast, ciphering/deciphering of user and control plane data, and
compression/decompression of DL/UL user plane packet headers.
2.1.3 Evolved Packet Core
2.1.3.1 MME (Mobility Management Entity)
The Mobility Management Entity (MME) is the key control-node for the LTE access network.
It is responsible for idle mode UE tracking and paging procedures including retransmissions.
It is involved in the bearer activation/deactivation process and is also responsible for
choosing the SGW for the UE at the initial attach and at the time of intra-LTE handover
involving Core Network node relocation. It is responsible for authenticating the user (in
conjunction with the HSS). The NAS (Non-Access Stratum) signalling terminates at the
MME which is also responsible for the generation and allocation of temporary identities to
the UEs. The MME validates the permission of the UE to camp on the service provider’s
PMN and enforces UE roaming restrictions. The MME is the termination point in the network
for ciphering/integrity protection for NAS signalling and handles security key management.
Lawful interception of signalling is also a function provided by the MME. The MME provides
the control plane function for mobility between LTE and 2G/3G access networks and
interfaces with the home HSS for roaming UEs.
2.1.3.2 SGW (Serving Gateway)
The SGW routes and forwards user data packets, while also acting as the mobility anchor
for the user plane during inter-eNodeB handovers and as the anchor for mobility between
LTE and other 3GPP technologies (terminating the S4 interface and relaying the traffic
between 2G/3G systems and PGW). For idle state UEs, the SGW terminates the DL data
path and triggers paging when the DL data arrives for the UE. It manages and stores UE
contexts and performs replication of the user traffic in case of lawful interception. The SGW
and PGW functions could be realized as a single network element.
V1.0
Official Document
2.1.3.3 PGW (Packet Data Network Gateway)

The PGW provides connectivity between the UE and external packet data networks. It
provides the entry and exit point of traffic for the UE. A UE may have simultaneous
connectivity with more than one PGW for accessing multiple Packet Data Networks. The
PGW performs policy enforcement, packet filtering for each user, charging support, lawful
interception and packet screening. The SGW and PGW functions could be realized as a
single network element.
2.1.3.4 HSS (Home Subscriber Server)
The HSS is a network database that holds both static and dynamic data elements related to
subscribers. The HSS provides user profile information to the MME and IMS core during UE
attach and IMS registration.
2.1.3.5 PCRF (Policy Charging and Rules Function)
The PCRF provides policy control decisions and flow based charging controls. The PCRF
determines how a service data flow shall be treated in the enforcement function (PGW in
this case) and ensure that the user plane traffic mapping and treatment is in accordance
with the user’s profile.
2.1.4 IMS
IMS is the control infrastructure for supporting next generation IP Multimedia Services and
consists of many separate elements which are listed below.
2.1.4.1 P-CSCF (Proxy Call Session Control Function)
The P-CSCF is the initial point of contact for session signalling for the IMS-enabled VoLTE
UE. The P-CSCF behaves as a SIP proxy by forwarding SIP messages between the UE
and the IMS Core Network, maintains the security associations between itself and the
VoLTE UE, and incorporates the Application Function aspect of PCC to enable binding of
the IMS session with the bearer for applying dynamic policy and receiving notifications of
bearer level events. The P-CSCF may be implemented in an Access Session Border
Controller which may also incorporate the IMS-ALG/IMS-AGW.
2.1.4.2 I-CSCF (Interrogating Call Session Control Function)
The I-CSCF is the contact point within an operator's network for all connections destined to
a user of that network. On IMS registration, it interrogates the HSS to determine which
suitable S-CSCF to route the request for registration. For mobile terminating calls, it
interrogates the HSS to determine which S-CSCF the user is registered on.
2.1.4.3 S-CSCF (Serving Call Session Control Function)
The S-CSCF provides session set-up, session tear-down, session control and routing
functions. It generates records for billing purposes for all sessions under its control, and
invokes Application Servers based on IFCs received from the HSS. The S-CSCF acts as
SIP registrar for VoLTE UEs that the HSS and I-CSCF assign to it. It queries the HSS for
the applicable subscriber profiles and handles calls involving these end points once they
have been registered.
2.1.4.4 Telephony Application Server (TAS)
The TAS is an IMS Application Server providing support for a minimum set of mandatory
MultiMedia Telephony (MMTel) services as defined by 3GPP e.g. supplementary service
functionality, and profiled within GSMA PRD IR.92 [54].
2.1.4.5 MRF (Media Resource Function)
The MRF is a common media resource function, for use by IMS Application Servers and I/S-
CSCFs, to provide media plane processing independent of application types, e.g.
V1.0
Official Document
transcoding, multiparty conferencing, network announcements/tones, etc. under the control

of IMS Application Servers (VoLTE AS) as well as basic media processing functions to
CSCFs. The control plane interfaces to MRFs are defined by the 3GPP references Mr, Mr’,
and Cr interfaces (SIP/SDP and XML encoded media service requests) while the media
plane interfaces to MRFs are defined by 3GPP reference Mb for RTP/RTCP transport.
2.1.4.6 IBCF/TrGW (Interconnection Border Control Function/Transition Gateway)
The IBCF/TrGW is responsible for the control/media plane at the network interconnect point
to other PMNs. The IBCF/TrGW may be implemented in an Interconnect Session Border
Controller.
2.1.4.7 IMS-ALG/IMS-AGW (IMS Application Level Gateway/IMS Access Gateway)

The IMS-ALG/IMS-AGW is not a stand-alone function, but is located with the P-CSCF. The
IMS-ALG/IMS-AGW is responsible for the control/media plane at the access point to the
IMS network. It provides functions for Gate Control & Local NAT, IP realm indication and
availability, Remote NAT traversal support, Traffic Policing, QoS Packet Marking, IMS
Media Plane Security, etc.
2.1.4.8 MGCF/IMS-MGW (Media Gateway Control Function / IMS Media Gateway)

The MGCF/IMS-MGW is responsible for the control/media plane interworking at the network
interconnect point to circuit-switched networks. This includes interworking to CS Networks
based on BICC/ISUP/SIP-I and may include transcoding of the media plane.
2.1.4.9 BGCF (Breakout Gateway Control Function)

The BGCF is responsible for determining the next hop for routing of SIP messages. This
determination is based on information received within the SIP/SDP and routing configuration
data (which can be internal configuration data or ENUM/DNS lookup). For CS Domain
terminations, the BGCF determines the network in which CS domain breakout is to occur
and selects the appropriate MGCF. For terminations in peer IMS networks, the BGCF
selects the appropriate IBCF to handle the interconnect to the peer IMS domain. The BGCF
may also provide directives to the MGCF/IBCF on which Interconnect or next network to
select . Such directives may be given by inclusion of a route header pointing to the next
network ingress node.
2.1.5 Additional Network Functionality

2.1.5.1 ENUM
This functionality enables translation of E.164 numbers to SIP URIs using DNS to enable
message routing of IMS sessions. In the above figure, a single ENUM Server is shown that
is accessible from either PMN as well as IPX. Please refer to GSMA PRD IR.67 [52] for
further information.
2.1.5.2 IPX
This is the IP Packet Exchange transit network providing an interconnect capability between
PMNs. Please refer to GSMA PRD IR.34 [48] for further information.
2.1.5.3 Diameter Agent
The Diameter Agent defined by IETF RFC 3588 [59] and utilised by GSMA PRD IR.88 [53],
is a network element that controls Diameter signalling, enabling the seamless
V1.0
Official Document
communication and control of information between network elements within LTE or IMS
networks and across network borders. A Diameter Agent reduces the mesh of Diameter
connections that negatively impacts network performance, capacity and management.
2.1.5.4 SEG (Security Gateway)
The SEG may be used to originate and terminate secure associations between the eNodeB
and the Evolved Packet Core network. IPsec tunnels are established with pre-shared
security keys, which can take a number of different formats. IPsec tunnels enforce traffic
encryption, for added protection, according to the parameters exchanged between the two
parties during tunnel setup. This enables secure communications between the eNodeB and
EPC across the S1-MME, S1-U and X2 interfaces.
2.2 VoLTE Interface Description
The main interfaces of the VoLTE architecture are defined by 3GPP and are described
below. Further information can be viewed in 3GPP TS 23.002 [1].
2.2.1 LTE-Uu Interface (UE – eNodeB)
LTE-Uu is the radio interface between the eNodeB and the User Equipment. It is defined in
3GPP TS 36.300 [44] series of documents.
2.2.2 S1-MME Interface (UE – MME)
S1-MME is the control plane interface between EUTRAN and MME. The protocols used
over this interface are the Non-access stratum protocols (NAS) defined in 3GPP TS 24.301
[10].
2.2.3 S1AP Interface (eNodeB – MME)
S1AP is the S1 application protocol between the EUTRAN and MME and is defined in
3GPP TS 36.413 [45].
2.2.4 S1-U Interface (eNodeB – SGW)
S1-U is the interface between EUTRAN and the S-GW for per-bearer user plane tunnelling
and inter-eNodeB path switching during handover. The transport protocol over this interface
is GPRS Tunnelling Protocol-User plane (GTPv1-U) defined in 3GPP TS 29.281 [32].
2.2.5 X2 Interface (eNodeB – eNodeB)
X2 is the interface between eNodeB's and is used for X2-based Handover and some Self-
Organising Network (SON) capabilities. The signalling protocol (X2 Application Protocol) is
defined in 3GPP TS 36.423 [46] and the user plane (GTPv1-U) is defined in 3GPP TS
29.281 [32].
2.2.6 S5 Interface (SGW – PGW)
The S5 interface provides user plane tunnelling and tunnel management between SGW and
PGW. The SGW and PGW may be realized as a single network element in which case the
S5 interface is not exposed. The control plane protocol (GTPv2-C) is defined in 3GPP TS
29.274 [31] and the user plane protocol (GTPv1-U) is defined in 3GPP TS 29.281 [32].
2.2.7 S6a Interface (HSS – MME)
The interface enables the transfer of subscription and authentication data for
authenticating/authorizing user access. The protocol used on the S6a interface is Diameter
and is defined in 3GPP TS 29.272 [30].
2.2.8 S9 Interface (H-PCRF – V-PCRF)
The S9 interface provides policy and charging rules and QoS information between the
Home PMN and the Visited PMN in order to support PCC roaming related functions. The
V1.0
Official Document
protocol used on the S9 interface is Diameter and is defined in 3GPP TS 29.215 [27]. The
S9 interface is optional and deployed by bilateral agreement between the Home and Visited
Operators. The policy and charging rules for roaming subscribers may be realised by local
configuration data in the Visited PCRF. However, for completeness, S9 interaction is shown
for all appropriate flows in this document.
2.2.9 S10 Interface (MME – MME)
The S10 interface provides for MME – MME information transfer and is used to enable MME
relocation. The protocol used on the S10 interface is GPRS Tunnelling Protocol-Control
plane (GTPv2-C) and is defined in 3GPP TS 29.274 [31].
2.2.10 S11 Interface (MME – SGW)
The S11 interface is between the MME and S-GW to support mobility and bearer
management. The protocol used on the S11 interface is GPRS Tunnelling Protocol-Control
plane (GTPv2-C) and is defined in 3GPP TS 29.274 [31].
2.2.11 Gx Interface (PCRF – PGW)
The Gx interface is between the PCRF and the PGW, allowing the PCRF direct control over
the policy enforcement functions of the PGW. The protocol used on the Gx interface is
Diameter and is defined in 3GPP TS 29.212 [25].
2.2.12 Rx Interface (PCRF – P-CSCF)
The Rx interface is between the appropriate Application Function (the P-CSCF in the case
of VoLTE) and the PCRF allowing the Application Function to request the application of an
appropriate policy for a session. The protocol used on the Rx interface is Diameter and is
defined in 3GPP TS 29.214 [26].
2.2.13 SGi Interface (PGW – P-CSCF)
The SGi interface is between the PGW and the P-CSCF within the IMS Network. The Gm
reference point from the UE to P-CSCF is tunnelled within SGi for VoLTE services. SGi is
IP-based and is defined within 3GPP TS 29.061 [22].
2.2.14 Cx Interface (I/S-CSCF – HSS)
The Cx interface is between the I/S CSCF and HSS to enable IMS registration and passing
of subscriber data to the S-CSCF. The protocol used on the Cx interface is Diameter and is
defined in 3GPP TS 29.228 [28] and 3GPP TS 29.229 [29].
2.2.15 Sh Interface (VoLTE AS – HSS)
The Sh interface is between the VoLTE Application Server and HSS to enable service and
subscriber related information to be passed to the Application Server or stored in the HSS.
The protocol used on the Sh interface is Diameter and is defined in 3GPP TS 29.328 [33]
and 3GPP TS 29.329 [34].
2.2.16 Gm Interface (UE – P-CSCF)
The Gm interface is between the UE and the P-CSCF and enables connectivity between the
UE and the IMS network for registration, authentication, encryption, and session control.
The protocol used on the Gm interface is SIP/SDP and is defined within 3GPP TS 24.229
[9] and profiled within GSMA PRD IR.92 [54].
2.2.17 Ut Interface (UE – VoLTE AS)
The Ut interface is between the UE and the VoLTE Application Server and allows user
configuration of the supplementary services specified for VoLTE service. The protocol used
on the Ut interface is XCAP and is defined in 3GPP TS 24.623 [21].
V1.0
Official Document
2.2.18 Mx Interface (x-CSCF – IBCF)

The Mx interface is between CSCF and IBCF used for the interworking with another IMS
network. The protocols used on the Mx interface are SIP and SDP and are defined in 3GPP
TS 24.229 [9].
2.2.19 Mw Interface (x-CSCF – x-CSCF)
The Mx interface is between a x-CSCF and another x-CSCF within the IMS core network
(e.g. P-CSCF to I/S-CSCF). The protocols used on the Mw interface are SIP and SDP and
are defined in 3GPP TS 24.229 [9].
2.2.20 Mg Interface (xCSCF – MGCF)
The Mg reference point allows the MGCF to forward incoming SIP/SDP messages that the
MGCF has interworked from the CS Network to the CSCF. The protocols used on the Mg
interface are SIP and SDP and are defined in 3GPP TS 24.229 [9].
2.2.21 Mi Interface (xCSCF – BGCF)
The Mi reference point allows the Serving CSCF to forward the SIP/SDP messages to the
Breakout Gateway Control Function for the purpose of MGCF selection for interworking with
CS networks. The protocols used on the Mi interface are SIP and SDP and are defined in
3GPP TS 24.229 [9].
2.2.22 Mj Interface (BGCF – MGCF)
The Mj reference point allows the Breakout Gateway Control Function to exchange
SIP/SDP messages with the BGCF for the purpose of interworking with CS networks. The
protocols used on the Mj interface are SIP and SDP and are defined in 3GPP TS 24.229 [9].
2.2.23 ISC Interface (S-CSCF –TAS)
The ISC interface is between S-CSCF and Telephony Application Server and is used to
interact with the MMTel supplementary services implemented on the TAS. The protocol
used on the ISC interface is SIP and is defined in 3GPP TS 24.229 [9].
2.2.24 Mr Interface (S-CSCF – MRF)
The Mr interface is between the S-CSCF and the MRF to allow interaction with the media
resource for specific supplementary services (e.g. conference call). The protocol used on
the Mr interface is SIP/SDP and is defined in 3GPP TS 24.229 [9].
2.2.25 Mr’ Interface (TAS – MRF)
The Mr' interface is between the Telephony Application Server and the MRF to allow
interaction with the media resource for specific supplementary services (e.g. conference
call). The protocol used on the Mr' interface is SIP/SDP and is defined in 3GPP TS 24.229
[9].
2.2.26 Cr Interface (TAS – MRF)
The Cr interface is between the Telephony Application Servers and the MRF. And is used
for sending/receiving XML encoded media service requires (Cr) which are served by the
MRF. The protocol is defined in 3GPP TS 24.229 [9], 3GPP TS 24.147 [7], and 3GPP TS
24.247 [11].
2.2.27 Mb Interface (media bearer)
Mb interface is the media bearer plane between UEs and network elements that interact
with the bearer (e.g. MRF). The protocol is based on symmetric RTP/RTCP over UDP as
defined in IETF RFC 3550 [58], IETF RFC 768 [55], and IETF RFC 4961 [62].
V1.0
Official Document
2.2.28 Ici Interface (IBCF – IBCF)

Ici interface is between an IBCF and another IBCF or I-CSCF belonging to a different IMS
network. The protocols used on the Ici interface are SIP and SDP and are defined in 3GPP
TS 29.165 [24].
2.2.29 Izi Interface (TrGW – TrGW)
The Izi interface is between a TrGW and another TrGW or media handling node belonging
to a different IMS network. The protocols used on the Izi interface are RTP and MSRP and
are defined in 3GPP TS 29.165 [24].
2.3 Related GSMA Permanent Reference Documents
The following GSMA PRD's shown in Table 1 are utilised within the VoLTE architecture.
PRD Title Description

Inter-Service Provider The document provides a brief introduction to the requirement for
IR.34
IP Backbone IP interworking and the IPX. It covers the background to the
[48]
Guidelines forerunner of the IPX, the GRX.
This document defines a voice over HSPA IMS profile by profiling
a number of HSPA, (Evolved) Packet Core, IMS core, and UE
IR.58 IMS Profile for Voice
features which are considered essential to launch interoperable
[49] over HSPA
IMS based voice on HSPA. This document is based on the IMS
Voice and SMS profile described in GSMA PRD IR.92 [54].
This document provides guidelines for the centralization of IMS
IMS Service
IR.64 based services and IMS based service continuity for radio devices
Centralization and
[50] by listing a number of Evolved Packet Core, IMS core, and User
Continuity Guidelines
Equipment (UE) features on top of the features defined in IR.92.
This document gives common guidelines for IMS (IP Multimedia
Subsystem as specified by 3GPP) inter-operator connections in
order to prevent non-interoperable and/or inefficient IMS services
IMS Roaming and & networks. Areas covered in the document are IMS specific
IR.65
Interworking issues in roaming and interworking, addressing of users and
[51]
Guidelines network elements, routing of traffic, inter-operator related security
issues, IP version usage and requirements for inter-PLMN
backbone caused by IMS. Document concentrates on the network
level issues.
This document consists of an overview of DNS/ENUM in relation
DNS/ENUM to the successful interworking of MNO services, guidelines for
IR.67 Guidelines for Service general and MNO service specific configuration of DNS/ENUM
[52] Providers & GRX/IPX servers, and describes GSMA defined processes and procedures
Providers relating to configuration and usage of domain names, updates to
the GRX Root DNS Server and so on.
This document presents material about LTE Roaming. The
document addresses aspects which are new and incremental to
IR.88 LTE Roaming LTE. It recognises that much of the data-roaming infrastructure is
[53] Guidelines reused from GPRS and High-Speed Packet Access (HSPA)
Roaming, and for which information and specification is found in
other PRDs.
V1.0
Official Document
This document defines a voice over IMS profile by listing a

IR.92 IMS Profile for Voice number of Evolved Universal Terrestrial Radio Access Network,
[54] and SMS evolved packet core, IMS core, and UE features that are
considered essential to launch interoperable IMS based voice.
This document defines the terms and conditions for IPX which
AA.80 IP Packet eXchange
underpin the Service Level Agreement between the IPX Provider
[74] Service Agreement
and IPX Client.
Table 1: VoLTE GSMA Permanent Reference Documents
V1.0
Official Document
3 VoLTE Implementation - Single PMN

The initial deployments of LTE and indeed VoLTE will likely be self-contained within a single
MNO's domain, serving its own subscribers only (no inter-operator VoLTE interconnect or
roaming capability).
Interworking between a single operators VoLTE network and its CS network is also within
scope.
This section describes the implementation for these scenarios.
3.1 General
The VoLTE architecture for a single PMN deployment is shown in Figure 2.
Figure 2: Intra-PMN VoLTE deployment

NOTE: The Gm interface (UE to P-CSCF) is included in the Intra-PMN VoLTE deployment
although not shown in the above figure.
NOTE: The Ut interface (UE to VoLTE AS) is included in the VoLTE architecture although
not shown in the above figure.
3.2 VoLTE Basic Call Flows
The VoLTE basic call flows are in accordance with 3GPP specifications for E-UTRAN/EPC,
IMS, and PCC. Please refer to 3GPP TS 23.401 [6], 3GPP TS 23.228 [5], and 3GPP TS
23.203 [4] respectively for further detailed information.
The following sub-sections define the additional requirements for the VoLTE service.
References to specific functionality within GSMA PRDs (e.g. IR.92) and 3GPP
specifications will be made within each sub-section.
NOTE: The messages within the call flows within this section are not necessarily performed
in sequential order (e.g. there may be no sequential dependency on some SIP <-> Diameter
interactions). Reference to the 3GPP specifications for further detailed information is
recommended.
V1.0
Official Document
3.2.1 VoLTE UE Attachment and IMS Registration

3.2.1.1 General
A VoLTE UE, under LTE coverage, shall automatically perform an LTE Attach followed by
an IMS registration for VoLTE, if the network supports VoLTE (for further details on the
conditions for IMS registration see section 2.2.1 of GSMA IR.92 [54]). This ensures that the
VoLTE UE shall be available for VoLTE services (i.e. incoming calls, outgoing calls and
supplementary services), similar to the voice experience in today's CS network
deployments.
3.2.1.2 Message Sequence
‎Figure 3 shows the message sequence for the VoLTE UE Attachment and IMS Registration
for the case that the IMS APN is the default APN.
V1.0
Official Document
VoLTE UE eNB MME SGW PGW PCRF HSS P-CSCF I-CSCF S-CSCF TAS
RRC Connection Request

RRC Connection Setup
Attach Req
Attach req - Uplink NAS Transport
RRC Connection Setup complete
[ Dedicated NAS ] [ NAS PDU ]
[ PDN Connectivity Request] [ EPS SM: PDN Connectivity Request ]
Authentication/Security
Update Location Request
Update Location Answer
Create Session Request
IMS-APN; QCI:5; Create Session Request

ARP; APN AMBR IP@; QCI:5; ARP;
IMS-APN; QCI:5;
APN AMBR
ARP; APN AMBR
E-RAB Setup CCR
E-RAB Setup Request Create Session Response CCA
QCI:5; ARP; Create Session Response QCI:5; ARP;
UE AMBR IP@; QCI:5; ARP; APN AMBR;
IP@; QCI:5; ARP; APN AMBR
RRC Connection Reconfiguraton [ NAS PDU ] IP default rule TFT
APN AMBR
[ Activate Default Bearer Request ]
[ NAS PDU ]
[ EPS SM: Activate Default [IMS-APN; IP@; ]
Bearer Request ] [ QCI:5; APN AMBR]
RRC Connection
Reconfiguration Complete
E-RAB Setup
Uplink Direct Transfer E-RAB Setup Response
[ NAS PDU ]
[ EPS SM: Activate Default Bearer Accept ]
Attach complete - Uplink NAS Transport
[ NAS PDU ] Modify Bearer Request
[ Activate Default Bearer Accept ]
Modify Bearer Response
Default Bearer Established for IMS Signalling IMS Signalling over IP
SIP Register SIP Register
UAR
UAA
SIP Register
MAR
MAA
401 Unauthorised 401 Unauthorised 401 Unauthorised
UAR
UAA
SIP Register
SAR
SAA
200 OK 200 OK 200 OK
AAR
AAA SIP Register
UDR
UDA
200 OK
SIP Subscribe
200 OK (Subscribe)
SIP Subscribe SIP Subscribe
200 OK (Subscribe) 200 OK (Subscribe)
SIP Subscribe
SIP Notify
200 OK (Notify) 200 OK (Subscribe)
SIP Notify SIP Notify SIP Notify
200 OK (Notify) 200 OK (Notify) 200 OK (Notify)
Figure 3: VoLTE UE Attachment and IMS Registration message sequence
V1.0 Page 27 of 121

Official Document
NOTE: The Diameter Agent has not been included in this message sequence, although
Diameter messages shall route via the Diameter Agent. Usage of Diameter Agents is
described in section ‎3.6.
3.2.1.3 Detailed Description
3.2.1.3.1 VoLTE UE Attach
When a VoLTE UE attaches to LTE, it executes the normal attach procedure as defined in
3GPP TS 23.401 [6] section 5.3.2.
LTE Radio capabilities are described in GSMA PRD IR.92 [54]; radio bearer capabilities –
section 4.2.1, DRX mode of operation – section 4.2.2, and RLC configuration section 4.3.2.
The VoLTE UE initiates the Attach Request to the eNodeB, with mandatory information
including the EPS Attach Type, NAS key set identifier, IMSI, UE network capability, DRX
parameters, PDN Type (set to IPv4v6), PCO (P-CSCF IPv4 Address Request, P-CSCF
IPv6 Address Request, IPv4 Link MTU Request), Voice Domain Preference and UE’s
Usage Setting (indicating support of IMS voice), ESM message container, etc.
The eNodeB selects the MME from the RRC parameters and forwards the Attach Request
to the MME with the Selected Network and the TAI+ECGI location information of the cell
where it received the message.
Authentication and security mechanisms are performed to activate integrity protection and
NAS ciphering. The MME shall initiate the Security Mode Command to the UE containing
the Selected NAS algorithms, eKSI, ME Identity request, and UE Security Capability. The
UE responds with the Security Mode Complete with the NAS-MAC and ME Identity. After
the completion, all NAS messages are protected by the NAS security functions (integrity
and ciphering).
The MME performs an Update Location to the HSS to retrieve the subscriber profile.
Additional information includes the IMSI, MME Identity, ME Identity and MME capabilities
and homogenous support for IMS Voice over PS session when being able to determine
such support prior performing Update Location. The HSS confirms the Update Location to
the MME with the related IMSI and subscriber data containing a PDN subscription context
with a subscribed QoS profile and subscribed APN-AMBR (Aggregate Maximum Bit Rate).
The UE shall not provide the IMS APN in the initial attach (see clause 4.3.1 of IR.92). The
default APN configured in the HSS can be set as the IMS-APN, and the HSS returns the
IMS-APN name for establishment of the default bearer. The APN-OI information is inserted
by the MME.
If the IMS APN is not configured as default APN, and the UE has determined the need to
establish a PDN connection to the IMS APN, then the UE must establish a PDN Connection
to the IMS APN in a subsequent PDN connection request as specified in clause 4.3.1 in
IR.92.
The MME initiates a Create Session Bearer request to the SGW to create a default bearer
for VoLTE IMS signalling. This message contains the IMSI, MSISDN, IMS-APN, QCI=5,
ARP value, the APN-AMBR, user location information (e.g. TAI+ECGI), UE Time Zone,
RAT-type (EUTRAN), PCO, etc. The SGW creates a new entry in the EPS Bearer table,
allocating a relevant TEID for the control plane and the user plane, which enables it to route
GTP control plane traffic between the MME and the PGW, and forwards the request to the
PGW.
The PGW allocates an IP Address (which can be IPv4 or IPv6) for the UE and utilises
dynamic PCC to initiate a Credit Control Request to the PCRF to obtain the default PCC
rules for the default bearer to be used for IMS signalling. Included in the message are the
IMSI, UE IP Address, default bearer QoS parameters (i.e. QCI=5, ARP, APN-AMBR), user
V1.0
Official Document
location information, time zone information, RAT type (EUTRAN), etc. The PCRF binds the
related policy rules to the IP Address of the default bearer, and responds to the PGW with
the default TFT (traffic flow template) and potentially modified QoS parameters. In the
message to the PGW, the PCRF shall also subscribe to modifications related to the default
bearer in the PGW (e.g. RELEASE_OF_BEARER,
DEFAULT_EPS_BEARER_QOS_CHANGE, etc.).
The PGW creates a new entry in the EPS Bearer table, allocating relevant TEID for the
control plane and the user plane, which enables it to route user plane data between the
SGW and the IMS network with the related policy rules obtained from the PCRF applied.
The PGW sends a Create Session Response to the SGW with the IP Address for the UE,
QoS parameters, PCO, relevant TEID's for the GTP control plane and GTP user plane, etc.
The PGW maps the IMS-APN received in the request to a pre-configured IMS P-CSCF IP
address and inserts this into the PCO as described in GSMA PRD IR.92 [54] section 4.4.
The SGW returns the Create Session Response to the MME.
The MME sends an Attach Accept to the eNodeB with the IMS-APN, IP Address for the UE,
QoS parameters, PCO, IMS Voice over PS supported indication, TAI list, ESM message
container, etc. The eNodeB communicates with the UE to update the RRC configuration
and includes the information received from the core network as part of the create session
request.
The UE sends the Attach Complete message to the eNodeB, which forwards to the MME.
At this time, the UE is capable of sending uplink packets. The MME initiates a Modify Bearer
Request to the SGW including the EPS Bearer Identity, eNodeB address, and eNodeB
TEID. The SGW acknowledges the request to the MME and is capable of sending downlink
packets.
At this stage, the VoLTE UE is attached to the network via a default bearer that is
established for IMS Signalling.
3.2.1.3.2 VoLTE UE Initial IMS Registration
When a VoLTE UE performs the IMS registration, it executes the procedures as defined in
3GPP TS 23.228 [5] section 5.2.
GSMA PRD IR.92 [54] provides additional profiling of IMS Registration procedures – section
2.2.1, IMS authentication - section 2.2.2, IMS Addressing – section 2.2.3.
The VoLTE UE shall not use SIGCOMP as defined in GSMA PRD IR.92 [54] section 2.2.7.
Where an ISIM is present on the UICC, ISIM based authentication and IMS-AKA as
described in GSMA PRD IR.92 [54] section 2.2.2, shall be used. The ISIM application shall
be preconfigured with the related IMS Identities as defined in 3GPP TS 31.103 [37].
Where no ISIM is present on the UICC, USIM based authentication and IMS-AKA as
described in GSMA PRD IR.92 [54] section 2.2.2 shall be used. The UE shall generate the
Private User Identity and the Public User Identity from the IMSI as defined in 3GPP TS
23.003 [2].
The VoLTE UE initiates a SIP REGISTER to the P-CSCF, using the P-CSCF IP Address
that was made available during the LTE Attach. The registration request contains:-
o Within the Contact header, the IMS Communication Service Identifier's
(ICSI) for IMS Multimedia Telephony:-
 urn:urn-7:3gpp-service.ims.icsi.mmtel
 “+sip.instance” containing an IMEI URN
o The feature tag for SMS over IP:- +g.3gpp.smsip
V1.0
Official Document
o The IMS Public User Identity (as derived above) in one of the forms below:-
 Alphanumeric SIP-URI: e.g. user@example.com
 MSISDN as a SIP-URI: e.g.
sip:+447700900123@example.com;user=phone
 MSISDN as Tel-URI: e.g. tel:+447700900123
o The IMS Private User Identity as an NAI: e.g. username@realm
o P-Access-Network-Info with:-
 access-type= 3GPP-E-UTRAN-FDD or 3GPP-E-UTRAN-TDD
 UTRAN-cell-id-3gpp parameter
o Request-URI set to the SIP-URI of the domain name of the home network
o Related headers for IMS AKA parameters
o etc.
The P-CSCF receives the SIP REGISTER request from the UE and inserts a Path header
with a SIP-URI identifying the P-CSCF for routing, a P-Charging-Vector header with the icid-
value, a P-Visited-Network-ID to identify the P-CSCF's network domain and forwards the
request to the I-CSCF. The I-CSCF name is determined via a DNS query or may be pre-
configured within the P-CSCF.
The I-CSCF queries the HSS using the User Authorization Request for authorization and
obtaining the S-CSCF name for the Public User Identity. The HSS validates that the Public
User Identity and Private User Identity are valid and not barred. If there is not an S-CSCF
associated to the Public User Identity, the HSS may return information related to the S-
CSCF capabilities allowing the I-CSCF to select an appropriate S-CSCF. Once the S-CSCF
is identified, the I-CSCF forwards the SIP REGISTER request to the S-CSCF.
The S-CSCF identifies that the SIP REGISTER is part of an initial IMS registration with IMS-
AKA related security. The S-CSCF initiates a Multimedia Authentication Request to the HSS
to retrieve the authentication vectors to perform IMS-AKA security. The HSS stores the
related S-CSCF name for the Public User Identity being registered and returns the
authentication vectors to the S-CSCF.
Upon receipt of the IMS AKA authentication vectors, the S-CSCF stores the XRES and
replies to the SIP REGISTER request with a 401 Unauthorised response indicating that
AKAv1-MD5 is the security mechanism to be used. The RAND and AUTN parameters,
Integrity Key and Cipher Key are also included.
The P-CSCF removes the Cipher Key and Integrity Key from the 401 Unauthorised
response and binds these to the Private User Identity with a set of temporary security
associations for the result of the challenge. The P-CSCF then forwards the response to the
UE.
The UE extracts the RAND and AUTN parameters, calculates the RES, and derives the
Cipher Key and Integrity Key from the RAND. The UE creates a temporary set of security
associations based on parameters received from the P-CSCF (IPSec), and sends a new
REGISTER request to the P-CSCF with a populated Authorization header containing the
RES indicating that the message is integrity protected.
The P-CSCF checks the temporary security associations, and verifies the security related
information received from the UE. This P-CSCF forwards the SIP REGISTER request to the
I-CSCF with the RES included.
V1.0
Official Document
The I-CSCF uses the User Authorization Request message to retrieve the S-CSCF name
stored within the HSS, and forwards the request to the relevant S-CSCF.
The S-CSCF checks whether the RES received in the SIP REGISTER and the XRES
previously stored match. The S-CSCF then performs the Server Assignment Request
procedure to the HSS to download the relevant user profile and register the VoLTE UE.
The S-CSCF stores the route header of the P-CSCF and binds this to the contact address
of the VoLTE UE, this is used for routing to the VoLTE UE in future messages. Parameters
of the P-Charging-Vector header are stored, and the S-CSCF sends a 200 OK response to
the I-CSCF, including the user's display name (retrieved from the user profile in the HSS)
within the P-Associated-URI, which forwards the message to the P-CSCF.
On receipt of the 200 OK from the I-CSCF, the P-CSCF changes the temporary set of
security associations to a newly established set of security associations. It protects the 200
OK with these associations and sends the 200 OK to the VoLTE UE. All future messages
sent to the UE will be protected using the security associations.
Optionally, the P-CSCF sends an AAR message to the PCRF to perform application binding
to the default bearer (i.e. the P-CSCF is requesting to be informed in the event of the default
bearer being lost/disconnected in order to trigger an IMS de-registration). The PCRF
performs the binding and responds with a AAA message to the P-CSCF. Note that if this
message is not sent, then IMS relies on other mechanisms to detect loss of the underlying
default bearer, i.e., loss of connectivity (e.g. timeouts on trying to signal to the UE for an
incoming call or the UE registers in the IMS with a new IP address).
On receipt of the 200 OK, the UE changes the temporary security association to a newly
established set of security associations that will be used for further messages to the P-
CSCF.
The VoLTE UE is now registered with the IMS network for VoLTE services, with SIP
signalling being transported over the default EPC bearer.
The S-CSCF sends a third party SIP REGISTER to the VoLTE AS, as configured in the
initial filter criteria (iFC) within the subscriber profile. The TAS may use the User Data
Request procedure to read VoLTE data stored in the HSS.
The VoLTE UE, P-CSCF and TAS shall subscriber to the registration event package using
the SIP SUBSCRIBE message, in order to be notified on any change of registration state for
the public user identity. In turn, the S-CSCF shall send a SIP NOTIFY to the subscribing
entities informing them of the active registration status.
V1.0
Official Document
3.2.2 VoLTE UE Initiated Detach and IMS Deregistration

3.2.2.1 General
A VoLTE UE shall automatically deregister from IMS before performing an LTE Detach, if the UE is not moving to another access technology
that supports Voice over IMS. This ensures that the VoLTE subscriber can have any terminating services routed accordingly (e.g. terminating
call being routed directly to voicemail) rather than a failed attempt to route the call to the VoLTE UE. This behaviour is similar to the voice
experience in today's CS network deployments.
V1.0 Page 32 of 121

Official Document
VoLTE UE eNB MME SGW PGW PCRF HSS P-CSCF I-CSCF S-CSCF TAS
Default Bearer for IMS Signalling IMS Signalling over IP

SIP Register
SIP Register (expires = 0) (expires = 0)
UAR
SIP Register
UAA
(expires = 0)
SAR
SAA
Notify (reg event) Notify (reg event) Notify (reg event)

200 OK 200 OK
STR Notify (reg event)
200 OK
200 OK
STA
200 OK 200 OK 200 OK
UE is not IMS Registered

Detach Request Delete Session Request
Delete Session Response
Delete Session Request
Detach Accept Delete Session Response
Release Access Bearers Request CCR
Release Access Bearers Response

CCA
UE Context Release Command
RRC Connection Release
UE Context Release Complete
Default Bearer is Deleted and UE is not LTE Attached
Figure 4: VoLTE UE Initiated Detach and IMS Deregistration message sequence
V1.0 Page 33 of 121

Official Document
described in section 3.6.
NOTE: The I-CSCF is shown in the signalling path in figure 4. This is optional and the I-
CSCF may also be omitted.

3.2.2.3.1 IMS Deregistration
When a VoLTE UE performs the IMS deregistration, it executes the procedures as defined
in 3GPP TS 23.228 [5] section 5.3.
The VoLTE initiates a SIP REGISTER message towards the P-CSCF including the Public
User Identity, Private User Identity, the Request-URI with the SIP-URI of the domain name
of the home network, the P-Access-Network-Info, etc. The registration expiration interval
timer shall be set to zero.
The P-CSCF forwards the SIP-REGISTER to the I-CSCF.
The I-CSCF uses the User Authorization Request message to retrieve the S-CSCF name
stored within the HSS, and forwards the request to the relevant S-CSCF.
Upon receiving the SIP REGISTER (expiration time of zero), the S-CSCF shall initiate the
Server Assignment Request procedure to the HSS, indicating User Deregistration Store
Server Name. The HSS shall keep the S-CSCF name associated to the public user identity
for future use and to enable unregistered services to be applied (e.g. routing of a
terminating voice call to voicemail). Note that if the HSS does not keep the S-CSCF name,
then the HSS would need to assign a S-CSCF to handle a new terminating INVITE
message.
The S-CSCF shall send a SIP NOTIFY to the VoLTE UE, TAS and P-CSCF to notify them
of the change of the registration state (the UE, TAS and P-CSCF having previously
subscribed to the reg-event package). The VoLTE UE / TAS / P-CSCF respond with a 200
OK (NOTIFY). If application session binding had been performed at registration, the P-
CSCF (on being notified of the change of registration state) sends a STR message to the
PCRF to remove the session binding to underlying default bearer. The P-CSCF shall
remove the security associations that were established between the P-CSCF and the UE.
The S-CSCF shall send a 200 OK (REGISTER) to acknowledge the de-registration.
The P-CSCF shall forward the 200 OK (REGISTER) to the UE.
On receiving the 200 OK responses, the UE shall remove all the registration details for the
Public User Identity and delete the stored security associations. The UE shall consider the
subscription to the registration event package as cancelled.
The VoLTE UE is now de-registered from the IMS network for VoLTE services, no further
SIP signalling is being transported over the default EPC bearer.
3.2.2.3.2 VoLTE UE Detach
When a VoLTE UE detaches from LTE, it executes the normal detach procedure as defined
in 3GPP TS 23.401 [6] section 5.3.8.
The VoLTE UE initiates a Detach Request to the MME, via the eNodeB which includes the
location information (TAI+ECGI) of the cell the VoLTE UE is using.
The MME initiates a Delete Session Request to the SGW, including the ECGI and
timestamp, to deactivate the default bearer. The SGW releases the default bearer context
information and sends the Delete Session Response to the MME.
V1.0
Official Document
The SGW initiates a Delete Session Request to the PGW including the ECGI , Time Zone
and Timestamp. The PGW acknowledges with the Delete Session Response to the SGW.
The PGW initiates a Credit Control Request to the PCRF to indicate that the default bearer
is released. The user location information (i.e. ECGI) and the Time Zone information are
included.
The MME utilises the Release Access Bearer Request to release the connection between
the SGW and the eNodeB.
The Detach Accept is sent by the MME, and the radio resources between the UE and the
eNodeB are removed.
At this stage, the VoLTE UE is not attached to the network and the default bearer that was
established for IMS Signalling is removed.
3.2.3 Basic VoLTE UE to VoLTE UE Call Establishment – Originating Side
3.2.3.1 General
A VoLTE UE, shall perform call establishment by using the IMS network. The IMS
Signalling shall be sent over the default bearer, and a new dedicated bearer shall be
dynamically established for the voice traffic.
V1.0
Official Document

P-CSCF/ IMS-
VoLTE UE eNB MME SGW PGW PCRF ALG/ IMS-AGW S-CSCF TAS
IMS Signalling via Default Bearer IMS Signalling over IP

SIP Invite (SDP)
SIP Invite (SDP) SIP Invite (SDP)
SIP 100 Trying SIP 100 Trying SIP 100 Trying
SIP Invite (SDP)

SIP 100 Trying
SIP Invite (SDP)
SIP 100 Trying

E-RAB Setup
SIP 183 Progress (SDP)
E-RAB Setup Request
QCI:1; ARP}; SIP 183 Progress (SDP)
Create Bearer Create Bearer Request SIP 183 Progress(SDP)
RRC Conn Reconfiguraton MBR&GBR RAR AAR
Request
[Activate Dedicated Bearer Req] QCI:1; ARP; TFT, QCI:1; ARP; TFT codec AMR
[ Activate Dedicated Bearer Request ] QCI:1; ARP; TFT, MBR&GBR
[ IP@; QCI:1; ] MBR&GBR
[ TFT; MBR&GBR ]
RRC Conn Reconfiguraton Complete
RAA
[ Activate Dedicated Bearer Accept ] E-RAB Setup Create Bearer
Create Bearer Response AAA
E-RAB Setup Reponse Response
[ Activate Dedicated Bearer Res ]
SIP 183 Progress( (SDP)

SIP PRACK SIP PRACK SIP PRACK
SIP 200 OK (PRACK) SIP 200 OK (PRACK) SIP 200 OK (PRACK)
SIP UPDATE (SDP) SIP UPDATE (SDP)
SIP UPDATE (SDP)
SIP UPDATE (SDP)
SIP 200 OK (UPDATE) (SDP)

SIP 200 OK (UPDATE) (SDP) SIP 200 OK (UPDATE) (SDP)
SIP 180 Ringing
SIP 180 Ringing

SIP 180 Ringing SIP 180 Ringing
SIP 200 OK IINV)

SIP 200 OK (INV) SIP 200 OK (INV) SIP 200 OK (INV)
RAR AAR
RAA AAA
SIP ACK SIP ACK
SIP ACK
SIP ACK
VoLTE voice traffic via dedicated bearer VoLTE voice traffic over RTP via IMS media plans elements
Figure 5: Basic VoLTE UE to VoLTE UE Call Establishment - Originating Side message sequence
V1.0 Page 36 of 121

Official Document
NOTE: The figure shows the PCRF being invoked only once on receipt of the of the SDP
answer (uplink & downlink configuration); this is sufficient if the UE is using preconditions as
mandated in GSMA IR.92 [54]. It is also possible to invoke the PCRF twice, i.e. on receipt of
both the SDP Offer (downlink configuration) and SDP Answer (uplink configuration). Both
options are valid - – see 3GPP TS 29.213 ([79]) annex B.
NOTE: This figure shows a double offer/answer exchange supporting preconditions and
utilising the segmented status type as defined in RFC 3312 [70] and described in 3GPP TS
24.229 ([9]) clauses 5.1.3 and 6.1.2.
NOTE: The PRACK and 200 OK (PRACK) messages also traverse through the AS but this
is not shown.
When a VoLTE UE originates a voice call from LTE, it executes the normal mobile
origination procedure as defined in 3GPP TS 23.228 [5] section 5.6.2.
The VoLTE UE initiates a SIP INVITE request, containing the SDP offer with IMS media
capabilities as specified in GSMA PRD IR.92 [54] section 3. The SDP offer shall contain the
AMR Narrowband codec, and it is recommended that the AMR Wideband codec is included
to provide support for HD Voice and shall indicate that local preconditions for QoS are
desired but not yet met, using the segmented status type (as defined in RFC 3312 [70]) and
that the media stream is set to inactive as described in 3GPP TS 24.229 ([9]) clause 6.1.2
The desired QOS for the remote end are set to “none” as the originating UE is unaware of
the QOS requirements at the terminating side. The request is sent to the P-CSCF that was
discovered during the registration procedure. The INVITE request contains:-
o Within the Contact header and the P-Preferred-Service header, the IMS
Communication Service Identifier's (ICSI) for IMS Multimedia Telephony:-
 urn:urn-7:3gpp-service.ims.icsi.mmtel
o The IMS Public User Identity of the calling-party in one of the forms below:-
 Alphanumeric SIP-URI: e.g. user@example.com
 MSISDN as a SIP-URI: e.g.
sip:+447700900123@example.com;user=phone
 MSISDN as Tel-URI: e.g. tel:+447700900123
o P-Access-Network-Info with:-
 access-type= 3GPP-E-UTRAN-FDD or 3GPP-E-UTRAN-TDD
 UTRAN-cell-id-3gpp parameter
o Request-URI set to the SIP-URI or tel-URI of the called-party.
o Within the Supported header, the P-Early-Media, 100rel& precondition
option tags are present (see IETF RFC 5009 [69], IETF RFC 3312 [70] and
IETF RFC 3262 [71]). The timer option tag may also be present (RFC 4028
[80]) when SIP keep-alives are supported.
o etc.
The P-CSCF adds the P-Charging-Vector header and forwards the SIP INVITE to the S-
CSCF that was identified during the registration process.
V1.0
Official Document
If an IMS-ALG/AGW is deployed, then the P-CSCF will also invoke the IMS-AGW over the
Iq reference point (see 3GPP TS 23.334 [73]) to provide appropriate resources in the media
plane. The IMS-AGW is an IP-IP GW and serves as a border element in the media plane in
an IMS network at the access side. .
The P-CSCF forwards the SIP INVITE to the S-CSCF The offered SDP address shall reflect
the media pin-hole created in the IMS-AGW if applicable.
The S-CSCF receives the SIP INVITE from the P-CSCF, and invokes any VoLTE services
as triggered by the initial filter criteria within the subscriber profile that was received during
the IMS Registration. The S-CSCF checks the P-Preferred-Service header in the SIP
INVITE (e.g. MMTel ICSI) and verifies that the user is authorised for the service by
validating against the subscribed services that were retrieved in the service profile during
IMS Registration (Core Network Service Authorisation – Service ID). If the MMTel ICSI is
not in the subscribed services, the INVITE request shall be rejected (403 Forbidden). If
validated, the S-CSCF then adds the ICSI into the P-Asserted-Service header, and removes
the P-Preferred-Service header. Due to service logic within the user profile, and the
identification of the call as a VoLTE call (i.e. MMTel ICSI), the S-CSCF shall route the SIP
INVITE to the TAS at this point to invoke VoLTE supplementary services. The TAS invokes
any supplementary service logic and routes the SIP INVITE to the S-CSCF. The S-CSCF
determines that the Called-Party is within the home network (i.e. ENUM/DNS lookup/internal
configuration) and routes the SIP INVITE to the I-CSCF to determine the terminating S-
CSCF of the Called-Party (see section 3.2.4).
The called party's VoLTE UE will return an SDP answer in a SIP 183 Progress message.
The SDP answer should contain only one codec and indicates that preconditions are also
desired but not yet met at the terminating end and that a confirmation should be sent when
QOS preconditions have been met at the originating side and that the media stream is
inactive. This message is received by the S-CSCF and forwarded to the P-CSCF. The P-
CSCF uses the SDP answer to configure the IMS-AGW if deployed.
In addition, the P-CSCF analyses the SDP in the SDP Answer and sends the
Authorize/Authenticate-Request message to the PCRF with the related service information
(IP address, port numbers, information on media-type). The PCRF authorises the request
and associates the service information with the stored subscription related information
containing the information about the allowed service(s), QoS information and PCC Rules
information. The PCRF identifies the affected IP-CAN session (e.g. default bearer) that has
been established during the LTE Attach procedure, and initiates a Re-Auth-Request to the
PGW to initiate the creation of a dedicated bearer for voice with the related QoS parameters
(QCI=1, ARP) and the related traffic flow template. The PCRF shall also subscribe to
modifications related to the dedicated bearer in the PGW (e.g.
INDICATION_OF_RELEASE_OF_BEARER, etc.).
The PGW acknowledges the Re-Auth-Request to the PCRF, which then acknowledges the
Authorize/Authenticate-Request message sent from the P-CSCF. At this point the IMS SIP
session and the dedicated bearer used for voice are bound together via PCC.
The PGW sends the Create Bearer Request to the SGW to create the dedicated bearer for
VoLTE media. This message contains the dedicated bearer identity, Linked Bearer Identity
to identify the associated default bearer, the traffic flow template, and the associated QoS
parameters (QCI=1, ARP, GBR and MBR), etc. The SGW sends the request to the MME.
The MME sends a Bearer Setup Request message to the eNodeB with the dedicated
bearer identity, Linked Bearer Identity, the traffic flow template, and the associated QoS
parameters in order to activate the dedicated bearer for voice traffic.
The eNodeB maps the QoS parameters to those required for the radio bearer, and then
signals a RRC Connection Reconfiguration to the UE. The UE stores the dedicated bearer
V1.0
Official Document
identity and links the dedicated bearer to the default bearer indicated by the Linked EPS
Bearer Identity. The UE binds the TFT and associated QoS parameters to the dedicated
bearer, and acknowledges the request to the eNodeB, which then acknowledges the Bearer
Request Setup to the MME.
The MME sends the Create Bearer Response message to the SGW to acknowledge the
bearer activation. The message includes the dedicated bearer identity and User Location
Information (ECGI). This is then forwarded to the PGW.
The P-CSCF forwards the SIP 183 Progress response to the VoLTE UE. This message
shall also utilize 100rel and the originating UE shall generate a PRACK which is transited to
the terminating side of the call with an associated 200 OK (PRACK) being received.
The VoLTE UE shall reserve internal resources to reflect the SDP answer and shall confirm
resource reservation by sending a SIP UPDATE message with a new SDP Offer confirming
the selected codec, that local preconditions have been met at the originating end (due to the
establishment of the dedicated bearer) and that the media stream is now set to active. The
UPDATE message is forwarded via the P-CSCF and S-CSCF to the terminating leg of the
call. Note that if the SDP Answer in the 183 Progress message contained more than one
voice codec, then the UE would ensure only a single codec from that multiple list was
included in the new Offer in the UPDATE message (as described in clause 6.1.2. of 3GPP
TS 24.229 ([9]).
The 200 OK (UPDATE) response is received from the terminating leg of the call containing
the SDP answer containing a single voice codec and confirming that preconditions are also
met at the terminating side and that the media stream is active. This message is passed
onto the originating UE via the S-CSCF and P-CSCF.
As preconditions have been met, the terminating UE is now alerted and shall send a SIP
180 (Ringing) response that is received by the S-CSCF and onto the P-CSCF and
originating UE.
The P-Early-Media header is not present in the SIP 180 Ringing message and so the UE
will generate local ring tone to the subscriber. This message shall not utilize 100rel as there
is no SDP within the message.
When the called party's VoLTE UE has answered the call, it sends a 200 OK to the calling
party VoLTE UE. This is received by the S-CSCF and forwarded to the P-CSCF. The P-
CSCF invokes the PCRF with an AAA message to enable both the uplink and downlink of
the dedicated bearer. In turn the PCRF invokes the P-GW with a RAR message to enable
the media flows at the P-GW. The P-CSCF (IMS-ALG) invokes the IMS-AGW (if deployed)
to ensure that duplex media can be conveyed via IMS-AGW at this point.
The P-CSCF forwards the SIP 200 OK (INVITE) to the VoLTE UE.
The VoLTE UE receives the 200 OK, and sends a SIP ACK message to acknowledge that
the call has been established.
At this stage, the VoLTE UE has a call established with voice traffic sent over the dedicated
bearer and via the IMS-AGW. The IMS Signalling is sent over the default bearer. Support of
Robust Header Compression is mandated and described in GSMA PRD IR.92 [54] section
4.1.
3.2.4 Basic VoLTE UE to VoLTE UE Call Establishment – Terminating Side
3.2.4.1 General
A VoLTE UE, shall receive a call via IMS network. The IMS Signalling shall be sent over
the default bearer, and a new dedicated bearer is established by the network for the voice
traffic.
V1.0
Official Document

P-CSCF/
VoLTE UE eNB PGW PCRF IMS-ALG / S-CSCF
MME SGW TAS
IMS-AGW

SIP Invite (SDP)
SIP 100 Trying
SIP Invite (SDP)
SIP 100 Trying
SIP Invite (SDP)
SIP Invite (SDP) SIP 100 Trying
SIP Invite (SDP)
SIP 100 Trying
SIP 100 Trying
RAR AAR
E-RAB Setup Create Bearer
Create Bearer Request
E-RAB Setup Request Request QCI:1; ARP; TFT , codec AMR
QCI:1; ARP}; QCI:1; ARP; TFT, QCI:1; ARP; TFT,
RRC Conn Reconfiguraton
MBR&GBR MBR&GBR MBR&GBR
[ Activate Dedicated Bearer Request ] [Activate Dedicated Bearer Req] RAA
[ IP@; QCI:1; ] AAA
RRC Conn Reconfiguraton Complete [ TFT; MBR&GBR ]
E-RAB Setup Create Bearer SIP 183 Progress (SDP) SIP 183 Progress (SDP)
Create Bearer Response
[ Activate Dedicated Bearer Accept ] E-RAB Setup Reponse Response
[ Activate Dedicated Bearer Res ]
SIP PRACK
SIP PRACK SIP PRACK
SIP 200 OK (PRACK)
SIP 200 OK (PRACK)

SIP 200 OK (PRACK)
SIP UPDATE (SDP)
SIP UPDATE (SDP)

SIP 200 OK (UPDATE) (SDP) SIP 200 OK (UPDATE) (SDP) SIP 200 OK (UPDATE) (SDP)
SIP 180 Ringing SIP 200 OK (UPDATE) (SDP)

SIP 180 Ringing
SIP 200 OK(INV) SIP 200 OK(NV) SIP 200 OK (INV)
RAR AAR SIP 200 OK (INV)
RAA AAA SIP ACK
SIP ACK SIP ACK SIP ACK
VoLTE voice traffic via dedicated bearer VoLTE voice traffic over RTP
Figure 6: Basic VoLTE UE to VoLTE UE Call Establishment – Terminating Side message sequence
V1.0 Page 40 of 121

Official Document
answer (uplink & downlink configuration); this is sufficient if the UE is using preconditions as
mandated in GSMA IR.92 [54]. It is also possible to invoke the PCRF twice, i.e. on receipt of
both the SDP Offer (downlink configuration) and SDP Answer (uplink configuration). Both
options are valid - see 3GPP TS 29.213 ([79]) annex B.
utilising the segmented status type as defined in RFC 3312 [70] and described in 3GPP TS
24.229 ([9]) clauses 5.1.4 and 6.1.2.
is not shown.
When a VoLTE UE receives an incoming voice call request, it executes the normal mobile
termination procedure as defined in 3GPP TS 23.228 [5] section 5.7.2.
The S-CSCF receives a SIP INVITE containing an SDP Offer with IMS media capabilities as
specified in GSMA PRD IR.92 [54] section 3. The SDP offer shall contain the AMR
Narrowband codec, and optionally the AMR Wideband codec. The SDP indicates that
preconditions are applicable and that QOS preconditions are desired but not yet reserved at
the originating side. The media stream is set to inactive.
The S-CSCF invokes any VoLTE services as triggered by the initial filter criteria within the
subscriber profile that was received during the IMS Registration. The S-CSCF shall route
the SIP INVITE to the TAS at this point to invoke VoLTE supplementary services. The TAS
invokes any supplementary service logic and routes the SIP INVITE to the S-CSCF. The S-
CSCF routes the SIP INVITE to the terminating P-CSCF that was associated to the
subscriber during IMS registration.
If an IMS-ALG/AGW is deployed, then the P-CSCF (IMS-ALG) invokes the IMS-AGW to
reserve resources for the media connection. In this event, the SDP address in the INVITE is
over-written to reflect the media pin-hole created on the IMS-AGW.
The P-CSCF forwards the SIP INVITE to the VoLTE UE. When the VoLTE UE receives the
SIP INVITE it shall allocate resources for the call and select one voice codec from the SDP
Offer (as described in section 6.1.3 of 3GPP TS 24.229 ([9]). The UE shall send a SIP 183
Progress response containing the SDP Answer. The message shall indicate that 100rel is
required. The SDP Answer indicates that QOS preconditions are desired but not yet met at
the terminating side of the call. In addition, the SDP shall indicate that the originating side
should confirm when its local QOS preconditions have been met.
On receipt of the SIP 183 Progress message, the P-CSCF updates the IMS-AGW if
applicable with the SDP answer from the UE and sends the Authorize/Authenticate-Request
message to the PCRF with the related updated service information (IP address, port
numbers, information on media-type). The PCRF authorises the request and associates the
service information to the stored subscription related information containing the information
about the allowed service(s), QoS information and PCC Rules information. The PCRF
identifies the affected IP-CAN session (e.g. default bearer) that has been established during
the LTE Attach procedure, and initiates a Re-Auth-Request to the PGW to initiate the
creation of a dedicated bearer for voice with the related QoS parameters (QCI=1, ARP) and
the related traffic flow template. The PCRF shall also subscribe to modifications related to
V1.0
Official Document
the dedicated bearer in the PGW (e.g. LOSS_OF_BEARER,

INDICATION_OF_RELEASE_OF_BEARER, etc.).
The PGW acknowledges the Re-Auth-Request to the PCRF, which then acknowledges the
Authorize/Authenticate-Request message sent from the P-CSCF. At this point the IMS SIP
session and the dedicated bearer used for voice are bound together via PCC.
VoLTE media. This message contains the dedicated bearer identity, Linked Bearer Identity
to identify the associated default bearer, the traffic flow template, and the associated QoS
parameters (QCI=1, ARP, GBR and MBR), etc. The SGW sends the request to the MME.
The MME sends a Bearer Setup Request message to the eNodeB with the dedicated
bearer identity, Linked Bearer Identity, the traffic flow template, and the associated QoS
parameters in order to activate the dedicated bearer for voice traffic.
The eNodeB maps the QoS parameters to those required for the radio bearer, and then
signals a RRC Connection Reconfiguration to the UE. The UE stores the dedicated bearer
identity and links the dedicated bearer to the default bearer indicated by the Linked EPS
Bearer Identity. The UE binds the TFT and associated QoS parameters to the dedicated
bearer, and acknowledges the request to the eNodeB, which then acknowledges the Bearer
Request Setup to the MME.
The MME sends the Create Bearer Response message to the SGW to acknowledge the
bearer activation. The message includes the dedicated bearer identity and User Location
Information (ECGI). This is then forwarded to the PGW.
On receipt of the AAA response from the PCRF, the P-CSCF will convey the SIP 183
Progress (SDP) message to the S-CSCF. The contained SDP reflects the address of the
media pin hole in the IMS-AGW if applicable.
The PRACK message is transited from the originating side of the call.
The terminating side sends a 200 OK (PRACK) in response to the PRACK.
A second SDP Offer is now received from the originating leg of the call in a SIP UPDATE
message indicating that preconditions have been met at the originating side and that the
media stream is active.
The UPDATE is passed to the UE via the S-CSCF and P-CSCF. The UE sends a 200 OK
(UPDATE) response containing a SDP Answer confirming that QOS preconditions are also
satisfied at the terminating side (due to the establishment of the dedicated bearer) and that
the media stream is active. The 200 OK (UPDATE) message is sent to the originating leg of
the call via the P-CSCF and S-CSCF.
As preconditions are now met at both ends, the UE will alert the user and send a SIP 180
Ringing response. This message does not contain SDP and so will not utilize 100rel. The P-
Early-Media header is not present in this message.
The SIP 180 Ringing response is sent to the originating leg via the P-CSCF and S-CSCF.
When the call is answered, the VoLTE UE shall send a SIP 200 OK (INVITE) message to
the P-CSCF.
The P-CSCF invokes the PCRF with an AAA message to enable both the uplink and
downlink of the dedicated bearer to reflect the SDP exchange. In turn the PCRF invokes the
P-GW with a RAR message to enable the media flows at the P-GW.
The P-CSCF (IMS-ALG) shall also invoke the IMS-AGW to if applicable ensure that duplex
media can traverse the IMS-AGW.
The 200 OK is forwarded to the S-CSCF and then to the originating side of the call.
V1.0
Official Document
bearer via the IMS-AGW. The IMS Signalling is sent over the default bearer. Support of
Robust Header Compression is mandated and described in GSMA PRD IR.92 [54] section
4.1.
3.2.5 Basic VoLTE UE to VoLTE UE Call Clearing - Initiated
3.2.5.1 General
A VoLTE UE, shall perform call clearing by using the IMS network. The IMS Signalling shall
be sent over the default bearer, and the dedicated bearer that was dynamically established
for the voice traffic shall be removed.
V1.0
Official Document
VoLTE UE eNB MME SGW PGW PCRF P-CSCF S-CSCF TAS
SIP BYE
SIP BYE
SIP BYE
STR SIP BYE
RAR SIP BYE
Remove TFT rule
RAA SIP 200 OK
STA SIP 200 OK
SIP 200 OK
SIP 200 OK
SIP 200 OK
Delete Bearer Delete Bearer Request

E-RAB Release Request
E-RABToBeRlsdLst
[ Deact. Bearer Request ]
RRC Connection Reconfiguraton

[ Deactivate Dedicated Bearer Request ]

[ Deactivate Dedicated Bearer Accept ] E-RAB Release
E-RABRlsLstBrRelComp
Uplink NAS Transport Delete Bearer

Response
[ Deact. Bearer Accept ] Delete Bearer Response
Dedicated Bearer is Deleted
Figure 7: Basic VoLTE UE to VoLTE UE Call Clearing – Initiated message sequence

NOTE: The Diameter Agent has not been included in this message sequence, although Diameter messages shall route via the Diameter Agent.
Usage of Diameter Agents is described in section 3.6.
V1.0 Page 44 of 121

Official Document

When a VoLTE UE terminates a voice call from LTE, it executes the normal mobile
termination procedure as defined in 3GPP TS 23.228 [5] section 5.10.
The VoLTE UE sends a SIP BYE message to the P-CSCF. If applicable, the P-CSCF (IMS-
ALG) releases the resources in the IMS-AGW.
The P-CSCF also initiates a Session Termination Request to the PCRF to initiate the
process of removing the dedicated bearer that was established for the voice traffic. The
PCRF removes the binding between the stored subscription information and the IMS service
information, and initiates a Re-Auth-Request to the PGW to remove the dedicated bearer for
voice with the related QoS parameters (QCI=1, ARP) and the related traffic flow template.
The Delete Bearer Request, Bearer Release Request, and RRC Reconfiguration Request
are utilised to remove the dedicated bearer utilised for voice traffic.
The P-CSCF forwards the SIP BYE message to the S-CSCF which may invoke any VoLTE
service logic as triggered by the initial filter criteria within the subscriber profile that was
received during the IMS Registration. The S-CSCF shall forward the SIP BYE to the TAS at
this point where VoLTE supplementary services may have been invoked . The S-CSCF
routes the SIP BYE to the S-CSCF of the other party. The other party acknowledges the
SIP BYE with a 200 OK.
At this stage, the VoLTE UE has cleared the call and the dedicated bearer for voice traffic
has been removed.
3.2.6 Basic VoLTE UE to VoLTE UE Call Clearing - Received
3.2.6.1 General
A VoLTE UE shall perform call clearing by using the IMS network. The IMS Signalling shall
be sent over the default bearer, and the dedicated bearer that was dynamically established
for the voice traffic shall be removed.
V1.0
Official Document
VoLTE UE eNB MME SGW PGW PCRF P-CSCF S-CSCF TAS
SIP BYE
SIP BYE
SIP BYE
SIP BYE
SIP BYE
STR
RAR
Remove TFT rule
RAA
STA
SIP 200 OK
SIP 200 OK
Delete Bearer Request SIP 200 OK
Delete Bearer SIP 200 OK
E-RAB Release Request SIP 200 OK
E-RABToBeRlsdLst
[ Deact. Bearer Request ]
RRC Connection Reconfiguraton

[ Deactivate Dedicated Bearer Request ]

[ Deactivate Dedicated Bearer Accept ] E-RAB Release
E-RABRlsLstBrRelComp
Uplink NAS Transport Delete Bearer

Response
[ Deact. Bearer Accept ] Delete Bearer Response
Dedicated Bearer is Deleted
Figure 8: Basic VoLTE UE to VoLTE Call Clearing – Received message sequence

V1.0 Page 46 of 121

Official Document

When a VoLTE UE terminates a voice call from LTE, it executes the normal mobile
termination procedure as defined in 3GPP TS 23.228 [5] section 5.10.
A SIP BYE is received by the S-CSCF from the other party. The S-CSCF shall forward the
SIP BYE to the TAS at this point where VoLTE supplementary services may have been
invoked. The S-CSCF routes the SIP BYE to the P-CSCF which in turn forwards to the
VoLTE UE. The VoLTE UE acknowledges the call clearing by sending a 200 OK.
On receiving the SIP BYE, the P-CSCF (IMS-ALG) frees off the media resources in the
IMS-AGW if applicable. The P-CSCF also initiates a Session Termination Request to the
PCRF to initiate the process of removing the dedicated bearer that was established for the
voice traffic. The PCRF removes the binding between the stored subscription information
and the IMS service information, and initiates a Re-Auth-Request to the PGW to remove the
dedicated bearer for voice with the related QoS parameters (QCI=1, ARP) and the related
traffic flow template.
The Delete Bearer Request, Bearer Release Request, and RRC Reconfiguration Request
are utilised to remove the dedicated bearer utilised for voice traffic.
has been removed.
3.3 VoLTE-CS Interworking
Interworking of VoLTE services within the IMS domain and CS voice calls of a single
Operator are in accordance with 3GPP specifications for IMS and CS Interworking. Please
refer to TS 23.228 [5], 3GPP TS 29.163 [23] and 3GPP TS 29.235 [65] for further detailed
information.
The procedures for the Circuit Switched Core Network are defined within 3GPP TS 23.205
[66] for ISUP/BICC and 3GPP TS 23.231 [67] for SIP-I.
The message sequences in sections 3.2.3 to 3.2.6 of this document apply, with the
difference that the MGCF and IMS-MGW provide the interworking between the IMS
signalling and media plane to the CS core network. The following sections document the
interworking between VoLTE and the CS network via the MGCF/IMS-MGW. The
interactions with PCC and EPC for the establishment of the dedicated bearer for voice are
as detailed in sections 3.2.3 to 3.2.6.
3.3.1 Basic VoLTE UE to CS Call Establishment – Originating Side
3.3.1.1 General
For calls originating on VoLTE and breaking out to the CS network, the originating S-CSCF
shall recognise that the termination is not within the VoLTE domain and shall invoke a
BGCF to determine the target MGCF to break out to the CS network. The message
sequence in section 3.3.1.2 details the interactions of the S-CSCF, BGCF and MGCF and
builds on the message sequence in section 3.2.3.2 from where the S-CSCF propagates the
SIP INVITE message to the terminating leg of the call.
V1.0
Official Document
S-CSCF BGCF MGCF CS Network
IMS Signalling ove IP

ISUP IAM / BICC IAM / SIP-I INVITE (IAM)
SIP Invite (SDP) SIP Invite (SDP) (SDP)
SIP 100 Trying
SIP 100 Trying

SIP 100 Trying (if SIP-I)
SIP 183 Progress (SDP) SIP 183 Progress (SDP) SIP 183 Progress (SDP) (SIP-I)
SIP PRACK SIP PRACK SIP PRACK (if SIP-I)
SIP 200 OK (PRACK) SIP 200 OK (PRACK) SIP-I 200 OK (PRACK) (if SIP-I)
SIP UPDATE (SDP) SIP UPDATE (SDP) SIP UPDATE (SDP) (SIP-I)
COT (BICC / ISUP)
SIP 200 OK (UPDATE) (SDP) SIP 200 OK (UPDATE) (SDP) (SIP-I)

SIP 180 Ringing (SDP) SIP 180 Ringing (SDP) ISUPACM (alerting) / BICC ACM (alerting) / SIP-I 180 (ACM)
SIP 200 OK (INV)

SIP 200 OK(INV) ISUP ANM / BICC ANM / SIP-I 200 OK (ANM)
SIP ACK SIP ACK SIP ACK (if SIP-I)
VoLTE voice traffic over RTP via IMS media plane elements
Figure 9: Basic VoLTE UE to CS Call Establishment – Originating Side

utilising the segmented status type as defined in RFC 3312 [70].
NOTE: The BGCF may not remain in the signalling path having invoked the MGCF (see
clause 5.6.2 of 3GPP TS 24.229 [9]).
The S-CSCF determines that the Called-Party is within the operators CS network (i.e.
ENUM/DNS lookup/internal configuration) and forwards the SIP INVITE to the BGCF.
The BGCF is responsible for selecting an appropriate MGCF for breaking out to the CS
network. The BGCF may use ENUM/DNS or internal configuration data to analyse the
Request-URI to determine the optimum MGCF to select. The Request-URI can be a TEL
URI or SIP URI but will contain either an E.164 number or a telephone number qualified by
a phone-context URI parameter. The BGCF forwards the INVITE to the selected MGCF.
The MGCF is responsible for inter-working to the CS network both in the control plane (IMS
SIP to SIP-I/BICC/ISUP) and media plane via the IMS-MGW and shall follow the procedures
of 3GPP TS 29.163 [23] (for ISUP/BICC) or 3GPP TS 29.235 [65] (for SIP-I). The IMS-
MGW may be required to perform transcoding between AMR-NB/AMR/WB codecs and
other codecs supported within the CS network (e.g. GSM-HR, GSM-FR, GSM-EFR, etc.).
The MGCF selects the outgoing route to the CS network (e.g. (G)MSC-S). The outgoing
route to the CS network may be TDM or IP based and utilizes ISUP or SIP-I/BICC
respectively. The route selection in the MGCF is based on ENUM/DNS or internal
V1.0
Official Document
configuration data. Having selected a route, the MGCF shall invoke an IMS-MGW to
allocate and configure media resources for the call (see 3GPP TS 29.332 [35]).
The MGCF sends an ISUP IAM/BICC IAM/SIP-I INVITE (IAM) to the GMSC-S which in turn
interrogates the HLR to discover location of the MSC-S that the user is currently registered
on. Note that the MGCF may be co-located with a (G)MSC-S. The (G)MSC-S forwards the
request to the MSC-S that the user is registered on and call establishment is progressed as
defined within the 3GPP specifications.
Since the SDP offer from the originating leg indicates that QOS preconditions are desired
but not yet met at the originating side, the MGCF shall (for BICC / ISUP) set the continuity
indicator to “continuity check performed on previous circuit” / “COT to be expected” in the
IAM message for ISUP/BICC respectively.
For ISUP/BICC, the MGCF shall send a 183 Progress message containing the SDP answer
For SIP-I, the MGCF shall receive a 183 Progress message from the peer MSC containing
the SDP answer. In both cases, the SDP answer contains a single voice codec, utilizes
100rel and indicates that QOS preconditions are also desired but not yet met at the
terminating side. In addition, the SDP answer shall request confirmation of QOS
preconditions being met at the originating side.
The 183 Progress (SDP) is sent to the originating leg via the BGCF/S-CSCF.
The MGCF receives a PRACK from the originating side of the call and responds with a 200
OK (PRACK) for BICC/ISUP routes. In the case of SIP-I routes, the MGFCF shall transit
PRACK and 200 OK (PRACK).
The originating UE shall now send an UPDATE message with a new SDP offer confirming
the selected voice codec and indicating that QOS preconditions have been met at the
originating leg. The MGCF receives the UPDATE message and responds with a 200 OK
(UPDATE) for BICC/ISUP routes and transits the UPDATE/200 OK (UPDATE) for SIP-I
routes. The 200 OK (UPDATE) contains the SDP answer which indicates that QOS
preconditions are also met at the terminating side. Since QOS preconditions are now met at
both ends, the MGCF shall (for ISUP/BICC) send a COT message indicating “continuity
check successful”.
The terminating user in the CS network is now alerted and the MGCF receives an ACM
(alerting) message from ISUP/BICC or a SIP 180 Ringing (ACM) message from SIP-I. The
MGCF sends a SIP 180 Ringing message to the originating leg. This message shall not
utilize 100rel. It is strongly recommended that the MGCF includes the P-Early-Media header
in the SIP 180 (Ringing) message as described in 3GPP TS 29.163. At this point, the MGCF
shall also ensure that backward media (e.g. ring tone, progress indications) are conveyed
via the IMS-MGW..
The SIP 180 Ringing is forwarded to the VoLTE UE to indicate a ringing tone to the
subscriber..
When the CS network indicates that the call has been answered, the MGCF sends a 200
OK (INVITE) message to the IMS network. This message is forwarded to the originating leg
of the call and onto the VoLTE UE. The MGCF shall ensure that duplex media can be
conveyed via the IMS-MGW at this point.
the call has been established. The ACK is propagated through the IMS network to the
MGCF. The ACK message is forwarded to the CS network for SIP-I routes.
At this stage, the call is established with voice traffic sent over the dedicated bearer
between the VoLTE UE and the CS network via the IMS-MGW.
V1.0
Official Document
3.3.2 Basic VoLTE UE to CS Call Establishment – Terminating Side

3.3.2.1 General
For calls originating in the CS network and breaking into VoLTE, the call enters the VoLTE
domain via the MGCF. The MGCF routes the call to the I-CSCF in order to determine the S-
CSCF of the terminating user. The message sequence in section 3.3.2.2 details the
interactions of the MGCF, I-CSCF and S-CSCF and builds on the message sequence in
section 3.2.4.2 from where the S-CSCF is initially invoked with the SIP INVITE message for
the terminating leg of the call.
V1.0
Official Document
S-CSCF I-CSCF MGCF CS Network
SIP Invite (SDP) SIP Invite (SDP) ISUP IAM / BICC IAM / SIP-I INVITE (IAM)
SIP 100 Trying SIP 100 Trying (if SIP-I)

SIP 100 Trying
SIP 183 Progress (SDP) SIP 183 Progress (SDP) SIP 183 Progress (SDP) (SIP-I)
SIP PRACK SIP PRACK (if SIP-I)
SIP 200 OK (PRACK)

SIP 200 OK ( PRACK) (if SIP-I)
SIP UPDATE (SDP) (if SIP-I)
SIP UPDATE (SDP)

ISUP / BICC COT

SIP 200 OK (UPDATE) (SDP) (SIP-I)
ISUPACM (alerting) / BICC ACM (alerting) / SIP-I 180 (ACM)

SIP 200 OK (INV) SIP 200 OK(INV) ISUP ANM / BICC ANM / SIP-I 200 OK (ANM)
SIP ACK (if SIP-I)
SIP ACK
Figure 10: Basic VoLTE UE to CS Call Establishment – Terminating Side
NOTE: The I-CSCF is a “stateful proxy” and remains in the signalling path only for the
INVITE transaction.
The CS Network initiates the call establishment by sending an ISUP IAM/BICC IAM/SIP-I
INVITE (IAM) to the MGCF. The MGCF shall follow the procedures of 3GPP TS 29.163 [23]
(for ISUP/BICC) or 3GPP TS 29.235 [65] (for SIP-I).
The MGCF invokes the IMS-MGW to allocate resources for the call and to potentially
transcode between AMR-NB/AMR/WB codecs and other codecs supported within the CS
network (e.g. GSM-HR, GSM-FR, GSM-EFR, etc.).
The target user will be identified via a telephone number for BICC/ISUP and via a SIP or
TEL-URI for SIP-I. The MGCF will map the called party number of BICC/ISUP to a Request-
URI which can either be a TEL URI or a SIP URI with “user-=phone” and shall contain either
an E.164 number or else a national specific number qualified with a “phone-context” URI
parameter as defined IETF RFC 3966 [68].
If overlap signalling is used from the ISUP/BICC CS network, then the MGCF shall
determine when the complete number of address digits have been received (as specified in
3GPP TS 29.163 [23] and TS 29.235 [65]) prior to sending the INVITE message. It is noted
that 3GPP TS 29.163 [23] does permit (as a network option) the INVITE to be sent prior to
determining end of dialling. However, this option is not recommended to be used.
V1.0
Official Document
Contrary to 3GPP TS 29.163 [23] and 3GPP TS 29.235 [65], it is recommended that the
MGCF sends a SIP INVITE indicating that QOS preconditions are desired but not yet met at
the originating side. This occurs irrespective of whether the incoming ISUP/BICC IAM / SIP-I
INVITE indicates that preconditions are not yet met (i.e. via continuity check indicator for
ISUP/BICC or via SIP preconditions for SIP-I). This is done so that the message flows for an
originating CS call align with those of an originating VoLTE UE. Furthermore, preconditions
and SIP UPDATE are supported in IMS (see 3GPP TS 29.163 clause 7.2.3.2.1.2). In
addition, the MGCF shall typically reserve multiple, codecs on the IMS-MGW but will
eventually select one (based on the offer/answer exchange) and it is at this point that
resource reservation is finalised at the originating side (in conjunction with precondition
considerations in the CS network). Note that in figure 10, it is assumed that the incoming
ISUP/BICC IAM / SIP-I INVITE indicates that preconditions are not yet met (i.e. via
continuity check indicator for ISUP/BICC or via SIP preconditions for SIP-I).
The MGCF shall include a SUPPORTED header containing 100rel, precondition and P-
Early-Media in the SIP INVITE message. The INVITE will also contain and SDP offer
reflecting the media resources of the IMS-MGW. The SDP offer will contain multiple voice
codecs (including AMR and AMR-WB) with the media stream set to “inactive” and that QOS
preconditions are desired but not yet met at the originating side For voice calls ingressing
the IMS, the MGCF should insert the media feature tag for IP Voice in Contact header (set
to +g.3gpp.icsi-ref="urn%3Aurn-7%3A3gpp-service.ims.icsi.mmtel") in order to enable the
terminating S-CSCF to invoke the appropriate Application Server into the session.
The MGCF invokes the I-CSCF to enable the appropriate S-CSCF for the target user to be
found.
The I-CSCF interrogates the HSS to identify the S-CSCF where the user is registered and
forwards the INVITE to the S-CSCF. The S-CSCF invokes any VoLTE services as triggered
by the initial filter criteria and routes the SIP INVITE to the AS and terminating P-CSCF as
described in section 3.2.4.
Call establishment proceeds as in section 3.2.4 and the MGCF maps subsequent call
establishment messages from the VoLTE network to the CS network.
A SIP 183 Progress (SDP) message is received from the terminating leg. This message
shall utilize 100rel and the contained SDP answer contains a single voice codec and
indicates that QOS preconditions are desired but not yet met at the terminating side. The
MGCF interacts with the IMS-MGW to reflect the SDP answer by paring down the required
codec list to that of the selected voice codec.
For SIP-I, the 183 Progress (SDP) message is forwarded to the peer MSC and the
associated SIP PRACK message and 200 OK (PRACK) are transited from SIP-I to the
terminating leg of the call via the MGCF.
For ISUP/BICC, the MGCF generates a SIP PRACK message and terminates the related
200 OK (PRACK) message.
For SIP-I routes, an UPDATE (SDP) message shall be received with a new SDP Offer. This
is transited by the MGCF to the originating leg of the call. A 200 OK (UPDATE) message is
received from the originating leg containing an SDP answer and passed through to SIP-I.
For ISUP/BICC, if a COT message is expected, then the MGCF awaits receipt of a COT
message prior to sending the UPDATE message – else the MGCF generates an UPDATE
(SDP) message immediately - with a new SDP Offer. This is sent to the terminating leg of
the call. A 200 OK (UPDATE) message is received from the terminating leg containing an
SDP answer and terminated at the MGCF.
V1.0
Official Document
The second offer / answer exchange has resulted in a single voice codec being selected,
confirmation of preconditions having been met on both originating and terminating ends
and the media stream set to active.
The terminating UE is alerted and a SIP 180 (Ringing) message is received from the
terminating leg which is mapped to an ISUP/BICC ACM (alerting) or SIP-I 180 (ACM)
message. This message does not use 100rel. The P-Early-Media header is not present in
the 180 (Ringing) message, and so the MGCF shall apply ringing tone toward the CS
network and shall inhibit the backward media path through the IMS-MGW. If the P-Early-
Media header is present, then the MGCF enable a backward media path via the IMS-MGW
to convey that media.
When the IMS network indicates that the call has been answered, the MGCF sends an
ISUP/BICC (ANM) or SIP-I 200 OK (ANM) message to the CS network. The MGCF shall
disconnect the ring tone (if previously applied at the IMS-MGW) and ensure that duplex
media can be conveyed via the IMS-MGW at this point .
For SIP-I signalling, the MGCF will receive an ACK message from the CS network that is
propagated to the IMS network. Otherwise, the MGCF shall generate an ACK message
toward the IMS network.
between the VoLTE UE and the CS network via the IMS-MGW.
3.3.3 Basic VoLTE UE to CS Call Clearing - Initiated
3.3.3.1 General
This section describes the call clearing of an VoLTE UE to CS call where the release is
initiated by the VoLTE UE. The message sequence in section 3.3.3.2 details the interactions
of the S-CSCF, BGCF, MGCF and CS Network and builds on the message sequence in
section 3.2.5.3 where the S-CSCF propagates the SIP BYE message to the other leg of the
call.
V1.0
Official Document
S-CSCF BGCF MGCF CS Network
IMS Signalling over IP CS Signalling (ISUP/BICC/SIP-I)
VoLTE Voice traffic over RTP Voice traffic over CS Bearer
SIP BYE
SIP BYE
ISUP REL / BICC REL / SIP-I BYE (REL)
ISUP RLC / BICC RLC / SIP-I 200 OK (RLC)
SIP 200 OK
SIP 200 OK
No VoLTE Voice traffic CS Bearer for Voice is Released
Figure 11: Basic VoLTE UE to CS Call Clearing - Initiated
NOTE: The BGCF may not be in the signalling path (see clause 5.6.2 of 3GPP TS 24.229
[9]).
The call teardown is initiated by the VoLTE UE as described in section 3.2.5.2.
The S-CSCF propagates the SIP BYE message to the BGCF and onto the MGCF.
The MGCF releases the resources in the IMS-MGW and sends an ISUP REL/BICC
REL/SIP-I BYE (REL) to the CS network.
On receipt of the ISUP RLC/BICC RLC/SIP-I 200 OK (RLC), the MGCF sends a 200 OK to
the IMS network in response to the BYE.
At this stage, the VoLTE UE to CS call is released and the dedicated bearer has been torn
down as described in section 3.2.5.2.
3.3.4 Basic VoLTE UE to CS Call Clearing - Received
3.3.4.1 General
This section describes the call clearing of an VoLTE UE to CS call where the release is
initiated in the CS network. The message sequence in section 3.3.4.2 details the
interactions of the S-CSCF, I-CSCF, MGCF and CS Network builds on the message
sequence in section 3.2.6.3 where the S-CSCF receives the SIP BYE message from the
other leg of the call.
V1.0
Official Document
S-CSCF I-CSCF MGCF CS Network
IMS Signalling over IP CS Signalling (ISUP/BICC/SIP-I)
VoLTE voice traffic over RTP Voice traffic over CS Bearer
ISUP REL / BICC REL / SIP-I BYE (REL)

SIP BYE
SIP 200 OK
ISUP RLC / BICC RLC / SIP-I 200 OK (RLC)
No VoLTE Voice traffic CS Bearer for Voice is released
Figure 12: Basic VoLTE UE to CS Call Clearing - Received

INVITE transaction.
The CS Network initiates the call teardown by sending an ISUP REL/BICC REL/SIP-I 200
OK (REL) to the MGCF.
The MGCF releases the resources in the IMS-MGW and sends a SIP BYE to the I-CSCF
and onto the S-CSCF. The call release and release of the dedicated bearer at the UE side
is as described in section 3.2.6.3.
The S-CSCF sends a 200 OK to the BYE to the I-CSCF and onto the MGCF.
The MGCF sends an ISUP RLC/BICC RLC/SIP-I 200 OK (RLC) to the CS network.
At this stage, the VoLTE UE to CS call is released and the dedicated bearer has been torn
down as described in section 3.2.6.3.
3.4 Supplementary Services

3.4.1 General
GSMA PRD IR.92 [54] section 2.3 specifies a mandatory subset of MMTel supplementary
services as defined by 3GPP TS 24.173 [8] for VoLTE. These are listed in Table 2.
Supplementary Service 3GPP Specification Additional Information

Originating Identification 3GPP TS 24.607 [16]
N/A
Presentation
Terminating Identification 3GPP TS 24.608 [17]
N/A
Presentation
Originating Identification 3GPP TS 24.607 [16] Recommended options are specified in
Restriction IR.92 Section 2.3.6
Terminating Identification 3GPP TS 24.608 [17] Recommended options are specified in
Restriction IR.92 Section 2.3.7
Communication Forwarding 3GPP TS 24.604 [13] Recommended options are specified in
Unconditional IR.92 Section 2.3.8
V1.0
Official Document
Communication Forwarding on 3GPP TS 24.604 [13] Recommended options are specified in

not Logged IR.92 Section 2.3.8
Busy IR.92 Section 2.3.8
not Reachable IR.92 Section 2.3.8
No Reply IR.92 Section 2.3.8
3GPP TS 24.611 [19] Recommended options are specified in
Barring of All Incoming Calls
IR.92 Section 2.3.9
Barring of All Outgoing Calls
IR.92 Section 2.3.9
Barring of Outgoing 3GPP TS 24.611 [19] Recommended options are specified in
International Calls IR.92 Section 2.3.9
Barring of Outgoing 3GPP TS 24.611 [19
Recommended options are specified in
International Calls – ex Home
IR.92 Section 2.3.9
Country
Barring of Incoming Calls - 3GPP TS 24.611 [19] Recommended options are specified in
When Roaming IR.92 Section 2.3.9
Communication Hold 3GPP TS 24.610 [18] N/A
Message Waiting Indication
IR.92 Section 2.3.5
Communication Waiting
IR.92 Section 2.3.4
Ad-Hoc Multi Party Conference
IR.92 Section 2.3.3
Table 1VoLTE Mandatory Supplementary Services

VoLTE supplementary service configuration (i.e. XCAP root URI, XCAP APN) is performed
using XCAP at the Ut reference point as described in GSMA PRD IR.92 [54] section 2.3.2.
The APN for XCAP requests must be provisioned by the home operator in the UE as
described in GSMA PRD IR.92 ([54]) section 4.3.1.Note that the XCAP APN is distinct from
the well-known IMS APN.
3GPP has defined a generic baseline XML document (see 3GPP TS 24.623 [21]) which
provides a flexible and extensible framework to enable the management of specific MMTel
services via service specific extensions to the baseline document. There is also the
capability to access an XML document in its entirety down to accessing a sub-set (including
a single item) of an XML document. All UEs and Application Servers shall support the
defined XML documents for the mandatory supplementary service set in Table 2.
Authentication at the XCAP Server via the Ut reference point for secure configuration of
VoLTE supplementary services shall be performed as described in GSMA PRD IR.92 [54]
section 2.2.2. It is recommended that the UE supports the Generic Authentication
Architecture procedures for authentication.
V1.0
Official Document
3.5 ENUM/DNS
3.5.1 General
VoLTE subscribers shall be allocated and identified by a E.164 telephone number. The
E.164 numbering standard has been adopted globally and provides a standardised
interoperable numbering system upon which network operators can build networks and
customer relationship management systems. Usage of ENUM is described in GSMA PRD
IR.67 ([52]).
3.5.2 Number Portability

Historically, telephone numbers were allocated to operators in blocks of E.164 numbers
typically based on dial code. Individual numbers were then allocated to their subscribers
from within their block. Using a table it is possible to look up which operator owned the
number, based on the dialling code. On this basis C7/TDM based service platforms used
the E.164 based telephone number to route traffic.
The introduction of number portability in many countries changed this. It is no longer
possible to determine which operator owns a number based on the dialling code. As a
result, C7/TDM platforms generally request additional information about the full telephone
number from a number registry or database before routing. This process is frequently called
number portability correction and the information accessed is referred to as number
portability data.
3.5.3 IP Service Routing
The VoLTE service is provided by IMS which uses URIs and IP addresses for routing traffic.
IP addresses are associated with URIs through the DNS system. DNS enables the IP
address associated with the target URI to be looked up during the routing process.
In the case where IP services like VoLTE employ an E.164 number as the subscriber
identity the IP network platforms need to identify the target URI destination from the dialled
E.164 number when routing traffic to that number. This can be accomplished by accessing
an E.164/URI registry or database.
3.5.4 Number Resolution
It is required to resolve a dialled E.164 into a routable end destination address, correcting
for the effects of Number Portability. Number Resolution is the process of determining from
an E.164 number what network to direct traffic to and to what address. GSMA and IETF
recommend that both the task of translating the E.164 number into a SIP URI and correcting
for number portability is accomplished via a standardised number look up technology,
ENUM.
3.5.5 ENUM
ENUM is a telephone look up or number resolution technology, standardised by the Internet
Engineering Task Force (IETF) and Carrier ENUM is further specified by the GSMA in PRD
IR.67 [52]. It enables number registries to be queried and return lists of end point
information relating to a telephone number via a standard ENUM API.
ENUM registries can:
 Provide IP routable information against a E.164 phone number
 Provide a list of service/platform information against an E.164 number
 Provide the correct destination correcting for number portability
Benefits:
V1.0
Official Document
 ENUM is an open standard to enable routing of IP/IMS services using telephone

numbers.
 ENUM can be applied to any/every IP service, e.g. Voice, Text, MMS, IM, and
Video, etc. and so is architecturally efficient.
 The ENUM is an API within the IP family of protocols (an extension of DNS) and fits
well with IP infrastructure.
 Number Resolution is required for internal and external routing. Using a standard is
essential for interoperability where queries will take place to external ENUM data
sources on a global basis.
 ENUM is agnostic of the type of network or interconnect medium an operator
chooses to use.
3.5.5.1 ENUM Network Impact
When IP services are available and the network infrastructure is required to route a piece of
traffic to an E.164 number there is uncertainty over whether the destination number is;
 Hosted by IMS technology or the pervasive legacy C7/TDM technology
 An original number allocated to the service provider
 An imported number from another service provider
 A number exported to another service provider
An ENUM query is performed to receive information about the end destination and enable it
to take the correct routing decision. An ENUM query is designed to return a number
portability corrected URI for a particular E.164 which the network can resolve to an IP
address for routing.
3.5.5.2 ENUM Functionality
Figure 13 shows an example of how ENUM is used during the setup of a person to person
session. In this general example the ENUM database is shown as a cloud, as there are
options for the storage and management of ENUM data. The originating subscriber
establishes a session towards the destination party. The originating infrastructure resolves
the number before taking a routing decision by accessing an ENUM database. The ENUM
database returns the URI associated with the destination subscriber. The originating
infrastructure resolves the destination URI into an IP address via DNS. Routing takes place
based on the destination IP address.
V1.0
Official Document
1. Query: +447777123456
ENUM
2. Result: Voice, SIP,
Cloud
447777123456@gateway.spdomain
Call
+447777123456 3. Obtain IP address:
447777123456@gateway.spdomain
Originating DNS
Application Cloud
4. IP address: 212.43.25.107
Originating IMS
SP
5. Route session gateway
212.43.25.107
+447777123456
Destination IMS
Fig 1. ENUM based routing SP
Figure 13: ENUM Based Routing
3.5.5.3 ENUM Registry Architecture
Subscribers can dial numbers to any destination in the world at any time. It is useful to
classify routing scenarios as;
 Internal
 External National
 External International
Different strategies and regulations may apply to the associated ENUM approach.
In order to resolve numbers, operators need to arrange access to an ENUM registry, or
registries, that contain the information for the above routing scenarios. Typically operators
have
 Full knowledge of the numbers they serve
 Full, partial or no knowledge of the numbers ported in their country depending on
local number portability procedures
 Little or no knowledge of the numbers existing and ported in other countries
As such operators should adopt a strategy whereby internal ENUM records can be
accessed and external ENUM registries can be accessed.
3.5.5.4 Internal ENUM Registries
Typically an operator may wish to establish an ENUM registry for internal use. This may
comprise of two elements, records concerning the subscribers it supports and a cache of
frequently dialled external numbers. The IP switching infrastructure is connected to this
registry in order to resolve numbers. In the event that a number is not present or cannot be
V1.0
Official Document
resolved by the local registry then the IP infrastructure needs to employ an alternative
strategy to query an external registry system. The internal registry is populated by the
operator. The operator associates URIs against its E.164 numbers in this registry that will
result in successful internal routing. Cached results are provided from external sources and
should be stored as received otherwise routing will be unsuccessful. The cache .time (also
known as “time to live”) provided by external sources for returned results should be
respected since routing based on out of date data could be unsuccessful e.g. cause a call
set-up failure. The Internal database could be managed by the operator, outsourced as with
other network components or be part of a national system depending on local number
portability regulations. Selection of the best approach will vary from country to country and
network to network and local analysis is required to select the right approach.
3.6 Diameter Signalling
3.6.1 General
The VoLTE architecture utilises Diameter interfaces between components of the EPC, PCC,
and IMS core network. 3GPP has defined a number of Diameter Applications based on the
Diameter Base Protocol is defined within IETF RFC 3588 [59].
3.6.2 Diameter Agents
In order to support scalability, resilience and maintainability of the Diameter interfaces, the
usage of a Diameter Agent is recommended. The Diameter Agent reduces the mesh of
different diameter connections within the network to aid in routing of Diameter messages,
provides load-balancing functionality for handling of signalling congestion, and provides
protocol interworking functionality (e.g. Diameter application AVP's, transport protocol, etc.).
Further description of the Diameter Agent is described within GSMA PRD IR.88 [53] section
3.1.3.
3.6.3 Diameter Transport
The Diameter Base protocol defines the clients must support either TCP or SCTP as the
transport protocol, and that the servers and agents must support both TCP and SCTP.
A number of the Diameter interfaces defined by 3GPP (i.e. Cx, Sh, S6a, S9) have profiled
this further by stating that SCTP must be supported by the relevant nodes. Whilst other
Diameter interfaces defined by 3GPP (i.e. Gx, Rx) allow support of SCTP or TCP.
It is recommended that SCTP is utilised as the transport protocol across all Diameter
interfaces utilised in the VoLTE architecture, or that a Diameter Agent is deployed to
perform the interworking between SCTP and TCP.
3.6.4 Diameter Peer Discovery
To enable Diameter routing within the home network, a Diameter node needs to discover
which peer to route messages to for a specific application. This may be performed by using
manual configuration of Diameter nodes within each node. However by allowing for a
dynamic discovery mechanism (NAPTR query), it allows for a simpler, scalable, and robust
deployment.
GSMA PRD IR.88 [53] section 3.1.3.4 describes the peer discovery mechanism in further
detail.
Note that the dynamic discovery mechanism becomes more necessary, as the scale of the
connections increases and topology hiding becomes more important.
3.6.5 Diameter Capability Exchange
As the Diameter Agent acts as a proxy for the network elements supporting a Diameter
interface, it must perform a Capability Exchange as defined within IETF RFC 3588 [59] to all
V1.0
Official Document
of its Diameter peers within the network. Therefore the Diameter Agent shall support the
superset of all the Diameter interfaces required for VoLTE.
3.6.6 Diameter Routing
All Diameter enabled elements in a network shall route their requests and responses via the
Diameter Agent. The Diameter Agent routes the requests/responses to the correct
destination based on the host and realm identity in the message as follows:-
 If the Diameter Client knows the address/name of the Diameter Server (e.g. pre-
configured), the request shall include both the Destination-Realm and Destination-
Host and forward the request message to the Diameter Agent. The client will add its
own Origin-Host and Origin-Realm information.
 If the Diameter Client does not know the address/name of the Diameter Server, it
shall forward the request to the Diameter Agent which will determine the destination
address/name by analysing the received Diameter Application ID, Destination
Realm, and its internal routing table information established during peer discovery
and capability exchange. The Diameter Agent shall insert the Destination-Host and
forward the message to the destination node.
A Diameter Server shall store the Origin-Host and Origin-Realm within the request, to be
used for future messages, in the respective destination host/realm parameters. The
Diameter response message from the server shall include its own host/realm parameters in
the Origin-Host and Origin-Realm, and forward the response message to the Diameter
Agent which in turn proxies the response to the Diameter Client. The Diameter Client shall
store the Host/Realm parameters for future usage.
The Diameter Agent may optionally overwrite the host/realm information (e.g. topology
hiding) but this requires a mapping table to be maintained within the Diameter Agent and is
not required for a single network deployment.
3.7 Traffic Management and Policy
3.7.1 General
In order to meet the requirements of the VoLTE services in regards to guaranteed bit-rate
for the voice data, sensitivity to jitter and delay, operators need to implement a traffic
handling policy. This requires a standardised mechanism that specifies the relative priority
for the VoLTE service. It ensures that end-users receive a high level of Quality of
Experience for VoLTE, by adapting the handling of the applications within the acceptable
bounds of the Quality of Service characteristics.
In the VoLTE architecture, this is performed in two ways:-
 Policy and Charging Control (PCC)
 DiffServ
3.7.2 Policy and Charging Control
Policy and Charging Control functionality encompasses Flow Based Charging (including
charging control and online credit control) and Policy control (e.g. gating control, QoS
control, QoS signalling, etc.).
This is achieved by three logical network components; the Application Function (AF) –
incorporated within the P-CSCF, the Policy Charging and Rules Function (PCRF), and the
Policy Control Enforcement Function (PCEF) – incorporated within the PGW.
The P-CSCF provides information related to the control-plane signalling of a particular
application to the PCRF. New application requests may invoke requests to the PCRF in
order to modify or create new IP-bearers for that application.
V1.0
Official Document
The PCRF provides policy control decisions and flow based charging controls which are
enforced by the PGW. The PCRF determines how a service data flow shall be treated, per-
subscriber, in the PGW and ensures that the user plane traffic mapping and treatment is in
accordance with the user’s profile. The PCRF also receives information related to control-
plane signalling and application control from the P-CSCF, which may result in a modification
of policy rules at the enforcement level. It is also capable of feeding back information
related to the user-pane to the application level, e.g. loss of bearer.
The PGW supports flow gating, rate limiting, policing, shaping, Differentiated Services
(DiffServ) marking, and other features. The control of applying these functionalities is
dependent on a per-subscriber basis and dependent on the policy rules received from the
PCRF. In addition the PGW supports deep packet inspection (DPI) capabilities, QoS
controls, and advanced reporting capabilities to the PCRF including support for volume-
based reporting.
Through the usage of the PCC architecture, and the binding between the control-plane
signalling over the default bearer and the user-plane media over a dedicated bearer is
achieved with the related QoS enforced.
Policy rules within PCC are defined on the service data flow level by defining a Quality of
Service rule to be applied to detect the type of service data flow and apply the correct QoS
parameters to the session. 3GPP TS 23.203 [4] standardises a set of QoS Class Identifiers
(QCI) for different services with related QoS parameters defined (i.e. resource type, priority
level, packet delay budget, packet error loss rate).
QCI Resource Priority Packet Packet Example Services

Type Delay Error
Budget Loss
Rate
1 2 100 ms 10-2 Conversational Voice
2 4 150 ms 10-3 Conversational Video (Live
GBR Streaming)
3 3 50 ms 10-3 Real Time Gaming
4 5 300 ms 10-6 Non-Conversational Video

(Buffered Streaming)
5 1 100 ms 10-6 IMS Signalling
6 Video (Buffered Streaming)

6 300 ms 10-6 TCP-based (e.g., www, e-mail,
chat, ftp, p2p file sharing,
progressive video, etc.)
7 Non- Voice,
-3
GBR 7 100 ms 10 Video (Live Streaming)
Interactive Gaming
8
8 Video (Buffered Streaming)
300 ms 10-6 TCP-based (e.g., www, e-mail,
chat, ftp, p2p file sharing,
9 9 progressive video, etc.)
V1.0
Official Document
Table 2 : Standard QCI Values and related QoS Parameters

VoLTE utilises QCI=5 for IMS Signalling on the default bearer and QCI=1 for conversational
voice on dedicated bearers.
The access network (e.g. eNodeB) utilises the QCI specific parameters that control packet
forwarding treatment (e.g. scheduling weights, admission thresholds, queue management
thresholds, link layer protocol configuration, etc.), that have been pre-configured by the
operator. The QCI, and associated parameters, are also mapped to the relevant DSCP
marking on IP packets at the transport layer.
3.7.3 DiffServ
DiffServ (Differentiated Services) is used to differentiate between different services utilising
the same IP bearer, to enable IP-routers and other nodes to handle specific traffic in a
differentiated way. For instance, IP packets that are related to a specific service may
require a higher level of buffer management and packet scheduling mechanisms for the
traffic that it carries.
In order to provide this functionality within a service-aware node, the IETF have defined an
IP packet header field to allow these nodes to "mark" IP Packets with a class of service.
The class of service is referred to as the DiffServ Code Point (DSCP). As the "marking" of
these packets is performed at the IP layer, IP routers and other network components that
are not service-aware can perform the relevant treatment of data traffic, dependent on the
value of the DSCP value per IP packet. For example, the utilisation of DSCP may enable
the VoLTE service that requires a low level of latency to be routed through an IP network
with a higher priority compared to best-effort internet traffic. This is accomplished by all IP-
nodes in the data path recognising the DSCP value, performing the relevant traffic
management, without information of what the data traffic pertains to.
It should be noted that the usage of DiffServ is widely deployed within the industry, mobile
and fixed networks, and forms a basis for base IP-routing network components (i.e. IP
switches and routers).
3.7.4 Mapping between QCI and DiffServ
GSMA PRD IR.34 [48] section 6.2 describes the different traffic classes that are used.
These are shown in Table 3.
QoS Information IP transport
QCI Signalling Diffserf

Traffic Class THP DSCP
indication PHB
1
2 Conversational N/A N/A EF 101110
3
4 Streaming N/A N/A AF41 100010
5 1 Yes AF31 011010

(see note)
Interactive
6 No AF32 011100
7 2 No AF21 010010
V1.0
Official Document
8 3 No AF11 001010
9 Background N/A N/A BE 000000
Table 3: EPC QoS information and mapping to DSCP

VoLTE utilises QCI=5 for IMS Signalling on the default bearer and QCI=1 for conversational
voice on the dedicated bearer.
3.8 Session Border Controllers
The Session Border Controller (SBC) has primarily been deployed within the context of IP-
Interworking at the edge of the Operators network for Interconnect to the IPX or direct
connection to another Operator's network. Within the VoLTE architecture, SBC's are
commonly used on the IMS NNI incorporating the IBCF/TrGW functionality. SBC's on the
access to IMS are becoming more prevalent, incorporating specific IMS functionality i.e. P-
CSCF, IMS-ALG, and IMS-AGW. It is important to understand that the SBC receives both
the Control Plane signalling and the User Plane data for data sessions.
The SBC can be seen as the point for ingress and egress for the Network Operator, a
signalling and user plane gateway. As part of the functions that the SBC provides are
Security (i.e. Firewall, topology hiding), Control-plane Interworking between different
protocols, Network Address Translation, Transcoding between different user plane data-
types, load-balancing and routing, etc.
However, in terms of traffic management, SBC's provide the functionality of analysing the
control plane signalling to ensure that operator policies (ingress and egress) are adhered to
across a network interconnect. Dependent on the service-level aspect derived from the
control plane signalling, and with Operator configuration, the SBC is capable of setting the
correct DSCP value for that service.
Additionally, the majority of SBC's also incorporate a Deep Packet Inspection solution. In
this respect, the SBC provides the functionality of analysing the class of service for data
traffic on the IP Bearer. Therefore SBC's are capable of differentiating traffic (e.g. voice,
video, internet, etc.) at the bearer level and setting the relevant DSCP accordingly, enabling
the correct handling of IP-traffic via intermediate nodes, to ensure end-to-end Quality of
Service from the Interconnect perspective.
3.9 Emergency Call
The VoLTE UE and network are required to support emergency calls in IMS as described in
GSMA PRD IR.92 [54] section 5.2. However support of emergency calls in the CS domain
may be a local regulatory requirement. Therefore, in areas where LTE radio coverage
and/or IMS network is not available, the network shall be able to reject an IMS emergency
attempt and the VoLTE UE shall support CS emergency calls as used today. GSMA PRD
IR.92 [54] Annex A.5 describes this in further detail.
3.10 Lawful Intercept
The architecture for providing Lawful Intercept for VoLTE services is defined in 3GPP TS
33.107 [41] section 7A with call scenarios shown in Annex C. In a VoLTE deployment, the
Access Session Border Controller (P-CSCF, IMS-ALG, IMS-AGW) may be used as the
point of intercept, due to the available access to both the IMS signalling and the media.
3.11 Security
3.11.1 General
The security architecture for LTE is defined in 3GPP TS 33.401 [43].
V1.0
Official Document
3.11.2 Security Gateway

Within the VoLTE architecture the Security Gateway is an optional network node. It is
particularly relevant for deployments where RAN sharing may be utilised. The Security
Gateway is used to originate and terminate secure associations between the eNodeB and
the Evolved Packet Core network. IPsec tunnels are established with pre-shared security
keys, which can take a number of different formats. IPsec tunnels enforce traffic encryption,
for added protection, according to the parameters exchanged between the two parties
during tunnel setup. This enables secure communications between the eNodeB and EPC
across the S1-MME, S1-U and X2 interfaces.
3.11.3 IMS Media Plane Security
IMS Media Plane Security is defined within 3GPP TS 33.328 [42]. It provides security for
RTP and MSRP based media and is used in the following ways:-
 User plane security between UE and IMS access (IMS-ALG, IMS-AGW), commonly
referred to as e2ae media protection.
 User plane security end to end between UE's, commonly referred to as e2e media
protection.
For the VoLTE services using RTP, media plane security is not profiled in GSMA PRD IR.92
[54] but may optionally be provided utilising SRTP (Secure RTP) as defined in IETF RFC
3711 [60] / 3GPP TS 33.328 [42]. The key management solutions are described in 3GPP
TS 33.328 [42].
3.12 SMS over IP

VoLTE architecture supports the delivery of SMS over IP as described in GSMA PRD IR.92
[54] section 2.5 and is based on the functionality and procedures defined in 3GPP TS
24.341 [12]. The VoLTE UE is required to support the SM-over-IP sender/receiver and an
IP-SM-GW is required within the IMS core network.
Where interaction with legacy SMS services are required, these are implemented as
described in GSMA PRD IR.92 [54] Annex A.7.
3.13 Support of Legacy Proprietary CS Services
The VoLTE Architecture and baseline 3GPP specifications do not provide a standardised
solution to interwork with legacy proprietary services that are commonly utilised within
Operators CS networks.
Three possibilities exist for legacy CS proprietary services within the VoLTE Architecture:-
1. Cease support of the legacy proprietary services for VoLTE users:- This may be
regarded as being a sub-optimal approach as some services are required under
local regulation and other services a deemed to provide distinct value-add to the end
user.
2. Develop the legacy proprietary services within the TAS:- This approach may be
regarded as being not cost-effective and the time required to develop each service,
on a case by case basis, would not enable a timely deployment of VoLTE.
3. Develop interworking between the VoLTE architecture and the existing legacy
proprietary CS network services:- A standardised approach would take time to
agree on a solution and therefore reliance on proprietary solutions may be required.
Potential proprietary solutions may involve the deployment of a gateway to interwork
between the IMS SIP architecture and the legacy CS network (e.g. IM-SSF). This
gateway may be collocated with the TAS.
V1.0
Official Document
3.14 Complementing VoLTE with 2G/3G Voice

3.14.1 SRVCC
As LTE radio access may not be deployed in a ubiquitous manner, the operator may wish to
complement a VoLTE network with their existing CS radio access for voice.
In order to enable seamless voice service between VoLTE and CS voice, GSMA PRD IR.92
[54] Annex A describes how Single Radio Voice Call Continuity (SRVCC) procedures are
used when performing a handover between LTE and GSM/UMTS coverage.
The enhanced SRVCC functionality defined within 3GPP Release 10 also supports the
handover of calls that have a mid-call feature being applied (e.g. calls on hold, calls that are
in a call waiting state), support for calls in alerting state and enhancements to minimise
voice interruption delay and is therefore recommended for an SRVCC deployment.
3.14.2 PS Handover
As an alternative to SRVCC, PS Handover may be preferable in order to keep the user
utilising VoLTE service in the IMS domain. GSMA PRD IR.58 [49] defines the profile IMS
profile for Voice over HSPA (VoHSPA) and the description of the Inter-RAT PS handover to
provide mobility to/from LTE in section 3.5.
3.14.3 IMS Service Centralization and Continuity
3GPP has standardised the principle for centralisation and continuity of services in IMS to
provide consistency to the end user regardless of which access technology is utilised. This
is further profiled within GSMA PRD IR.64 [50].
GSMA PRD IR.92 [54] Annex A.4 mandates the usage of the Ut interface for setting the
supplementary service configuration when the UE is using the CS network for voice service.
This implies that the services are centralised and calls anchored within the VoLTE IMS
domain by using one of the solutions as described in GSMA PRD IR.64 [50] (e.g. an MSC-S
enhanced for ICS, utilising CAPv2, or CS-interworking.
However, the initial deployments of VoLTE IMS networks may not offer service
centralisation, with specific issues defined within GSMA PRD IR.64 [50] Annex A.
As VoLTE mandates Ut for supplementary service configuration, the network needs to
ensure that supplementary service settings are the same in both VoLTE and CS networks.
This can be done by one of the following two ways:
1. Avoiding synchronisation and accepting the limitation of a subscriber having to
manually configure supplementary services in both domains.
2. Synchronization between the CS and IMS/MMTEL subscription data which ensures
that the service settings in IMS (over Ut) are set the same as their CS service
equivalents. This has been studied in 3GPP but finally no solution was standardized
due to the complexity and different ways that such data is stored internally within the
likes of the HSS/HLR and VoLTE MMTel AS. A potential solution could be to utilise
User Data Convergence (UDC) architecture as defined within 3GPP TS 23.335 [75]
for HLR/HSS integration.
3.15 Charging
VoLTE charging architecture for Online and Offline Charging is defined in 3GPP TS 32.240
[38] and 3GPP TS 32.260 [39] which defines the IMS Charging architecture. The Diameter
protocol is utilised for the Ro interface for online charging and the Rf interface for offline
charging.
V1.0
Official Document
VoLTE charging may be based on the utilisation of Charging Data Records (CDR's)
generated by EPC and IMS nodes. The format and content of the CDR's for the SGW,
PGW, P-CSCF, S-CSCF and the TAS are defined in 3GPP TS 32.298 [40].
3.16 Codecs
GSMA PRD IR.92 [54] has mandated AMR / AMR-WB codecs to be used for VoLTE and
these codecs shall be implemented by all equipment manufactures to ensure good voice
quality on VoLTE as well as facilitating inter-operability and avoiding transcoding. Other
voice codecs may optionally be offered in addition to the AMR codecs.
3.17 IP Version & Transport
As described in GSMA PRD IR.92 [54] section 5.1, both IPv4 and IPv6 must be supported
by the UE and network for VoLTE. On attaching to the network, the UE must request the
PDN type (IPv4v6) – see section 3.2.3.1. If both IPv4 and IPv6 addresses are assigned for
the UE, the UE must prefer IPv6 when performing P-CSCF discovery.
As recommended in GSMA PRD IR.92 [54], both UDP and TCP shall be supported for SIP
message transport both by the UE and network. If UDP is used for SIP transport,
implementations should avoid fragmentation by obeying clause 18.1.1 of IETF RFC 3261
[69] and swap over to TCP for large messages. Note that the transport change is done on
per SIP message basis and not on a per SIP session basis.
3.18 Home eNodeB (HeNB)

It is also possible to deploy an eNode-B to provide Femtocell coverage. Such an element
resides in the CPE and is termed a Home eNodeB (HeNB). This option is shown in figure 14
below.
HeNBs are usually connected to the EPC via both a SEG and HeNB GW. The HeNB GW
serves as a concentrator for the C-Plane, specifically the S1-MME interface, between the
HeNB and MME. The HeNB GW appears to the MME as an eNodeB and appears to the
HeNB as a MME. The S1-U interface from the HeNB may also be terminated at the HeNB
GW or may bypass the HeNB GW.
S1-U S1-U
HeNB
S1- S1-
MME MME
X2 HeNB
S1- GW
S1- EPC
MME MME
SeGW
HeNB S1-U S1-U
HeNB
Mgmt
System
Figure 14: HeNB Architecture

V1.0
Official Document
V1.0
Official Document
4 VoLTE Implementation - Interconnect

This section builds on the single PMN scenario in section 3 and considers inter-operator
VoLTE interconnect between PMNs. All users are attached and registered in their
respective home networks (i.e. no roaming) and calls are established between users in
different PMNs via the NNI.
Two ways for inter-connect are considered to be in scope, namely a bi-lateral interconnect
and interconnect via an IPX.
4.1 General
The VoLTE architecture for an interconnect deployment is shown in Figure 15.
IMS Core
Telephony
MRF
Server
I Mx
Mr
Cx S Mi
C
P- CSCF/
IMS- AGW
S6a Sh
ENUM
Ici/Izi
X2 Rx Rx
SGi
Diameter Agent
VoLTE UE MME PCRF
( DRA/ DEA)
LTE-Uu
S1- MME Gx
SEG S11
Gx
S1-U
eNodeB SGW S5 PGW
PLM1
ENUM / IPX
Service ENUM
Provider ENUM IPX
Server
Network
ENUM
Ici/Izi Ici/Izi
PLM2 Cx
P- CSCF/
S6a Diameter Agent
Rx IMS- ALG/ Mw I/S- CSCF IBCF/ TrGW
( DRA/ DEA)
IMS- AGW
I Mi
Sh Mr
S6a
Cx S Mx
C
MME PCRF Rx Sh Telephony
HSS Application Mr’ MRF BGCF
S1- MME Gx Server
SEG S11
VoLTE UE LTE-Uu Gx
S1-U
SGW S5 PGW SGi
eNodeB
IMS Core
Figure 15: Interconnect PMN VoLTE deployment

4.2 VoLTE Interconnect

IMS interconnect is defined in 3GPP 29.165 [24].
V1.0
Official Document
Users are assumed to have attached and registered in their home network as shown in
section 3. Calls are then established between users in different PMNs across the
interconnect NNI.
The message sequences in sections 3.2.3 to 3.2.6 of this document apply, with the
difference that the IBCF and TrGW provide the interworking between the IMS network and
the NNI to the peer IMS network. The call flows in this section build on those in section 3
and document the additions/differences from sections 3.2.3 to 3.2.6 to interconnect with
peer IMS networks.
4.2.1 Basic VoLTE UE to Peer IMS Call Establishment – Originating Side
4.2.1.1 General
For calls originating on IMS and breaking out to a peer IMS network, the difference is that
the originating S-CSCF shall recognise (via internal configuration data or ENUM) that the
termination is not within this IMS network and may optionally invoke the BGCF to determine
the destination network and to invoke the IBCF. The message sequence in section 4.2.1.2
thus portrays the S-CSCF, BGCF and IBCF and may be combined with the flow in section
3.2.3.2 starting at the point at which the S-CSCF propagates the SIP INVITE message to
the terminating leg of the call.
V1.0
Official Document
BGCF IBCF/ TrGW Peer IMS

S-CSCF
Network
SIP Invite (SDP)

SIP Invite (SDP)
SIP Invite (SDP)
SIP 100 Trying
SIP 100 Trying

SIP 100 Trying
SIP 183 Progress)(SDP) SIP 183 Progress)(SDP) SIP 183 Progress)(SDP)
SIP PRACK SIP PRACK SIP PRACK
SIP 200 OK(PRACK) SIP 200 OK(PRACK) SIP 200 OK(PRACK)
SIP UPDATE (SDP) SIP UPDATE (SDP) SIP UPDATE (SDP)
SIP 200 OK(UPDATE) (SDP) SIP 200 OK(UPDATE) (SDP) SIP 200 OK(UPDATE) (SDP)
SIP 180 Ringing SIP 180 Ringing SIP 180 Ringing
SIP 200 OK (INV) SIP 200 OK(INV) SIP 200 OK(INV)
SIP ACK
SIP ACK
SIP ACK
Figure 16: Basic VoLTE UE to Peer IMS Call Establishment – Originating Side
V1.0 Page 71 of 121

Official Document
NOTE: The interaction with the PCC and the establishment of the dedicated bearer is as
shown in figure 5.
NOTE: The BGCF may not remain in the signalling path. (see clause 5.6.2 of 3GPP TS
24.229 [9]).

The S-CSCF determines that the Called-Party is not within this IMS network (i.e. ENUM
lookup/internal configuration) and optionally forwards the SIP INVITE to the BGCF.
Alternately, the S-CSCF may itself identify the correct IBCF as the egress point from this
network. The rest of this section assumes that a BGCF is invoked by the S-CSCF.
The BGCF is responsible for selecting the appropriate control element for breaking out of
this IMS network. The BGCF may use ENUM or internal configuration data to analyse the
Request-URI to determine whether the destination is in a CS network or peer IMS network.
In this case, the destination is in a peer IMS network and the BGCF is responsible for
selecting the IBCF to handle the interconnect. The Request-URI can be a TEL URI or SIP
URI but will contain an E.164 number The BGCF forwards the INVITE to the selected IBCF.
The IBCF is responsible for inter-working to the peer IMS network both in the control plane
and media plane and shall follow the procedures of 3GPP TS 29.165 [24].
The IBCF selects the outgoing route to the peer network and controls one or more TrGWs
over the Ix reference point (see 3GPP TS 29.238 [73]) to provide appropriate resources in
the media plane. The selection in the IBCF is based on DNS or internal configuration data.
Having selected a route, the IBCF shall invoke a TrGW to allocate and configure media
resources for the call. The TrGW is an IP-IP GW and serves as a border element in the
media plane in an IMS network.
The IBCF sends a SIP INVITE to the peer network, having modified the SIP headers as
described in 3GPP TS 29.165 (e.g. Record-Route, Via, other local information that should
not cross the trust boundary etc.) as well as overwriting the associated SDP to reflect the
media pin-hole newly created on the TrGW.
The IBCF remains in the control path and transits all further SIP session establishment
messages between the peer network and the home IMS network.
The IBCF transits the received 183 Progress (SDP) message followed by the associated
PRACK and 200 OK (PRACK) messages (the 183 Progress message utilizes 100rel).
The IBCF transits the UPDATE (SDP) message and associated 200 OK (UPDATE) (SDP)
message.
The IBCF uses the SDP offer/answer exchanges to modify the media pin-hole in the TrGW.
On receipt of a SIP 180 (Ringing) response from the peer network, the IBCF shall modify
the TrGW to ensure a backward transmission path if the P-Early-Media header is present
(and thus indicating that ring tone is being applied from the far end).
The 180 Ringing message is forwarded by the IBCF to the IMS network and is received by
the S-CSCF and forwarded to the VoLTE UE to indicate a ringing tone to the subscriber.
Note that the 180 Ringing message does not utilize 100rel.
When the peer network indicates that the call has been answered, the IBCF forwards the
200 OK (INVITE) message to its IMS network. This message is received by the S-CSCF
V1.0
Official Document
and forwarded to the P-CSCF. The IBCF shall ensure that duplex media can be conveyed
via the TrGW at this point.
the call has been established. The ACK is propagated through the IMS network to the IBCF
and onto the peer network.
At this stage, the IMS-CS call is established with voice traffic sent over the dedicated bearer
between the VoLTE UE and the peer network via the TrGW.
4.2.2 Basic VoLTE UE to Peer IMS Call Establishment – Terminating Side
4.2.2.1 General
For calls originating in a peer IMS network, the only difference is that the call enters the IMS
network of the target user via an IBCF rather than a P-CSCF. The IBCF invokes an I-CSCF
in order to determine the S-CSCF of the terminating user. The message sequence in
section 4.4.2.2 thus portrays only the IBCF, I-CSCF and S-CSCF and may be combined
with the flow in section 3.2.4.2 and starting at the point at which the S-CSCF is initially
invoked with the SIP INVITE message.
V1.0
Official Document
I-CSCF IBCF/ TrGW Peer IMS

S-CSCF
Network

SIP Invite (SDP)
SIP Invite (SDP)

SIP 100 Trying

SIP 100 Trying
SIP 183 Progress (SDP) SIP 183 Progress (SDP) SIP 183 Progress (SDP)
SIP PRACK SIP PRACK
SIP 200 OK (PRACK)

SIP 200 OK (PRACK)
SIP 200 OK (UPDATE) (SDP) SIP 200 OK (UPDATE) (SDP)
SIP 200 OK (INV) SIP 200 OK (INV) SIP 200 OK (INV)
SIP ACK
SIP ACK
Figure 17: Basic VoLTE UE to Peer IMS Call Establishment – Terminating Side
V1.0 Page 74 of 121

Official Document
NOTE: The interaction with the PCC and the establishment of the dedicated bearer is as
shown in figure 6.
INVITE transaction.

The peer IMS Network initiates the call establishment by sending a SIP INVITE to the IBCF.
The IBCF shall follow the procedures of 3GPP TS 29.165 [24].
The IBCF invokes the TrGW over the Ix reference point (see 3GPP TS 29.238 [73]) to
allocate media resources for the call.
The IBCF may modify the content of a number of the SIP headers in the SIP INVITE that
are not required in the terminating network (e.g. Record-Route, Via etc.) as described in
3GPP TS 29.165 as well as overwriting the associated SDP to reflect the media pin-hole
newly created on the TrGW.
The target user may be identified via a SIP or TEL-URI. In the former case, the SIP URI
may contain a “user=phone” URI parameter and shall contain an E.164 number.
The IBCF invokes the I-CSCF to enable the appropriate S-CSCF for the target user to be
found.
The I-CSCF interrogates the HSS to identify the S-CSCF where the user is registered and
forwards the INVITE to the S-CSCF. The S-CSCF invokes any VoLTE services as triggered
by the initial filter criteria and routes the SIP INVITE to the AS and terminating P-CSCF as
described in section 3.2.4.
Call establishment proceeds as in section 3.2.4 and the IBCF forwards subsequent call
establishment messages from the S-CSCF to the peer IMS network potentially modifying
the content of SIP headers (e.g. Record-Route, Via, other local information that should not
cross the trust boundary etc.) as described in TS 29.165 and modifying any SDP to reflect
the media pin-hole in TrGW.
The IBCF transits the received 183 Progress (SDP) message followed by the associated
PRACK and 200 OK (PRACK) messages (the 183 Progress message utilizes 100rel).
The IBCF transits the UPDATE (SDP) message and associated 200 OK (UPDATE) (SDP)
message.
The IBCF uses the SDP offer/answer exchanges to modify the media pin-hole in the TrGW.
If the P-Early-Media header is present in the 180 (Ringing) message, then the IBCF shall
ensure that backward media (e.g. ring tone, progress indications) are conveyed via the
TrGW. This message does not utilize 100rel.
When the IMS network indicates that the call has been answered, the IBCF shall ensure
that duplex media can be conveyed via the TrGW at this point.
between the VoLTE UE and the peer IMS network via the TrGW.
V1.0
Official Document
4.2.3 Basic VoLTE UE to Peer IMS Call Teardown - Initiated

4.2.3.1 General
This section describes the teardown of an interconnect call where the release is initiated by
the VoLTE UE in this IMS network. The message sequence in section 4,2.3.2 portrays the
S-CSCF, BGCF, IBCF and Peer IMS Network and may be combined with the flow in figure
7 and starting at the point at which the S-CSCF propagates the SIP BYE message to the
terminating leg of the call.
V1.0
Official Document
BGCF IBCF / TrGW Peer IMS

S-CSCF
Network
IMS Signalling over IP
SIP BYE
SIP BYE
SIP BYE
SIP 200 OK
SIP 200 OK
SIP 200 OK
Dedicated bearer has been releaseed.
Figure 18: Basic VoLTE UE to Peer IMS Call Teardown – initiated
V1.0 Page 77 of 121

Official Document
NOTE: The interaction with the PCC and the release of the dedicated bearer is as shown in
figure 7.
NOTE: This figure shows a single offer/answer exchange supporting preconditions and
utilising the segmented status type as defined in RFC 3312 [71]. A further offer/answer
exchange could be used if required (e.g. if multiple codecs were returned in the SDP
answer and the originating side wishes to specify a single codec). .
NOTE: The BGCF may not be in the signalling path. (see clause 5.6.2 of 3GPP TS 24.229
[9]).
The call teardown is initiated in this IMS network via the VoLTE UE as described in section
3.2.6.2.
The S-CSCF propagates the SIP BYE message to the BGCF and onto the IBCF.
The IBCF releases the resources in the TrGW and sends a SIP BYE to the Peer IMS
network (modifying SIP headers as described in TS 29.165).
On receipt of the 200 OK (BYE), the IBCF forwards the message to the IMS network
(modifying headers as described in TS 29.165).
At this stage, the interconnect call is released and the dedicated bearer has been torn down
as described in section 3.2.6.2.
4.2.4 Basic VoLTE UE to Peer IMS Call Teardown - Received
4.2.4.1 General
This section describes the teardown of an interconnect call where the release is initiated in
the Peer IMS network. The message sequence in section 4.2.4.2 portrays the S-CSCF,
IBCF and Peer IMS Network and may be combined with the flow in figure 8 and starting at
the point at which the S-CSCF receives the SIP BYE message from the terminating leg of
the call.
V1.0
Official Document
I-CSCF IBCF / TrGW Peer IMS

S-CSCF
Network
SIP BYE
SIP BYE
SIP 200 OK
200 OK
Dedicated bearer has been released
Figure 19: Basic VoLTE UE to Peer IMS Call Teardown - Received
V1.0 Page 79 of 121

GSM Association Confidential
Official Document
NOTE: The interaction with the PCC and the release of the dedicated bearer is as shown in
figure 8.
The Peer IMS Network initiates the call teardown by sending a SIP BYE to the IBCF.
The IBCF releases the resources in the TrGW and forwards the SIP BYE to the S-CSCF
(modifying headers as described in TS 29.165). The call release and release of the
dedicated bearer at the UE side is as described in section 3.2.6.3.
The S-CSCF sends a 200 OK to the BYE to the IBCF.
The IBCF sends the SIP 200 OK to the Peer IMS network (modifying SIP headers as
described in TS 29.165).
At this stage, the interconnect call is released and the dedicated bearer has been torn down
as described in section 3.2.6.3.
4.3 Bi-lateral Interconnect
In this model, an IMS network has an IP route to its peer network. In architectural terms, the
interconnect is realised via a point-point route between the border elements (IBCF/TrGW) in
each network as described in 3GPP TS 29.165 [24]). The interconnect interface is realised
via the Ici/Izi interfaces in the control/media plane respectively. In this scenario, both
operators are responsible for screening of SIP/SDP traffic in accordance with the bi-lateral
interconnection agreement.
4.3.1 Physical Configuration of Bi-lateral Interconnect

The physical interconnect may be realised via a direct leased line (e.g. Frame Relay or ATM
based) or VPN connection (e.g. IPSEC connection) over the public internet. In either case, it
is important that there is a SLA between the parties and that the connection is able to meet
the requirements of that SLA. Such a SLA has been defined in PRD AA.80 [75] for IPX.
However, the requirements for interconnect as defined for IPX are equally applicable to a
bilateral interconnect.
4.3.2 Usage of ENUM/DNS
In order to be able to route calls between peer IMS networks, with a bi-lateral interconnect,
translation between an E.164 number to SIP-URI and SIP-URI to IP address is required.
This can be done via internal configuration data within network nodes or via ENUM.
When using ENUM, an ENUM look up within the the originating IMS network (e.g. from
TAS, S-CSCF or BGCF) is used to retrieve the domain name of the terminating network.
The Domain name is then together with configuration data used to determine the correct
IBCF and the optimum IP Route to the target IMS Peer network.
In general, due to a combination of n/w transformation (i.e. porting of subscriber numbers
from CS to IMS) and number portability, the correlation of destination telephone number and
location (e.g. CS Network, peer IMS network etc.) is non-trivial and liable to be in a state of
flux. The ENUM look up is constantly influenced by activities in other networks and thus
may be centralised and managed by a trusted 3rd party (i.e. Carrier ENUM). Usage of
ENUM is described in GSMA PRD IR.67 ([52]).
V1.1 Page 80 of 121

Official Document
4.3.3 Usage of Session Border Controllers

4.3.3.1 General
The combination of an IBCF and TrGW realises a Session Border Control (SBC) at the
edge of an IMS network and responsible for managing the Ici/Izi interfaces. The SBC
checks the SIP/SDP to ensure conformance to the bi-lateral interconnection agreement.
4.3.3.2 Control Plane
The IBCF manages the control plane at an interconnect. Clause 6 of 3GPP TS 29.165 [24]
describes procedures to be performed in the control plane although these procedures may
be modified/extended for a given bi-lateral interconnection agreement. SBCs can also
typically modify SIP signalling/headers in order to facilitate interworking between different
implementations of SIP.
4.3.3.3 Media Plane
The TrGW manages the media plane at an interconnect, under the control of an IBCF.
Clause 7 of 3GPP TS 29.165 [24] describes procedures to be performed in the medial
plane. These include the creation of media pin holes to enable NAT of the media packets as
well as policing the media flows to ensure that they are consistent with the associated
control plane signalling and related bi-lateral interconnection agreement. DiffServ marking of
media packets is also performed at the TrGW (under the control of the IBCF) – see section
3.7 for further information on Diffserv codepoints. Based on network policy and the bi-
lateral agreement with the peer network, the TrGW may also perform transcoding. It is
recommended that transcoding be avoided in the TrGW if possible.
4.4 IPX-Based Interconnect
The IPX is defined in GSMA PRD IR.34 [48] and provides an IP interconnect network to
enable connectivity between Operator’s IMS networks. Thus, the use of an IPX requires
only a 1:1 relationship to be configured by an N/W Operator rather than 1:N relationships
and many bilateral inter-connects to other peer networks. Therefore, an IPX based
interconnect is the recommended option. In this scenario, the IPX can perform screening of
SIP/SDP traffic on behalf of the Operator in accordance with the per-operator
interconnection agreement.
4.4.1 Configuration of IPX-based Interconnect

IR.34 defines three modes for IPX based interconnect, namely :-
 Transport-Only Connectivity - , a bilateral agreement between Service providers
using the IPX transport layer with guaranteed QOS (IPX-edge to IPX-edge).
 Bilateral Service Transit Connectivity - , a bilateral agreement between Service
providers using the IPX Proxy Functions and IPX transport layer with guaranteed
QOS end-end. This model provides the opportunity to include service based
interconnect charging to the transport charging of the previous model
 Multilateral Service Hub Connectivity – this model provides multi-lateral interconnect
with guaranteed end-end QOS and service based charging. Traffic may be routed
from one Service Provider to many destinations via a single agreement with the IPX
provider. The hub functionality is provided by IPX Proxies.
Due to the complexity associated with network transformation and number portability, as
well as the emergence of multi-services via IMS, it makes sense in the long term to utilise
the multi-lateral service hub IPX option. This option simplifies the interconnect mapping for
the relatively many IPX clients as well as insulating the IPX clients from being directly
impacted due to numbers being ported between destination networks as well as the
introduction of new services.
V1.1 Page 81 of 121

Official Document
4.4.2 Usage of ENUM/DNS

This is as for section 4.3.2.. In some cases, it may be possible to dispense with the ENUM
dip the originating network due to all off-net calls being handed to the IPX as default. Given
the flux of mapping destination addresses to destination networks, centralization of the
ENUM data for inter-operator look-up is recommendedso as to avoid changes rippling
through local configuration data as numbers become relocated (i.e. utilising Carrier ENUM).
Usage of ENUM is described in GSMA PRD IR.67 ([52]).
4.4.3 Usage of Session Border Controllers

4.4.3.1 General
As for section 4.4.2.1,with the SBC (IBCF/TrGW) interfacing between then Operator and
the IPX. The Operator may choose to outsource the deployment of the SBC to the IPX
Provider, deploy an SBC within its own network or combine both.
Where an IPX deploys an SBC, the screening of SIP/SDP by the IPX on a per operator
basis provides a more scalable approach than the bi-lateral case as the IPX Provider
manages the inter-Operator connectivity. It is an operator policy decision as to the extent
that its own SBCs perform SIP/SDP screening when such screening is being performed by
the IPX.
4.4.3.2 Control Plane
As for section 4.4.2.2 but done on a per operator basis
4.4.3.3 Media Plane
As for section 4.4.2.3 but done on a per operator basis. .
4.5 CS Interconnect
It is also possible that a given endpoint resides in the CS Network of another operator. In
this case, the call flows would be identical to those in section 3.3. This scenario is
mentioned here only for completeness.
4.6 Charging
Charging within each network is as for section 3.15. In addition, each peer at an
interconnect will also perform route accounting at the interconnect point to enable the
relative call total in each direction to be billed appropriately. This mechanism is identical to
the current mechanisms used at operator interconnect points in the CS Network.
V1.1 Page 82 of 121

Official Document
5 VoLTE Implementation - Roaming

This section builds on the single PMN scenario in section 3 and considers inter-operator
VoLTE roaming between PMNs. The roaming user attaches in a visited network and
registers for IMS services in its respective home network. As recommended in GSMA PRD
IR.65 [51], it is assumed that the local breakout model is used whereby the full EPC and P-
CSCF are located in the visited network for the roaming user. The P-CSCF then interacts
with the S-CSCF and TAS in the home network for the provision of VoLTE services. Local
breakout also places requirements on the Diameter routing between the home and visited
networks to convey the S6a and optionally S9 interfaces respectively.
Note:- S9 may be optionally used to convey Rx messages between the home and visited
networks for dynamic policy control and the subsequent text and message flows assume
dynamic policy control. The deployment of the S9 interface is determined on a bi-lateral
operator agreement. It is also possible to use static policy control where the local
configuration data is stored in the V-PCRF, in which case the S9 interaction does not occur.
However, in this document, all message flows show S9 being utilised.
Note: The subsequent text and message flows in this section assume a separate P-CSCF
and IBCF. However, it is also noted that the P-CSCF may be co-located with an IBCF (i.e.
combined access and network side SBC).
5.1 General
The VoLTE architecture for a PMN deployment supporting roaming is shown in Figure 20.
IMS Core
Telephony
MRF
Server
I Mx
Mr
Cx S Mi
C
P- CSCF/
IMS- AGW
S6a Sh
Ici/Izi
X2 ENUM
Rx Rx
SGi
Diameter Agent
VoLTE UE MME H- PCRF S9
(DRA/ DEA)
(Home) LTE-Uu
S1-MME Gx
SEG S11
Gx
S1-U
eNodeB SGW S5 PGW
HPLMN
ENUM / IPX
Service Diameter App ENUM
Provider ID=0 ENUM IPX
Server
Network
ENUM
Mw
Mw
VPLMN Mw
P- CSCF/
S6a Diameter Agent
Rx IMS- ALG/ Mw I-CSCF IBCF/ TrGW
S9 ( DRA/ DEA)
IMS- AGW
Mi
a Mx
MME V- PCRF Rx
BGCF
S1- MME Gx
SEG S11
VoLTE UE LTE-Uu Gx
(Roaming)
S1-U
SGW S5 PGW SGi
eNodeB
IMS Core
Figure 20: Roaming VoLTE deployment
V1.1 Page 83 of 121

Official Document
5.2 VoLTE Roaming Basic Call Flows
The VoLTE basic call flows are in accordance with 3GPP specifications for E-UTRAN/EPC,
IMS, and PCC. Please refer to 3GPP TS 23.401 [6], 3GPP TS 23.228 [5], and 3GPP TS
23.203 [4] respectively for further detailed information.
The following sub-sections define the additional requirements for the VoLTE service.
References to specific functionality within GSMA PRDs (e.g. IR.92, IR.65, IR.80) and 3GPP
specifications will be made within each sub-section. It is assumed that there is a roaming
agreement between the home network of the UE and the visited network in which the UE is
and appropriate configuration data has been set up in each network.
In particular, this section will highlight the key differences from the corresponding flows for a
non-roaming UE as described in section 3.2 and (where possible) shall refer to text that is
common for roaming and non-roaming scenarios.
5.2.1 Roaming VoLTE UE Attachment and IMS Registration
5.2.1.1 General
As section 3.2.1.1, apart from the UE roaming and thus attaching to a visited network.
‎Figure 3 show the message sequences for the LTE Attachment and IMS Registration for a
roaming UE.
V1.1 Page 84 of 121

Official Document
PCRF PCRF HSS

VoLTE UE eNB MME SGW PGW (V) (H) (H)
RRC Connection Request
RRC Connection Setup
Attach Req
Attach req - Uplink NAS Transport
RRC Connection Setup complete
[ Dedicated NAS ] [ NAS PDU ]
[ PDN Connectivity Request] [ EPS SM: PDN Connectivity Request ]
Authentication/Security
Update Location Request
Update Location Answer
Create Session Request
IMS-APN; QCI:5; Create Session Request

ARP; APN AMBR
IMS-APN; QCI:5;
ARP; APN AMBR
CCR CCR
IP@; QCI:5; ARP; IP@; QCI:5; ARP;

APN AMBR APN AMBR
CCA CCA
Create Session Response QCI:5; ARP; QCI:5; ARP;

E-RAB Setup APN AMBR; APN AMBR;
E-RAB Setup Request IP default rule TFT IP default rule TFT
QCI:5; ARP; IP@; QCI:5; ARP;
RRC Connection Reconfiguraton Create Session Response APN AMBR
UE AMBR
[ NAS PDU ] IP@; QCI:5; ARP;
[ NAS PDU ]
[ EPS SM: Activate Default APN AMBR
[ Activate Default Bearer Request ]
Bearer Request ] [IMS-APN; IP@; ]
[ QCI:5; APN AMBR]
RRC Connection E-RAB Setup

Reconfiguration Complete E-RAB Setup Response
Uplink Direct Transfer

[ NAS PDU ]
[ EPS SM: Activate Default Bearer Accept ]
Attach complete - Uplink NAS Transport
[ NAS PDU ]
[ Activate Default Bearer Accept ]
Modify Bearer Request
Modify Bearer Response
Default Bearer Established for IMS Signalling IMS Signalling over IP
Figure 21: Roaming VoLTE UE Attachment
V1.1 Page 85 of 121

Official Document
P-CSCF PCRF IBCF IBCF PCRF I-CSCF S-CSCF TAS (H)

VoLTE UE HSS (H)
(V) (V) (V) (H) (H) (H) (H)
SIP Register
SIP Register
SIP Register
SIP Register
UAR
UAA
SIP Register
MAR
MAA
401 Unauthorised 401 Unauthorised 401 Unauthorised 401 Unauthorised 401 Unauthorised
SIP Register

SIP Register
UAR
UAA
SIP Register
SAR
SAA
200 OK (REGISTER)
200 OK (REGISTER)
200 OK (REGISTER) SIP Register
200 OK (REGISTER) 200 OK (REGISTER)
UDR
AAR
AAR UDA
200 OK
AAA AAA
SIP Subscribe
200 OK (Subscribe)
SIP Subscribe SIP Subscribe SIP Subscribe SIP Subscribe
SIP Notify
200 OK (Subscribe) 200 OK (Subscribe) 200 OK (Subscribe)
200 OK (Subscribe) 200 OK (Notify)
SIP Subscribe
SIP Subscribe SIP Subscribe SIP Subscribe SIP Subscribe
200 OK (Subscribe) 200 OK (Subscribe) 200 OK (Subscribe) 200 OK (Subscribe) 200 OK (Subscribe)
SIP Notify SIP Notify SIP Notify SIP Notify

200 OK (Notify) 200 OK (Notify) 200 OK (Notify) 200 OK (Notify)
SIP Notify SIP Notify SIP Notify SIP Notify SIP Notify
200 OK (Notify) 200 OK (Notify) 200 OK (Notify) 200 OK (Notify) 200 OK (Notify)
Figure 22: Roaming VoLTE UE Initial IMS Registration message sequence
V1.1 Page 86 of 121

Official Document
NOTE: The Diameter Agents have not been included in this message sequence, although
Diameter messages shall route via the Diameter Agents in the home and visited networks.
Usage of Diameter Agents is described in section ‎3.6.
5.2.1.3.1 Roaming VoLTE UE Attach
As section 3.2.1.3.1, with differences specific to the roaming scenario described in this
section.
The UE has a list of foreign/visited networks to which it can attach. The UE initiates an
attach to the eNodeB as in the single network scenario.
The MME shall validate that the UE is permitted to attach and shall send the Update
Location to the HSS in the UE’s home network. This message shall be sent via the
DRA/DEA in each network.
As in the single network case, if the initial LTE attach is for VoLTE service, there shall not
be an APN included in the attach request. The default APN configured in the HSS shall be
set as the IMS-APN, and the HSS returns the IMS-APN name for establishment of the
default bearer. If the initial LTE attach is not for VoLTE service (e.g. internet), the UE shall
specify the IMS-APN in a subsequent PDN connection request as specified in clause 5.10.2
of 3GPP 23.401 [6]) and defined within IR.88 [53]. In either case, the APN-OI information is
inserted by the MME which enables the UE to attach to a P-CSCF in the visited network.
The MME initiates a Create Session Bearer request to the SGW to create a default bearer
for VoLTE IMS signalling as for the non-roaming case.
The PGW allocates an IP Address for the UE and utilises dynamic PCC to initiate a Credit
Control Request to the PCRF (in the visited network) to obtain the default PCC rules for the
default bearer to be used for IMS signalling. In the roaming scenario, the PCRF in the
visited network notes that the IMSI is for a visiting UE and sends the CCR message to the
PCRF in the corresponding home network. This message is used to request default QOS
parameters from the home network and also to establish the S9 interface between the
visited and home network PCRFs (see 3GPP TS 29.215 [27]). The home PCRF responds
with a CCA message and returns the default QOS parameters and indicates that S9 is now
established. The S9 interface shall be used to tunnel subsequent Rx messages between the
PCRFs for session establishment and teardown. The CCA is forwarded to the PGW by the
visited network PCRF.
Note :- The above description of the S9 interface assumes that dynamic policy rules must
be exchanged between the home and visited network. It is also possible (as documented in
GSMA IR.88 section 3.3. to use static policy control in which case Rx need not be tunnelled
between the home and visited networks.
On receipt of the CCA from the visited network PCRF into the PGW, the default bearer is
established as described for the non-roaming case.
At this stage, the roaming VoLTE UE is attached to the visited network via a default bearer
that is established for IMS Signalling.
5.2.1.3.2 Roaming VoLTE UE Initial IMS Registration
section.
The roaming VoLTE UE initiates as SIP REGISTER to the P-CSCF, which is located in the
visited network.
The P-CSCF receives the SIP REGISTER request from the UE and inserts a Path header
with a SIP-URI identifying the P-CSCF for routing, a P-Charging-Vector header with the icid-
V1.1 Page 87 of 121

Official Document
value, a P-Visited-Network-ID to identify the P-CSCF's network domain. In the roaming

case, the domain in Request-URI is recognised as being a foreign domain and the P-CSCF
thus forwards the request to the IBCF via local configuration data or DNS as described in
clause 5.2.2.1 of 3GPP TS 24.229 ([9]). .
The visited network IBCF passes the REGISTER message to its peer in the home network
of the roaming UE. The IBCF may modify the SIP headers in the REGISTER message (e.g.
for topology hiding) as described in 3GPP TS 29.165 [24]. In particular, the IBCF shall
modify the Path header to add its own SIP-URI.
The REGISTER is received by the IBCF in home network of the roaming UE. The home
network IBCF modifies the SIP REGISTER in a similar manner to its peer in the visited
network. The IBCF passes the REGISTER to the I-CSCF.
As described in section 3.2.1.3.2, the I-CSCF queries the HSS to perform validation checks
on the user identity and obtain the identity of the appropriate S-CSCF. The I-CSCF invokes
the S-CSCF which retrieves authentication data from the HSS and sends a 401
Unauthorised response to the REGISTER.
The 401 Unauthorised response is transited via the IBCFs to the P-CSCF which modifies
the response as described in section 3.2.1.3.2 and creates a temporary set of security
associations prior to passing the response to the roaming UE.
As described in section 3.2.1.3.2., the UE re-sends the SIP REGISTER populated with the
Authiorization header.
As in section 3.2.1.3.2, the P-CSCF checks the temporary security associations, and
verifies the security related information received from the UE prior to forwarding the
REGISTER to the IBCF.
The IBCF sends the REGISTER to its peer in the UE’s home network which invokes the I-
CSCF. The I-CSCF uses the User Authorization Request message to retrieve the S-CSCF
name stored within the HSS, and forwards the request to the relevant S-CSCF.
The S-CSCF validates the security parameters in the REGISTER message and then
downloads user profile data from the HSS. The S-CSCF binds the P-CSCF address to the
user’s contact (to be used to route future requests to the user) and returns a 200 OK
response containing a SERVICE-ROUTE header.
The 200 OK (REGISTER) response traverses the IBCFs and is conveyed to the P-CSCF. In
accordance with 3GPP TS 24.229, the IBCFs must not modify the SERVICE-ROUTE
header and pass it on unchanged as received from the S-CSCF.
As in section 3.2.1.3.2, the P-CSCF changes the temporary set of security associations to a
newly established set of security associations. It protects the 200 OK with these
associations and sends the 200 OK to the VoLTE UE. All future messages sent to the UE
will be protected using the security associations.
The P-CSCF optionally sends an AAR message to the PCRF to perform application binding
to the default bearer (i.e. the P-CSCF is requesting to be informed in the event of the default
bearer being lost/disconnected in order to trigger an IMS de-registration). In the roaming
scenario, the visited network P-CSCF passes the message to the visited PCRF and onto
home PCRF over the S9 interface if implemented. The AAA is sent back from the home
network PCRF over S9, if implemented, to the visited network PCRF. The visited network
PCRF sends the AAA onto the P-CSCF. Note that if application session binding is not
performed, then IMS learns of the loss of the default bearer by other means (e.g. timeout
when attempting to send a SIP message to the UE).
As in section 3.2.1.3.2, the UE changes the temporary security association to a newly
established set of security associations that will be used for further messages to the P-
CSCF.
V1.1 Page 88 of 121

Official Document
As in section 3.2.1.3.2, 3rd party registration occurs from the S-CSCF to the TAS and the
UE, P-CSCF and TAS shall subscribe to the registration event package to be notified of any
change of registration state of the public user identity and be notified of the registration state
via a SIP NOTIFY message.
The SUBSCRIBE and NOTIFY messages between the UE and P-CSCF to the S-CSCF
shall be conveyed via the pair of IBCFs.
The VoLTE UE is now registered with the IMS network for VoLTE services, with SIP
signalling being transported over the default EPC bearer. The signalling path from the UE to
the S-CSCF in the home network traverses the P-CSCF and IBCF pair (as determined by
the SERVICE-ROUTE header returned in the 200 OK (REGISTER) response.
V1.1 Page 89 of 121

Official Document
5.2.2 Roaming VoLTE UE Initiated Detach and IMS Deregistration

5.2.2.1 General
As for the single network case, a roaming VoLTE UE shall automatically deregister from IMS before performing an LTE Detach, if the UE is not
moving to another access technology that supports Voice over IMS.
P-CSCF PCRF IBCF IBCF PCRF I-CSCF S-CSCF TAS (H)
VoLTE UE HSS (H)
(V) (V) (V) (H) (H) (H) (H)
SIP Register (exp =0) SIP Register (exp=0) SIP Register (exp=0) SIP Register (exp=0) UAR
UAA
SIP Register (exp=0)
SAR
SAA
Notify (reg event)
Notify (reg event) Notify (reg event) Notify (reg event) Notify (reg event)
200 OK (Notify) 200 OK (Notify) 200 OK (Notify) 200 OK (Notify) 200 OK

Notify (reg event) Notify (reg event) Notify (reg event)
200 OK (Notify) 200 OK (Notify) 200 OK (Notify)
STR STR
STA STA
200 OK (REGISTER)
200 OK (REGISTER)
200 OK (REGISTER)
200 OK (REGISTER)
200 OK (REGISTER)
Figure 23: Roaming VoLTE UE IMS Deregistration message sequence
V1.1 Page 90 of 121

Official Document
PCRF PCRF HSS

VoLTE UE eNB MME SGW PGW
(V) (H) (H)
Detach Request

Detach Acceot
CCR CCR
Release Access Bearer

Request
CCA CCA
Release Access Bearer

Response
UE Context Release Command
RRC Connection Release
UE Context Release Complete
Default Bearer is Deleted and UE is not LTE Attached
Figure 24: Roaming VoLTE UE Initiated Detach
V1.1 Page 91 of 121

Official Document

5.2.2.3.1 IMS Deregistration (Roaming)
section.
The SIP REGISTER (exp=0) message is forwarded to the I-CSCF via the P-CSCF and
IBCFs. The I-CSCF uses the User Authorization Request to retrieve the S-CSCF name
stored within the HSS and forwards the REGISTER (exp=0) to the S-CSCF.
The S-CSCF informs the HSS of the de-registration and the HSS shall keep the S-CSCF
name associated to the public user identity for future use and to enable unregistered
services to be applied (e.g. routing of a terminating voice call to voicemail) as in the single
network case.
The S-CSCF shall send a SIP NOTIFY to the P-CSCF, TAS and VoLTE UE to notify them
of the change of the registration state.
The P-CSCF, on being notified of the de-registration, optionally sends a STR message to
the PCRF to remove the application binding to the default bearer (if application session
binding was done previously). In the roaming scenario, the STR is sent over the S9
interface, if implemented, to the PCRF in the home network. The home network PCRF
responds with a STA over S9 if implemented to the visited PCRF. The visited PCRF sends
the STA to the P-CSCF. The P-CSCF shall also remove the security associations that
were established between the P-CSCF and the UE as in the single network case.
The S-CSCF shall send a 200 OK to acknowledge the de-registration.
The P-CSCF shall forward the 200 OK to the UE.
On receiving the 200 OK responses, the UE shall remove all the registration details for the
Public User Identity and delete the stored security associations. The UE shall consider the
subscription to the registration event package as cancelled.
The roaming VoLTE UE is now de-registered from the IMS network for VoLTE services, no
further SIP signalling is being transported over the default EPC bearer.
5.2.2.3.2 Roaming VoLTE UE Detach
The detach procedure is as described in section 3.2.2.3.2. Where the S9 interface is
implemented, the CCR message from the visited network PCRF is forwarded to the home
PCRF over S9. The CCR message shall also result in the termination of the S9 session
between the visited and home PCRFs (see TS 29.215 [27]). The home PCRF shall respond
with a CCA to the visited PCRF.
The default bearer is released as in the single network case and the VoLTE UE is no longer
attached to the network.
5.2.3 Roaming VoLTE UE to VoLTE Call Establishment – Originating Side
5.2.3.1 General
As section 3.2.3.1 with differences specific to the roaming scenario described in this section.
The essential differences from the single network scenario are that the Mw reference point
is conveyed via the IBCFs between the P-CSCF in the visited network and the S-CSCF in
the home network. Where S9 interface is deployed, the Rx messages related to the
establishment of the dedicated bearer are sent via the visited and home network PCRFs.
V1.1 Page 92 of 121

Official Document
The interactions between the visited network PCRF and the EPC/UE are identical to the
single network case. Therefore, the message flows in this section will show the SIP
signalling and Diameter Rx and Gx signalling only – the latter to indicate the points at which
the EPC is invoked. .
As for the single network case, a roaming VoLTE UE, shall perform call establishment by
using the IMS network. The IMS Signalling shall be sent over the default bearer, and a new
dedicated bearer shall be dynamically established for the voice traffic.
V1.1 Page 93 of 121

Official Document
PCRF P-CSCF/ IMS- PCRF

VoLTE UE PGW IBCF (V) IBCF (H) S-CSCF TAS
(v) ALG/ IMS-AGW (H)
SIP Invite (SDP)
SIP 100 Trying
SIP Invite (SDP)
SIP Invite (SDP) SIP Invite (SDP) SIP Invite (SDP)
SIP 100 Trying

SIP 100 Trying
SIP Invite (SDP)
SIP 100 Trying
SIP Invite (SDP)
SIP 100 Trying

AAR SIP 183 Progress (SDP) SIP 183 Progress (SDP) SIP 183 Progress (SDP) SIP 183 Progress (SDP)
AAR
RAR AAA
RAA AAA

SIP PRACK SIP PRACK SIP PRACK SIP PRACK SIP PRACK
SIP 200 OK (PRACK) SIP 200 OK (PRACK) SIP 200 OK (PRACK) SIP 200 OK (PRACK) SIP 200 OK (PRACK)
SIP UPDATE (SDP) SIP UPDATE (SDP) SIP UPDATE (SDP) SIP UPDATE (SDP) SIP UPDATE (SDP) SIP UPDATE (SDP)
SIP UPDATE (SDP)
SIP 200 OK (UPDATE) (SDP) SIP 200 OK (UPDATE) (SDP) SIP 200 OK (UPDATE) (SDP) SIP 200 OK (UPDATE) (SDP) SIP 200 OK (UPDATE) (SDP)
SIP 180 Ringing

SIP 200 OK (INV) SIP 200 OK IINV)

SIP 200 OK (INV) SIP 200 OK (INV) SIP 200 OK (INV) SIP 200 OK (INV)
AAR
AAR
RAR AAA
RAA AAA
SIP ACK SIP ACK SIP ACK SIP ACK SIP ACK
SIP ACK
VoLTE voice traffic via

dedicated bearer
Figure 25: Roaming VoLTE UE to VoLTE UE Call Establishment - Originating Side message sequence
V1.1 Page 94 of 121

Official Document
answer (uplink & downlink configuration). It is also possible to invoke the PCRF twice, i.e.
on receipt of both the SDP Offer (downlink configuration) and SDP Answer (uplink
configuration). Both options are valid – see 3GPP TS 29.213 ([79]) annex B.
is not shown.
As section 3.2.3.3 with roaming specific differences highlighted in this section.
The P-CSCF will also invoke the IMS-AGW over the Iq reference point (see TS 23.334 [74])
to provide appropriate resources in the media plane as for the single network scenario if an
IMS ALG/AGW is deployed.
The P-CSCF may modify the SDP as for the single network case (if applicable) and
forwards the SIP INVITE to the S-CSCF. In the roaming case, the S-CSCF is resident in the
home network of the user. Therefore, the INVITE traverses the pair of IBCFs in the visited
and home networks respectively prior to being received by the S-CSCF. Each IBCF shall
invoke its respective TrGW to allocated media resources for the session and shall modify
the SDP in accordance with the newly created media pin-holes. The IBCFs shall also modify
the SIP headers as described in 3GPP TS 29.165.
On receipt of the INVITE, the S-CSCF behaves as for the single network case and invokes
the TAS apply VoLTE supplementary services prior to invoking the terminating leg of the
call. In this case, the S-CSCF determines that the Called-Party is within the home network
(i.e. ENUM lookup/internal configuration) and routes the SIP INVITE to the I-CSCF to
determine the terminating S-CSCF of the Called-Party (see section 3.2.4 for terminating call
establishment to a non-roaming VoLTE UE).
The called party's VoLTE UE sends a 183 Progress (SDP) message with the SDP answer.
This is received by the S-CSCF and forwarded to the visited network P-CSCF via the IBCFs
in the home and visited networks. On receipt of a SIP 183 Progress response containing the
SDP answer, each IBCF shall modify its TrGW to reflect the SDP answer and update the
SDP answer to reflect the respective media pin-holes in the TrGWs. The 183 Progress
(SDP) message is forwarded to the P-CSCF.
The SDP answer in the 183 Progress response shall indicate a single voice codec and that
QOS preconditions are required but not yet met at the terminating side. As in the single
network scenario, the P-CSCF uses the SDP answer to configure the IMS-AGW (if
deployed) and sends the Authorize/Authenticate-Request message to the visited network
PCRF with the related updated service information (IP address, port numbers, information
on media-type) specifying the access facing IP address of the IMS-AGW.
The PCRF in the visited network sends the AAR message to its peer in the home network of
the user via the S9 reference point if implemented. In this case, the home network PCRF
authorises the request and responds to the visited PCRF with an Authorize/Authenticate-
Answer (AAA) and both PCRFs associate the service information to the stored subscription
related information containing the information about the allowed service(s), QoS information
and PCC Rules information. The visited network PCRF identifies the affected IP-CAN
session (e.g. default bearer) that has been established during the LTE Attach procedure,
V1.1 Page 95 of 121

Official Document
and initiates a Re-Auth-Request to the PGW to initiate the creation of a dedicated bearer as
described in the single network scenario.
The PGW acknowledges the Re-Auth-Request to the visited network PCRF, which then
acknowledges the Authorize/Authenticate-Request message sent from the P-CSCF. At this
point the IMS SIP session and the dedicated bearer used for voice are bound together via
PCC.
VoLTE media as described in the single network scenario.
The P-CSCF forwards the SIP 183 Progress response to the VoLTE UE. This message
shall also utilize 100rel and the UE shall generate a PRACK which is transited to the
terminating side of the call via the IBCFs .
A 200 OK (PRACK) is received from the terminating side of the call via the IBCFs.
The VoLTE UE now sends a SIP UPDATE message containing a new SDP offer as in the
single network case. This message is transited via the P-CSCF to the S-CSCF via the
IBCFs and onto the terminating leg of the call.
A SIP 200 OK (UPDATE) message containing the SDP answer is sent from the terminating
leg of the call to the S-CSCF and onto the VoLTE UE via the P-CSCF and IBCFs.
The terminating UE is now alerted and a SIP 180 Ringing message is sent from the
terminating leg of the call. This message is received by the S-CSCF and sent through to the
originating UE via the pair of IBCFs and P-CSCF. This message does not utilize 100rel. The
P-Early-Media header is not present and so the UE will generate local ring tone to the
subscriber.
When the called party's VoLTE UE has answered the call, it sends a 200 OK to the calling
party VoLTE UE. This is received by the S-CSCF and forwarded to the P-CSCF via the
IBCFs. Each IBCF shall ensure that duplex media can be conveyed via their respective
TrGWs.
The P-CSCF sends an AAR message to the visited PCRF to enable the uplink and downlink
media flows. This message is conveyed to the home PCRF over S9 if implemented. In this
case, the home PCRF authorizes the request and responds with an AAA message to the
visited PCRF which invokes the P-GW to enable the media flows. The visited network
PCRF sends the AAA message to the P-CSCF. As in the single network scenario, the P-
CSCF(IMS-ALG) also invokes the IMS-AGW (if deployed) to ensure that duplex media can
be conveyed via IMS-AGW at this point.
The P-CSCF forwards the SIP 200 OK (INVITE) to the VoLTE UE.
The VoLTE UE receives the 200 OK, disconnects ring tone and sends a SIP ACK message
to acknowledge that the call has been established. The ACK message is sent to the
terminating leg of the call via the IBCFs.
bearer and via the IMS-AGW and pair of TrGWs. Support of Robust Header Compression is
mandated and described in GSMA PRD IR.92 [54] section 4.1.
.The IMS Signalling is sent over the default bearer.
5.2.4 Roaming VoLTE UE to VoLTE UE Call Establishment – Terminating Side

5.2.4.1 General
As section 3.2.4.1, with differences specific to the roaming scenario described in this
section.
V1.1 Page 96 of 121

Official Document
The essential differences from the single network scenario are that the Mw reference point
is conveyed via the IBCFs between the P-CSCF in the visited network and the S-CSCF in
the home network and that the Rx messages related to the establishment of the dedicated
bearer are sent via the visited and home network PCRFs over the S9 reference point. The
interactions between the visited network PCRF and the EPC/UE are identical to the single
network case. Therefore, the message flows in this section will show the SIP signalling and
Diameter Rx and Gx signalling only – the latter to indicate the points at which the EPC is
invoked. .
As for the single network case, a roaming VoLTE UE, shall perform call establishment by
using the IMS network. The IMS Signalling shall be sent over the default bearer, and a new
dedicated bearer shall be dynamically established for the voice traffic.
V1.1 Page 97 of 121

Official Document

PCRF P-CSCF/ PCRF
VoLTE UE PGW IMS-ALG / IBCF (V) IBCF (H) S-CSCF TAS
(V) (H)
IMS-AGW
SIP Invite (SDP)

SIP 100 Trying
SIP Invite (SDP)
SIP 100 Trying
SIP Invite (SDP)
SIP 100 Trying
SIP Invite (SDP) SIP Invite (SDP) SIP Invite (SDP)
SIP 100 Trying
SIP 100 Trying
AAR
AAR
RAR AAA
RAA AAA SIP 183 Progress (SDP) SIP 183 Progress (SDP) SIP 183 Progress (SDP) SIP 183 Progress (SDP)
SIP PRACK SIP PRACK SIP PRACK SIP PRACK SIP PRACK
SIP 200 OK (PRACK) SIP 200 OK (PRACK) SIP 200 OK (PRACK) SIP 200 OK (PRACK)
SIP 200 OK (PRACK)
SIP UPDATE (SDP)
SIP UPDATE (SDP) SIP UPDATE (SDP) SIP UPDATE (SDP) SIP UPDATE (SDP) SIP UPDATE (SDP)
SIP 200 OK (UPDATE) (SDP) SIP 200 OK (UPDATE) (SDP) SIP 200 OK (UPDATE) (SDP) SIP 200 OK (UPDATE) (SDP) SIP 200 OK (UPDATE) (SDP)
SIP 180 Ringing SIP 180 Ringing SIP 180 Ringing SIP 180 Ringing SIP 180 Ringing
SIP 180 Ringing
SIP 200 OK(INV) SIP 200 OK(INV) SIP 200 OK(INV) SIP 200 OK(NV) SIP 200 OK(INV)
AAR SIP 200 OK (INV)

AAR
RAR AAA
RAA AAA
SIP ACK SIP ACK

SIP ACK SIP ACK SIP ACK SIP ACK
VoLTE voice traffic via

dedicated bearer
Figure 26: Roaming VoLTE UE to VoLTE UE Call Establishment – Terminating Side message sequence
V1.1 Page 98 of 121

Official Document
answer (uplink & downlink configuration). It is also possible to invoke the PCRF twice, i.e.
on receipt of both the SDP Offer (downlink configuration) and SDP Answer (uplink
configuration). Both options are valid - – see 3GPP TS 29.213 ([79]) annex B.
is not shown.
As for section 3.2.4.3, with differences due to roaming highlighted in this section.
As in the single network scenario, the S-CSCF receives a SIP INVITE from the originating
leg of the call, invokes VoLTE services and routes the INVITE to the P-CSCF that was
associated to the subscriber during the IMS registration. In this case, the INVITE is
forwarded to the P-CSCF in the visited network via the pair of IBCFs. Each IBCF shall
invoke its respective TrGW to allocated media resources for the session and shall modify
the SDP in accordance with the newly created media pin-holes. The IBCFs shall also modify
the SIP headers as described in 3GPP TS 29.165.
As in the single network scenario, the P-CSCF (IMS-ALG) invokes the IMS-AGW (if
deployed) to reserve resources for the media connection. The SDP address in the INVITE is
over-written to reflect the media pin-hole created on the IMS-AGW.
As in the single network scenario, the P-CSCF forwards the SIP INVITE to the VoLTE UE.
The VoLTE UE shall allocate resources for the call and sends a 183 Progress response
containing an SDP answer with a single voice codec and indicating the preconditions are
desired but not yet met at the terminating end. This message also utilizes 100rel. The P-
CSCF updates the IMS-AGW (if deployed) with the SDP answer from the UE and sends the
Authorize/Authenticate-Request message to the PCRF in the visited network with the
related updated service information (IP address, port numbers, information on media-type)
for the dedicated bearer. The AAR message is conveyed to the PCRF in the home network
of the user via the S9 interface if implemented. In this case, the home network PCRF
authorises the request and responds with an Authorize/Authenticate-Answer (AAA)
message to the visited network PCRF and both PCRFs associate the service information to
the stored subscription related information containing the information about the allowed
service(s), QoS information and PCC Rules information. The visited network PCRF
identifies the affected IP-CAN session (e.g. default bearer) that has been established during
the LTE Attach procedure, and initiates a Re-Auth-Request to the PGW to initiate the
creation of a dedicated bearer as in the single network case. The visited network PCRF
shall also subscribe to modifications related to the dedicated bearer in the PGW (e.g.
LOSS_OF_BEARER, INDICATION_OF_RELEASE_OF_BEARER, etc.).
The PGW acknowledges the Re-Auth-Request to the visited network PCRF, which then
acknowledges the Authorize/Authenticate-Request message sent from the P-CSCF. At this
point the IMS SIP session and the dedicated bearer used for voice are bound together via
PCC.
VoLTE media as in the single network scenario.
V1.1 Page 99 of 121

Official Document
On receipt of the AAA response from the visited network PCRF, the P-CSCF will convey the
SIP 183 Progress (SDP) message to the S-CSCF in the user’s home network via the IBCFs.
The contained SDP reflects the address of the media pin hole in the IMS-AGW (if
deployed). In turn, each IBCF shall modify its TrGW to reflect the SDP answer and update
the SDP answer to reflect the respective media pin-holes in the TrGWs.
The PRACK message is transited from the originating side of the call via the IBCFs.
The terminating side sends a 200 OK (PRACK) in response to the PRACK which is
conveyed via the IBCFs.
The originating leg now sends a new SDP Offer in a SIP UPDATE message. The new offer
indicates that preconditions have been met at the originating side and that the media stream
is now active. The UPDATE message is conveyed to the terminating UE via the S-CSCF,
IBCF pair and the P-CSCF.
The terminating UE sends a SIP 200 OK (UPDATE) response containing the SDP answer.
The terminating UE notes that preconditions have been met at both ends, alerts the
subscriber and sends a SIP 180 Ringing message to the P-CSCF. This message is
transited to the originating leg of the call via the IBCF pair and S-CSCF. This message does
not utilize 100rel. The P-Early-Media header is not present in the message.
When the call is answered, the VoLTE UE shall send a SIP 200 OK (INVITE) message to
the P-CSCF. The P-CSCF sends an AAR message to the visited PCRF to enable the
uplink and downlink media flows. This message is conveyed to the home PCRF over S9 if
implemented. In this case, the home PCRF authorizes the request and responds with an
AAA message to the visited PCRF which invokes the P-GW to enable the media flows. The
visited network PCRF sends the AAA message to the P-CSCF. The P-CSCF(IMS-ALG)
also invokes the IMS-AGW (if deployed) to ensure that duplex media can be conveyed via
IMS-AGW at this point.
The SIP 200 OK (INVITE) message is forwarded to the S-CSCF via the IBCFs and then to
the originating side of the call. On receipt of the 200 OK (INVITE), each IBCF shall ensure
that duplex media can be conveyed via their respective TrGWs.
At this stage, the roaming VoLTE UE has a call established with voice traffic sent over the
dedicated bearer via the IMS-AGW and pair of TrGWs. Support of Robust Header
Compression is mandated and described in GSMA PRD IR.92 [54] section 4.1. The IMS
Signalling is sent over the default bearer.
5.2.5 Roaming VoLTE UE to VoLTE UE Call Clearing - Initiated
5.2.5.1 General
A roaming VoLTE UE, shall perform call clearing by using the IMS network. The IMS
Signalling shall be sent over the default bearer, and the dedicated bearer that was
dynamically established for the voice traffic shall be removed.
V1.1 Page 100 of 121

Official Document
VoLTE UE PCRF P-CSCF/ IMS- PCRF S-CSCF TAS

PGW IBCF (V) IBCF (H)
(V) AGW/ IMS-ALG (H)
SIP BYE SIP BYE SIP BYE SIP BYE
STR
SIP BYE
STR
SIP BYE
SIP BYE
RAR STA
SIP 200 OK
RAA STA
SIP 200 OK
SIP 200 OK
SIP 200 OK
SIP 200 OK
SIP 200 OK
SIP 200 OK
Figure 27: Roaming VoLTE UE to VoLTE UE Call Clearing – Initiated message sequence
V1.1 Page 101 of 121

Official Document
5.2.5.3 Detailed Description.

As section 3.2.5.3, with additional details specific to roaming added in this section.
The VoLTE UE sends a SIP BYE message to the P-CSCF. The P-CSCF (IMS-ALG)
releases the resources in the IMS-AGW and issues
A Session Termination Request to the PCRF in the visited network. The Session
Termination Request is sent to the PCRF in the home network of the user which responds
with a Session Termination Answer message over the S9 interface if implemented. The V-
PCRF and optionally the H-PCRF (if S9 is implemented) remove the binding between the
stored subscription information and the IMS service information, and the visited network
PCRF initiates a Re-Auth-Request to the PGW to remove the dedicated bearer as in the
single network case. .
The P-CSCF forwards the SIP BYE message to the S-CSCF in the home network of the
user via the IBCFs. The IBCFs will free off any media resources allocated in their respective
TRGws. The S-CSCF routes the SIP BYE to the S-CSCF of the other party. The other party
acknowledges the SIP BYE with a 200 OK.
The 200 OK (BYE) is signalled back to the UE via the IBCFs and P-CSCF.
has been removed.
5.2.6 Roaming VoLTE UE to VoLTE Call Clearing - Received
5.2.6.1 General
A roaming VoLTE UE, shall perform call clearing by using the IMS network. The IMS
Signalling shall be sent over the default bearer, and the dedicated bearer that was
dynamically established for the voice traffic shall be removed.
V1.1 Page 102 of 121

Official Document
VoLTE UE PCRF P-CSCF/ IMS- PCRF S-CSCF TAS

PGW IBCF (V) IBCF (H)
(V) AGW/ IMS-ALG (H)
SIP BYE
SIP BYE
SIP BYE
SIP BYE SIP BYE SIP BYE SIP BYE
STR
STR
RAR STA
RAA
STA
SIP 200 OK SIP 200 OK
SIP 200 OK
SIP 200 OK SIP 200 OK
SIP 200 OK
SIP 200 OK
Figure 28: Roaming VoLTE UE to VoLTE Call Clearing – Received message sequence
V1.1 Page 103 of 121

Official Document

As section 3.2.5.3 with additional details specific to roaming added in this section..
A SIP BYE is received by the S-CSCF from the other party. The S-CSCF shall forward the
SIP BYE to the TAS as in the single network case. The S-CSCF routes the SIP BYE to the
P-CSCF in the visited network via the IBCFs. The IBCFs shall free off any media resources
allocated to the call in their respective TrGWs. The P-CSCF forwards the BYE to the
VoLTE UE. The VoLTE UE acknowledges the call clearing by sending a 200 OK.
On receiving the SIP BYE, the P-CSCF (IMS-ALG) frees off the media resources in the
IMS-AGW. The P-CSCF also initiates a Session Termination Request to the V-PCRF to
initiate the process of removing the dedicated bearer that was established for the voice
traffic. The V-PCRF sends the Session Termination Request to its peer in the user’s home
network over the S9 interface if implemented. In this case, the PCRF in the user’s home
network responds with a Session Termination Answer to the visited network PCRF. The V-
PCRF and optionally the H-PCRF (if S9 implemented) remove the binding between the
stored subscription information and the IMS service information. The visited network PCRF
and initiates a Re-Auth-Request to the PGW to remove the dedicated bearer for voice
The 200 OK (BYE) is signalled back to the terminating leg of the call via the IBCFs and S-
CSCF.
has been removed.
5.3 Roaming Architecture for Voice over IMS with Local break-out (RAVEL)
As recommended in GSMA PRD IR.65 [51], it shall also be possible to utilize VPMN
Routing to replicate the CS charging model for VoLTE roaming scenarios as recommended
in GSMA PRD IR.65 ([51]) and described in 3GPP TS 23.228 ([5]), TS 24.229 ([9]) and TS
29.165([24]).
To apply this functionality, loopback routing procedures are required whereby the SIP
session signalling can be looped back from the home network to the visited network whilst
the media is anchored in the visited network. From that point, both the signalling and media
are routed to the terminating party’s home network as per current roaming CS routing
principles.
In addition to the roaming architecture in figure 19, the Transit & Roaming Function (TRF) is
required within the visited network to handle the looped back routing between the visited
and home networks of the originating party prior to forwarding the session request to the
home network of the terminating party. The procedures required for this are as follows:-
 The P-CSCF inserts the Feature-Caps header field with the “+g.3gpp.trf” header field
parameter set to the URI of the TRF.
 The IBCF in the visited network initiates OMR procedures (see 5.3), which also must
be supported in all other IBCFs that are traversed along the signalling loop. The
HPLMN decides, based on local policy whether loopback routing shall be applied. If
so, then the HPLMN routes the INVITE request back to the TRF in the visited
network including a Feature-Caps header field with a “+g.3gpp.loopback” feature-
capability indicator.
 The IBCFs traversed on during the loop-back will by means of the OMR procedures
determine that a media short-cut can be establish, and thus that media shall not be
anchored in the corresponding TrGWs.The TRF routes the SIP INVITE to the home
network of the destination party via the IBCFs and ensures that OMR procedures
are not further activated.
V1.1 Page 104 of 121

Official Document
 When 200 OK (INVITE) is sent along the signalling loop from the TRF in the VPLMN
via the S-CSCF in the HPLMN and back to the P-CSCF in the VPLMN, the traversed
IBCFs will be informed that a media short cut is established and that no media will
pass through the respective TrGWs. IBCFs that previously have not become aware
of this (in particular IBCFs between P-CSCF and S-CSCF) will release the allocated
media resources in the corresponding TrGWs.
 If loop-back routings shall not be applied, The HPLMN terminates the OMR
procedures to ensure that media anchoring in the HPLMN will occur..
In VPMN Routing, Optimal Media Routing (OMR) us used to determine the optimal media
path between the visited network of the originating party and terminating network without
passing through the home network of the originating party.
5.4 Optimal Media Routing
When VPMN routing as recommended in GSMA PRD IR.65 ([51]) is supported, The OMR
procedures described in 3GPP TS 29.079 ([77]) must also be supported. The Reason for
applying OMR for VPMN routing is, to maintain the media path in the visited network and
avoid needlessly tromboning the media path via the home network. Note that the signalling
shall always be conveyed to the home network,
The respective IBCFs through which the IMS signalling traverses would support the SDP
visited-realm media attribute extensions defined in 3GPP TS 24.229 ([9]) resulting in OMR
being recognized as being applicable and resulting in media being negotiated to flow
between the respective IMS-AGWs.
The message flows documented in sections 5.2.3 are unchanged by OMR. However, if
OMR is applicable, then the IBCF frees off the media resources in the TrGW during session
establishment which means that there is subsequently no need for the IBCF to
communicate with the TrGW during session teardown.
See annex C of 3GPP TS 29.079 ([77]) for example message flows illustrating OMR.
5.5 Diameter Signalling

Roaming introduces the added complexity of there being an additional capability required to
provide a PMN-edge Diameter function to reduce the export of network topologies and
support scalability and resilience. Such a Diameter Edge Agent is described in GSMA PRD
IR.88 [53] section 3.1.3.
The Diameter Edge Agent needs to discover which peer to route messages to for a specific
application to enable Diameter routing between networks. This may be done manual
configuration of Diameter nodes within each node. However a dynamic discovery
mechanism (NAPTR query), would allow for a simpler, scalable, and robust deployment.
GSMA PRD IR.88 [53] section 3.1.3.4 describes the peer discovery mechanism in further
detail.
The Diameter Edge Agents use application id=0 in their Capability Exchange to indicate that
all applications are supported. This approach enables new applications to be added in each
network without impacting on the Diameter Edge Agent.
The roaming scenario requires Diameter messages needing to be routed between the
respective Diameter Edge Agents. As in the single network scenario, all Diameter enabled
elements in a network shall route their requests and responses via the local Diameter
Agent. For the roaming scenario, the Diameter Agent routes the requests/responses via the
local Diameter Edge Agent based on the realm identity in the message.
V1.1 Page 105 of 121

Official Document
A Diameter Edge Agent may apply topology hiding to reduce the export of topology
information across a network boundary. The Diameter Edge Agent may optionally overwrite
the host/realm information but this requires a mapping table to be maintained within the
Diameter Edge Agent.
5.6 Traffic Management and Policy
Dynamic or static policy control may be applied between the home and visited networks. For
dynamic policy control, Rx messages are sent between the visited network PCRF and home
network PCRF via the S9 interface. For static policy control, rules are configured locally in
the V-PCRF.
In terms of DiffServ marking, TrGWs are present in the end-end media path and are thus
able to modify the DSCPs under the control of the IBCF. It is expected that DSCP settings
by the IBCF are consistent with table 4.
5.7 Session Border Controllers
For the roaming scenario, both signalling and media traverse the network boundary
between the visited and home networks via network side SBCs performing the role of an
IMS IBCF/TrGW and providing capabilities such as topology hiding, firewall, NAT traversal
etc.
It is also noted that is possible for an access and network side SBC to be co-located in a
single physical element. In this case, the message flows documented in section 5 may be
simplified as a single element would provide the IMS P-CSCF/IMS-ALG/IMS-AGW and
IBCF/TrGW functions such that messages between the P-CSCF and IBCF (in the control
plane) and IMS-AGW and TrGW (in the media plane) would be internal to that element.
5.8 IMS Emergency Call
For a roaming UE, there is a requirement for the emergency call to be connected to a PSAP
in the visited network. The P-CSCF in the visited network shall recognise the emergency
service URN (see RFC 5031 [78]) and route the request to an E-CSCF in the visited
network. The requirements of IMS emergency calls are discussed in 3GPP TS 23.167
([79]) and the procedures of the E-CSCF are described in section 5.11 of 3GPP TS 24.229
([9]).
Note that if IMS Emergency Call is not supported, then CS Emergency Call is required.
5.9 Lawful Intercept
For the roaming scenario, the Access Session Border Controller (P-CSCF/IMS-ALG/IMS—
AGW) in the visited network is a mandatory point of intercept (see 3GPP TS 33.107 [41]
section 7A). .
5.10 Security
For the roaming scenario, additional security is required between the Diameter Edge
Agents. See section 6.5.1 of GSMA PRD IR.88 ([53]). .
5.11 Charging
For VoLTE calls in the roaming scenario with local break-out with the P-CSCF in the visited
network, then the VPMN is service aware and there is scope for service based charging to
be deployed between the visited and home networks and to apply flow based charging
mechanisms. However, this is topic is still under study and it is likely that charging rates for
voice will (for the moment) continue to be independent of the underlying technology.
V1.1 Page 106 of 121

Official Document
6 Implementation Guidelines
This section provides the highlights of the issues discovered during Interoperability testing
(IOT) and Operators commercial deployments. It contains the guidelines for the VoLTE
related protocol implementations in order to achieve seamless interoperability of VoLTE
products and accelerate their time-to-market (TTM).
The section is separated in 5 main sections:-
 Open Implementation Issues
 VoLTE Device Implementation Guidelines
 LTE/EPC Implementation Guidelines
 VoLTE IMS Implementation Guidelines
 Other Implementation Guidelines
6.1.1 Open Implementation Issues
Ref. ID Title Priority Category Status

Diameter signalling Critical Other On-going work in 3GPP
ID_Other_ overload and congestion based on IETF work.
01 Recommendation to deploy a
Diameter Agent.
6.1.2 VoLTE Device Implementation Guidelines

6.1.2.1 Downloadable VoLTE Client non-compliancy – Requirement for VoLTE UE
support of an Embedded IMS stack
Downloadable Client non-compliancy – Requirement for UE support of

Title Embedded IMS stack
Reference ID ID_Device_01
Priority Critical
Date Submitted 08/07/2013

Date Modified 08/07/2013
a) Some downloadable VoLTE clients are not providing

information required for VoLTE authentication and services
Overview (e.g. IMSI in the IMS private ID, IMEI, etc.). Utilisation of
downloadable clients not able to provide the correct QoS.
Status Closed
Some downloadable VoLTE clients are non-compliant to the
capabilities described within GSMA PRD IR.92 [54]. Therefore the
Detailed Description
required registration, security mechanisms, and applied QoS are
failing, without having to provide proprietary solutions in the network.
Downloadable VoLTE clients must be compliant with GSMA PRD IR.92
Solution [54] and the capabilities described within.
The recommendation is for the UE to use native VoLTE functionality
V1.1 Page 107 of 121

Official Document
(with an embedded IMS stack) pre-installed on device.
6.1.2.2 VoLTE UE exceeds the link MTU-size – IP Layer fragmentation – Packets

dropped
VoLTE UE exceeds the link MTU-size – IP Layer fragmentation –

Title Packets dropped
Priority High
IP fragmentation may occur between the UE and the PGW if the UE
sends packets that exceed the maximum link MTU size that is
supported in the network as part of IP configuration. IP fragmentation
Overview
is not recommended by 3GPP due to significant transmission
overhead. If the UE exceeds the limit and IP fragmentation is not
supported in the EPC, the result is packet loss.
Status Closed
3GPP TS 23.060 [3] Annex C provides information related to Link MTU
considerations. The maximum size of the link MTU size is currently set
to 1500 octets. Taking into account the headers for GTP packets that
may be further encapsulated within an IPSec tunnel, the overall UE link
MTU size is set at 1358 octets.
The link MTU size of the network can be requested by the UE in the
Protocol Configuration Options (PCO) during LTE Attach. This enables
the UE to discover the link MTU size and be compatible with the
network IP configuration.
It has been discovered that not all UE's request the link MTU size, and
regularly exceed this limit when sending SIP messages (e.g.
particularly in downloadable clients). If the network does not support
procedures for IP fragmentation, then the packets are discarded which
in turn results in loss of VoLTE functionality.
VoLTE UE's shall request the link MTU size from the network
(requested in the PCO during attach) and utilise this value when
Solution transmitting data packets.
NOTE: For IPv6 implementations, the link MTU size is present in the
IPv6 Router Advertisement.
6.1.2.3 VoLTE UE's not supporting GBA for Ut Authentication
Title VoLTE UE's not supporting GBA for Ut Authentication

Priority High
V1.1 Page 108 of 121

Official Document
Support of GBA on VoLTE devices requires the support of alternative

Overview mechanisms for authentication of the Ut interfaces used for
Supplementary Service configuration.
Status Closed
GSMA PRD IR.92 [54] section 2.2.2 recommends that the UE supports
the Generic Authentication Architecture procedures for authentication,
alternative procedures as defined in 3GPP TS 24.623 [21] may also be
used.
GSMA recommends the support of GAA for Ut interface authentication.
As an interim solution for systems where Generic Authentication
Solution Architecture is not supported, the VoLTE UE and TAS shall support
HTTP authentication and TLS as defined in RFC 2617 [57] and IETF
RFC 2246 [56] according to ETSI TS 183 038 [47].
6.1.3 LTE/EPC Implementation Guidelines

6.1.3.1 Incorrect LTE Cause Code causing UMTS/GSM roaming failure
Title Incorrect LTE Cause Code causing UMTS/GSM roaming failure

Reference ID ID_LTE/EPC_01
Priority Critical
Whilst a Home Network Operator and a Visited Network Operator may
have an established UMTS/GSM roaming agreement, LTE roaming
may not yet be available. Without LTE roaming enabled between the
Home Network Operator and the Visited Network Operator,
UMTS/GSM inbound roamers with LTE-capable UMTS/GSM handsets
may be restricted from roaming on to the UMTS/GSM network unless
and until manual re-selection of the UMTS/GSM network is made.
Without LTE Roaming enabled the access control in the Visited
Network rejects the LTE Attach request with an applicable cause code
to allow the device to select another Radio Access Technology for the
Overview
same PNM (e.g. UMTS/GSM). The issue specified above is caused by
a non-recommended LTE attach reject cause code being returned by
some LTE networks, to roaming subscribers. The result is that
handsets will not attempt to roam onto the UMTS/GSM network and
therefore access is restricted to the roaming subscriber, even though
an existing roaming relationship for UMTS/GSM exists.
GSMA PRD IR.88 [53] (LTE Roaming Guidelines) recommends LTE
networks should return #15 (no suitable cells in tracking area),
however a number of LTE networks are returning #11 (PLMN not
allowed) or #14 (EPS services not allowed in this PLMN).
Status Closed
1. A handset tries to attach to LTE access as the first Radio Access
Detailed Description Technology in the visited network (normal behaviour of LTE-capable
handsets).
V1.1 Page 109 of 121

Official Document
2. As the LTE network has not implemented LTE-LTE roaming with

the HPMN of the roaming subscriber, the LTE MME returns an attach
reject with a cause value, e.g. #11 (or #14).
3. As the handset received cause code #11 (or #14), the handset
stores the LTE MCC-MNC into its forbidden PMN list (or its forbidden
PMN for GPRS service list).
4. The handset will never initiate a Location Update or a GPRS Attach
to the VLR/SGSN of the UMTS/GSM Network which is broadcasting
the same MCC-MNC, because the MCC-MNC is already stored in
forbidden PMN list (or forbidden PMN for GPRS service list).
5. As the result of above steps, unless and until the roaming
subscriber manually re-selects the UMTS/GSM Network, the handset
remains out-of-service in case of no other UMTS/GSM Network
available (or packet data does not work while voice/SMS may work).
If the VPMN (broadcasting the same MCC-MNC as its UMTS/GSM
network) already has an existing roaming agreement for other Radio
Access Technologies with the HPMN, the LTE Attach reject cause
Solution code should be changed to #15 at its earliest possible timeframe.
NOTE: This behaviour is described in GSMA PRD IR.88 [53] section
6.1.1.
6.1.3.2 User Location Information, TimeZone, Cause Code is not reported when
VoLTE call is dropped.
Updated User Location Information, TimeZone is not reported to EPC

in user bearer release and UE context release scenarios, and
RAN/NAS Cause Code is not reported to IMS Application Server when
Title VoLTE call is dropped
Reference ID ID_LTE/EPC_02
Priority High
When a VoLTE dedicated bearer or session is dropped, an operator
will not get current ULI/TimeZone nor the real failure cause in the S-
Overview CDRs and P-CDRs, nor at IMS level, and then will not be able to make
performance analysis, User QoE analysis and proper billing
reconciliation.
Status Closed
When an ERAB or a data session is dropped, an operator may need
to get the most up-to-date ULI information as well as detailed RAN
and/or NAS cause codes information from the access network to be
included in the S-GW and PDN GW CDRs for call performance
analysis, User QoE analysis and proper billing reconciliation. Also, for
IMS sessions, the operator may need to get the above information
available at P-CSCF.
ULI and UE Time Zone:

a) An operator may be only interested in where the ERAB or the
data session is established and where it is released or dropped.
Intermediate ECGI information may not be too much of
V1.1 Page 110 of 121

Official Document
interests. Using Location Reporting procedure would not be

easy because it only provides an on-demand mechanism (or
reporting at any cell change). Moreover, the eNB does not
provide the ECGI in current ERAB Release Response, ERAB
Release Indication, UE Context Release Request and UE
Context Release Complete messages;
b) The Delete Bearer Command and Release Access Bearer
Request messages currently don’t contain ULI and UE Time
Zone, which makes impossible for the SGW and PGW to get
this information for their CDRs without requiring significant
PCRF and PGW state machine changes (PGW and PCRF have
to keep their connection alive, and PCRF has to wait for PGW
before replying to the P-CSCF). See below implementations
details;
c) For VoLTE UEs, if a call is dropped due to RF conditions, the
VoLTE UE has no way to give the last seen ECGI to the P-
CSCF over the Gm interface, therefore this information should
be added to Gx IPCAN Session messages where missing.
Further implementation details for ULI and Time Zone:

- At specific UE or EPC events such as UE-initiated Detach,
MME/SGSN-initiated Detach, HSS-initiated Detach
procedure, UE or MME requested PDN disconnection, ULI
and TimeZone are systematically provided to PCRF in the
IPCAN Session Termination.
- However, at other specific UE or EPC events such as MME
Initiated Dedicated Bearer Deactivation, whereas ULI and
Time Zone are available in the MME when the procedure
starts, they are not provided immediately to the PGW and the
PCRF. They are delayed to the next PCEF-PCRF exchange,
and provided only if the PCRF has required it. This was
specified for the introduction of NPLI feature and significantly
impacts PGW and PCRF state machine.
- There is a need for simple mechanism, not requiring full NPLI
implementation, by which ULI and UE Time Zone are
provided to PCRF in an existing IPCAN Session Modification
upon a loss of bearer.
Change Requests to 3GPP specifications to add the ULI, TimeZone
and Cause Codes to the relevant messages have either been done or
Solution
are in the process of being done (expected completion date of June
2014) for 3GPP R12 onwards.
6.1.4 VoLTE IMS Implementation Guidelines

6.1.4.1 Compatibility of MGCF with VoLTE services utilising MMTel ICSI
Title Compatibility of MGCF with VoLTE services utilising MMTel ICSI

Reference ID ID_IMS_01
Priority High

3GPP TS 29.163 [23] does not mandate that the MGCF include the
Overview MMTEL ICSI in the INVITE message. This causes an issue as the
target S-CSCF looks for this information in the iFCs when invoking the
V1.1 Page 111 of 121

Official Document
MMTEL Application Server. When calls arrive from the CS network at

the MGCF, it can be assumed that for voice calls that MMTEL services
are applicable and thus the ICSI should be included and set to identify
VoLTE.
Status Closed
For voice calls ingressing to IMS via the MGCF, it can be assumed that
VoLTE services based on MMTel are applicable and thus the ICSI
needs to be included in order to enable the terminating S-CSCF to
invoke the appropriate Application Server into the session, The current
tables in 3GPP TS 29.163 [23] focusses on deriving/mapping between
legacy information in SIP-I/BICC/ISUP to IMS SIP and does not include
MMTel aspects. It is believed that the absence of the ICSI is an
Detailed Description oversight due to it being a pure IMS concept. The ICSI can be set up
for all ingress sessions to the MGCF with “m=audio” in the related
SDP. The MGCF should set the media feature tag for IP Voice in
Contact header (set to +g.3gpp.icsi-ref="urn%3Aurn-7%3A3gpp-
service.ims.icsi.mmtel"). In addition, the MGCF is acting as a proxy for
the originating (CS) user and thus should mimic the behaviour of an
originating UE as described in clause 5.1.2A.1.1 of 3GPP TS 24.229
[9].
Change request to 3GPP CT3 to change 3GPP TS 29.163 [23] agreed
and this behaviour is mandated as defined in section 7.2.3.2.2.5 of
3GPP TS 29.163 [23].
Solution
GSMA recommends that deployed MGCF's insert the ICSI within
relevant messages.
6.1.4.2 Correct support of dynamic codecs
Title Correct support of dynamic codecs

Priority High

Some endpoints do not properly support dynamic payloads for codecs.
Specifically, they mandate/.assume that AMR be payload type 96 at all
Overview
times rather than allowing dynamic codecs to be correctly assigned a
number from the dynamic number range.
Status Closed
Codecs that use dynamic payload numbers should be able to be
Detailed Description assigned to any of the permitted numbers from the dynamic range as
described in RFC 3550 [58] and (from AMR) RFC 4867 [61]. .
Endpoints should implement dynamic payload mapping properly and
Solution make no assumptions about fixing integer values to represent dynamic
payload types.
V1.1 Page 112 of 121

Official Document
6.1.4.3 Indication of Early media support
Title Indication of Early media support

Priority High

There is not a single consistent way to indicate support of early media.
Sometimes the P-Early-Media header is used, other implementations
Overview
use analysis of the SDP. This inconsistency can cause issues when
required to set up early dialogs and media.
Status Closed
The P-Early-Media header (defined in RFC 5009 [63]) was defined as
an explicit mechanism in SIP to indicate the support/presence of early
media. However, some implementations still do not support this header
and also specifications have (for historical reasons) been vague and
allowed its support to be optional rather than mandatory. This has led
to a situation where the support of the P-Early-Media is sporadic which
can lead to inter-operability issues when support of early media is
required.
All VoLTE endpoints and elements should support the P-Early-Media
Solution
header.
6.1.4.4 Assorted MGCF SIP Issues
Title Assorted MGCF SIP Issues

Priority High
Overview Assorted issues highlighted regarded non-compliance of MGCFs.

Status Closed
The highlighted issues include the following:-
 Inconsistent use & support of the SIP ROUTE header,
 Support of only TEL URIs and no support of SIP URIs with
user=phone,
These issues are likely to have arisen due to legacy implementations
being evolved to become a MGCF.
Solution MGCFs should correctly support the highlighted issues.
6.1.4.5 Support of legacy CS Services
Title Support of legacy CS Services

V1.1 Page 113 of 121

Official Document
Priority High

CS networks support a large number of legacy services. With the
Overview advent of IMS and service provision via a Telephony AS, concern has
been expressed regarding service equivalence.
Status Closed
In order to achieve the cited capability, there is a need to replicate the

(more) intelligent routing capability existing in MSC Servers. In the IMS
context, this capability is inherently more difficult to realize due to the
more fragmented nature of the overall architecture. The VoLTE
Architecture and baseline 3GPP specifications do not provide a
standardised solution to interwork with legacy proprietary services that
are commonly utilised within Operators CS networks. A number of
possibilities exist for legacy CS proprietary services within the VoLTE
Detailed Description Architecture:-
1. Cease support of the legacy proprietary services for VoLTE
users:
2. Develop the legacy proprietary services within the TAS
3. Develop interworking between the VoLTE architecture and the
existing legacy proprietary CS network services
Refer to clause 3.13 for further details of the above.
It is expected that the solution will vary between operators and based
Solution
on the nature of the services and the optimum option selected.
6.1.4.6 Support wildcard PSIs in HSS
Title Support wildcard PSIs in HSS

Priority High

HSSs have been observed that do not support wildcarded PSIs. Such
Overview
support is needed for dial-in conferencing.
Status Closed
For dial-in conferences, the conference URIs must be realized as

Detailed Description wildcarded PSIs within the IMS network for proper SIP request routing
to be successful.
Solution HSSs should provide support for wildcard PSI's.
6.1.4.7 SIP Overload Control
Title SIP Overload Control

V1.1 Page 114 of 121

Official Document
Priority High
Whilst most equipment provides for interpretation and remapping of

SIP Causes, there is still an issue around network congestion controls
Overview tied to specific numbers (tele-voting) and the need for some
standardized way in SIP to put backward controls on specific
destinations.
Status Closed
Legacy control protocols (e.g. BICC, ISUP) employ mechanisms to
enable an overload condition to be signalled as part of a session
rejection to enable the rate of future session requests to be reduced
(e.g. ISUP ACC). In addition, legacy nodes are also able to apply
Detailed Description network level services such as Call Gapping to prevent a focussed
overload to a given destination number (e.g. tele-voting). In the IMS
world, similar mechanisms are required in order to protect the network
from load spikes.
Legacy services such as Call Gapping should be realized in IMS via a

network level service on the AS.
At the current time, there is no agreed mechanism is SIP to convey an
overload condition although work is currently on-going in the IETF in
the SOC WG and an informational IETF RFC 6357 [64] has been
published discussing SIP overload design considerations. There are
Solution also a number of current internet drafts in progress (draft-ietf-soc-load-
control-event-package-13, draft-ietf-soc-overload-control-14 & draft-
ietf-soc-overload-rate-control-07). These drafts have been agreed as
SIP overload control mechanisms by 3GPP and referred to in 3GPP
TS 24.229 section 4.12 in Rel11. The work in IETF should be
monitored and reviewed when complete and published as RFCs.
These mechanisms for SIP overload control should be implemented.
6.1.4.8 XML Complexity for MMTel Services
Title XML Complexity for MMTel Services

Priority High
Concern was expressed at the complexity of the XML that is used for
Overview the configuration of the MMTel services over the XCAP Ut reference
point.
Status Closed
As stated in section 3.4.1, a flexible and extensible XML document
structure has been defined to support the management of XML
services. This includes the (generic) capability to access a given XML
document in its entirety as well as accessing a sub-set of the document
V1.1 Page 115 of 121

Official Document
(e.g. interrogating the destination address for the CFU supplementary

service).
All UEs and Application Servers shall support the XML documents
Solution
defined for the mandatory set of VoLTE services (see section 3.4.1).
6.1.4.9 SIP Fragmentation
Title SIP Fragmentation

Priority High
Related to section 6.1.2.2, the MTU for SIP messages can also be
Overview exceeded in the IMS core. On occasion, these messages have
become fragmented over UDP transport.
Status Closed
Even if the link MTU is not exceeded by the UE (see section 6.1.2.2),
SIP messages become extended as they traverse the chain of IMS
elements (e.g. VIA and RECORD-ROUTE headers growing in length)
and thus SIP fragmentation can occur if the MTU is exceeded and
Detailed Description UDP transport is used. The SIP, as defined in IETF RFC 3261 [70],
does discuss this issue in clause 18.1.1. and recommends that UDP is
not used if the message is within 200 bytes of the MTU or if the
message exceeds 1300 bytes and the MTU is unknown.
Implementations must avoid SIP fragmentation by obeying clause

18.1.1 of RFC 3261 and swapping over to TCP for the transport of
Solution
large messages. Note that the transport change is done on a per SIP
message basis and not on a per session basis.
6.1.5 Other Guidelines

6.1.5.1 Diameter signalling overload and congestion
Title Diameter signalling overload and congestion

Reference ID ID_Other_01
Priority Critical
Diameter requests are being dropped at the server due to queue
Overview
overload resulting in lack/degradation of VoLTE service.
Status Open
The existing overload control mechanisms in the Diameter base
protocol are too limited to efficiently prevent and react to signalling
V1.1 Page 116 of 121

Official Document
overload. These limitations are even more critical in large scale

networks in which multiple Diameter nodes, from various vendors, are
in the signalling path.
Solutions for Diameter overload prevention, detection, and overload
mitigation are required.
The IETF DiME working group has defined "Diameter Overload Control
Requirements" (draft-ietf-dime-overload-reqs-01) with proposed
solutions for "Diameter Overload Control Application" (draft-korhonen-
dime-ovl-00.txt) and "A Mechanism for Diameter Overload Control"
(draft-roach-dime-overload-ctrl-01).
3GPP have completed a Release 12 Study Item on Diameter Overload
Control Mechanisms to leverage the work within IETF and provide
Solution solutions for Overload Indication, Load-balancing, message
retransmission, message throttling, message prioritisation and
application prioritisation. See 3GPP TR 29.809 [36].
3GPP are now defining normative solutions, based on the IETF work,
which should be monitored and enacted within VoLTE deployments
when normative work has been agreed. In the interim, proprietary
mechanisms may also exist, e.g. within the implementation and
deployment of a Diameter Agent (see section 3.6.2).
6.1.5.2 Diameter Charging Interfaces (Rf and Ro) not Supported
Title Diameter Charging Interfaces (Rf and Ro) not Supported

Priority Medium

The network readiness for support of Diameter Rf and Ro interfaces for
Overview Online and Offline Charging is not available. This has led to interim
solutions.
Status Closed
As Diameter Rf and Ro interfaces are not supported in some networks,
interim solutions have been deployed for charging. These include
direct CDR generation (often using the MSC’s CDR format to prevent
billing domain changes) on each IMS node, other interfaces for online
Detailed Description charging (e.g. CAMEL, Radius, etc.).
There is an expectation that the IMS network shall be able to provide
all information that is currently provided in a MSC-generated CDR,
resulting in proprietary solutions to enable cause codes to be captured
(in CDRs, etc.) in a manner consistent with CS domain networks.
Interim solutions may apply, with the view for deployment of Rf and Ro
Solution
interfaces for Online and Offline charging.
V1.1 Page 117 of 121

Official Document
6.1.5.3 Lack of support for ENUM
Title Lack of support for ENUM

Priority Medium

Lack of ENUM support causes issues for support of Mobile Number
Overview
Portability and resolving E.164 numbers to a routable SIP URI.
Status Closed
Support of ENUM is required for resolving E.164 numbers to routable
SIP URI's within the VoLTE architecture. On a VoLTE call between
two Operator's networks, lack of support for ENUM in the originating
network may result in the terminating network performing sub-optimal
Detailed Description routing, i.e. trying to route a call to a non-IMS subscriber before having
to perform a breakout to the terminating CS network.
Lack of support for the usage of ENUM is leading to proprietary
solutions for performing Mobile Number Portability.
The recommendation is that ENUM is supported within the VoLTE
Solution
architecture.
6.1.5.4 IMS Lawful Intercept interfaces not defined
Title IMS Lawful Intercept interfaces not defined

Priority Medium
The IMS Lawful Intercept X1, X2 and X3 specifications are not defined.
This leads to prolonged deployment of solutions as IMS vendors and LI
Overview
vendors must agree on the interfaces to be used within each network
deployment.
Status Closed
The IMS Lawful Intercept X1, X2 and X3 specifications are not defined.
This leads to prolonged deployment of solutions as IMS vendors and LI
vendors must agree on the interfaces to be used within each network
deployment.
3GPP may define the X1, X2, X3 interfaces, although standardisation
may be a lengthy process.
Solution In the interim, IMS vendors and LI vendors must perform integration on
a per network basis to perform interception at the A-SBC/P-CSCF as
V1.1 Page 118 of 121

Official Document
6.1.5.5 Support for (e)SRVCC requires 3GPP Release 10
Title Support for (e)SRVCC requires 3GPP Release 10

Priority Medium

Support of 3GPP Release 10 functionality is not readily available which
Overview
is inhibiting the deployment of (e)SRVCC.
Status Closed
Whilst VoLTE describes the support of SRVCC as described in GSMA
PRD IR.92 [54] section A.3 which is based on 3GPP Release 8 (with
Release 9 aspects to support Emergency Call), there is a view that this
solution is not robust enough for handover between LTE coverage and
UMTS/GSM coverage. The enhanced SRVCC functionality defined
Detailed Description within 3GPP Release 10 also supports the mid-call feature during
SRVCC handover, support for calls in alerting state and enhancements
to minimise voice interruption delay.
Some HSS deployments do not support 3GPP Release 10, and the
resultant Sh interface usage to allow the update of the STN-SR in the
HSS.
Support of 3GPP Release 10 eSRVCC is recommended to
Solution complement VoLTE with CS voice, or to utilise Inter-RAT PS Handover
as described in section 3.14.
6.1.5.6 Other Diameter Issues
Title Other Diameter Issues

Priority High

A number of Diameter related issues were highlighted:-
 Inconsistent use of Diameter AVPs in requests (primarily on
whether to include the Destination-Host AVP) and related to
the addition of a DRA to a deployed network,
Overview  Some Diameter Servers do not allow multiple TCP / SCTP
connections from a single Origin-Host, which leads to having to
use multiple Origin-Host identities on a single network
element.
 Use of proprietary Diameter AVPs
Status Closed
The deployment of a Diameter Agent (see section 3.6.2) is
recommended. However, with a DRA present, there are options as to
whether to include the Destination-Host AVP and perform realm based
routing.
V1.1 Page 119 of 121

Official Document
It is undesirable and more complicated to have to use multiple Origin-

Host identities (e.g. HSS that supports multiple Diameter applications)
on a single network element. As a related point, it is also undesirable
for such an element to require multiple TCP/SCTP connections rather
than use a single connection and advertise multiple Diameter
interfaces via the Capability Exchange mechanism.
It is recommended to deploy a DRA as per section 3.6.2.
Network elements need only include a Destination-Host AVP when it is
available (e.g. pre-configured, obtained via a previous request/answer
exchange). The DRA shall support the routing of Diameter messages
when the Destination-Host AVP is not present.
As indicated in IETF RFC 3588 [59], a given Diameter instance of the
peer state machine must not use more than one transport connection
to communicate with a given peer, unless multiple instances exist on
the peer, in which, case a separate connection per process is
allowed. This means that a given Origin-Host may only be used on a
Solution
single connection. Also, it is noted that the use of a DRA provides a
point of inter-working to enable inter-operability between Diameter
elements both at a transport and application level.
Use of proprietary AVPs should be deprecated where possible (i.e.
standard AVP in existence to perform a given function). VoLTE service
does not rely on proprietary AVPs as standardised AVPs have been
defined for VoLTE functionality. Where proprietary AVPs are included
in Diameter messages, these should be defined as optional and
discarded by elements where they are not supported without rejecting
the message.
V1.1 Page 120 of 121

Official Document
Document Management
Document History
Version Date Brief Description of Change Approval Editor /

Authority Company
This PRD was reviewed and Dave Hutton &
1.0 05/12/2013 approved by FCPLT in December FCPLT Wayne Cutler
2013 GSMA
CR1001 VoLTE Service
Description and Implementation Dave Hutton &
1.1 26/03/2014 Guidelines and CR1002 VoLTE FCPLT Wayne Cutler
Service Description and GSMA
ImplementationGuidelines
Other Information
Type Description
Document Owner GSMA Future Communications Programme
Editor / Company David Hutton / GSMA, Wayne Cutler / GSMA
It is our intention to provide a quality product for your use. If you find any errors or
omissions, please contact us with your comments. You may notify us at prd@gsma.com
Your comments or suggestions & questions are always welcome.
V1.1 Page 121 of 121

Volte Service Description and Implementation Guidelines 26 March 2014

Uploaded by

Copyright:

Available Formats

Volte Service Description and Implementation Guidelines 26 March 2014

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Volte Service Description and Implementation Guidelines 26 March 2014

Uploaded by

Copyright:

Available Formats

GSM Association Non-confidential

Official Document FCM.01 - VoLTE Service Description and Implementation Guidelines

VoLTE Service Description and Implementation Guidelines

This is a Non-binding Permanent Reference Document of the GSMA

Security Classification: Non-confidential

V1.1 Page 1 of 121

2.2.29 Izi Interface (TrGW – TrGW) 23

3.13 Support of Legacy Proprietary CS Services 65

5.11 Charging 106

1.2 Relationship to Standards

 Section 2 VoLTE Architecture:- defines the logical architecture, functional node

I-SBC Interconnect Session Border Controller

MWI Message Waiting Indicator

SIP Session Initiation Protocol

1.5 Document Cross-References

Extensible Markup Language (XML) Configuration Access

Overall description; Stage 2

[72] 3GPP TS 29.238 Interconnection Border Control Function (IBCF) – Transition

Figure 1: VoLTE Logical Architecture

2.1.1 VoLTE UE (User Equipment)

2.1.3.3 PGW (Packet Data Network Gateway)

transcoding, multiparty conferencing, network announcements/tones, etc. under the control

2.1.4.7 IMS-ALG/IMS-AGW (IMS Application Level Gateway/IMS Access Gateway)

2.1.4.8 MGCF/IMS-MGW (Media Gateway Control Function / IMS Media Gateway)

2.1.4.9 BGCF (Breakout Gateway Control Function)

2.1.5 Additional Network Functionality

2.2.18 Mx Interface (x-CSCF – IBCF)

2.2.28 Ici Interface (IBCF – IBCF)

PRD Title Description

This document defines a voice over IMS profile by listing a

Table 1: VoLTE GSMA Permanent Reference Documents

3 VoLTE Implementation - Single PMN

This section describes the implementation for these scenarios.

Figure 2: Intra-PMN VoLTE deployment

3.2.1 VoLTE UE Attachment and IMS Registration

RRC Connection Request

Update Location Request

Update Location Answer

Create Session Request

IMS-APN; QCI:5; Create Session Request

Modify Bearer Response

Default Bearer Established for IMS Signalling IMS Signalling over IP

SIP Register SIP Register

Figure 3: VoLTE UE Attachment and IMS Registration message sequence

V1.0 Page 27 of 121

3.2.2 VoLTE UE Initiated Detach and IMS Deregistration

V1.0 Page 32 of 121

3.2.2.2 Message Sequence

Default Bearer for IMS Signalling IMS Signalling over IP

Notify (reg event) Notify (reg event) Notify (reg event)

UE is not IMS Registered

Delete Session Response

Delete Session Request

Detach Accept Delete Session Response

Release Access Bearers Request CCR

Release Access Bearers Response

RRC Connection Release

UE Context Release Complete

Default Bearer is Deleted and UE is not LTE Attached