Description
Voice over IP (VoIP) within the 3GPP framework is the standardized methodology for transmitting voice communications as data packets over IP-based networks, specifically the Evolved Packet Core (EPC) and 5G Core (5GC). It represents a fundamental shift from traditional circuit-switched telephony (like 2G/3G CS voice) to an all-IP, packet-switched architecture. The technology leverages the IP Multimedia Subsystem (IMS) as the core service delivery platform. IMS provides the necessary control functions for session establishment, management, and teardown using protocols like SIP (Session Initiation Protocol). The actual voice data is encoded using codecs such as AMR (Adaptive Multi-Rate) or EVS (Enhanced Voice Services), packetized into Real-time Transport Protocol (RTP) streams, and transported over the underlying IP bearer network.
Architecturally, VoIP service delivery involves several key network functions. The User Equipment (UE) must support IMS client functionality. The call signaling traverses the P-CSCF (Proxy-Call Session Control Function), S-CSCF (Serving-CSCF), and other IMS nodes for routing and service logic. Media resources, if needed for conferencing or transcoding, are handled by a Media Resource Function (MRF). The core network (EPC/5GC) provides the guaranteed QoS-enabled bearers, typically dedicated EPS bearers or QoS Flows with appropriate QoS Class Identifiers (QCIs/5QIs) for real-time conversational voice, ensuring low latency and packet loss.
VoIP's operation is tightly integrated with policy control. The Policy and Charging Rules Function (PCRF) in EPC or the Policy Control Function (PCF) in 5GC interacts with the IMS via the Rx interface. Upon session initiation, the IMS informs the policy controller of the required media parameters (codec, bandwidth, ports). The policy controller then provisions the corresponding rules in the Packet Gateway (PGW) or User Plane Function (UPF) to establish a dedicated bearer with the correct QoS profile, ensuring the voice packets receive priority treatment across the radio and transport networks. This end-to-end QoS management is critical for achieving carrier-grade voice quality, matching or exceeding that of legacy circuit-switched networks.
Purpose & Motivation
VoIP was introduced to address the limitations and inefficiencies of circuit-switched (CS) voice networks in the era of broadband data. Traditional CS voice dedicated an entire 64 kbps timeslot for the duration of a call, which was resource-inefficient and could not be easily integrated with emerging IP-based data services. The primary motivation was network convergence—creating a single, unified IP infrastructure capable of delivering all services (voice, video, data), thereby simplifying network architecture, reducing operational costs, and enabling seamless service integration.
Historically, 2G (GSM) and 3G (UMTS) networks relied on a separate circuit-switched core network for voice. With the development of high-speed packet access (HSPA) and the move towards LTE (a pure packet-switched radio technology), a new solution for voice was imperative. VoIP over IMS (commmercialized as VoLTE) was standardized to provide a future-proof, high-quality voice service native to the packet network. It solved the problem of the 'voice gap' in LTE and enabled advanced features like simultaneous voice and high-speed data (SVLTE was a workaround, not a solution), HD voice, and rich communication services (RCS) integration.
Furthermore, VoIP standardization ensured interoperability across different operators and device vendors, preventing fragmentation. It established a clear evolution path from CS voice to IP voice, which has been fully realized in 5G as Voice over New Radio (VoNR), where VoIP is the sole native voice solution. This transition supports network simplification, as the legacy CS core can be retired, and enables voice to be treated as just another data application, albeit with stringent QoS requirements, within a fully software-defined and cloud-native core network.
Key Features
- Packet-switched voice transmission using RTP/UDP/IP protocols
- Reliance on IMS for session control and service delivery
- Mandatory support for end-to-end QoS provisioning via policy control (PCRF/PCF)
- Utilization of efficient voice codecs (e.g., AMR, AMR-WB, EVS) for high quality and bandwidth adaptation
- Support for emergency calls (eT911/eT112) and lawful interception over IP
- Enabler for service innovation like HD voice, video calling, and Rich Communication Services (RCS)
Evolution Across Releases
Initial standardization of VoIP as the foundation for IMS-based voice in LTE (VoLTE). Defined the fundamental architecture integrating IMS with the EPC, specified the Rx interface for QoS policy interaction, and mandated support for the AMR codec. Established the baseline for SRVCC (Single Radio Voice Call Continuity) to handover calls to legacy 2G/3G CS networks.
Standardized the Enhanced Voice Services (EVS) codec for superior audio quality and robustness. Introduced IMS Centralized Services (ICS) enhancements for better service consistency. Defined further optimizations for battery life during VoIP calls (e.g., discontinuous reception DRX configurations).
Enhanced EVS codec with new operational modes and improved efficiency. Specified Voice over LTE via Generic Access (VoLGA) was not pursued; focus remained on native VoLTE and VoWiFi. Introduced further optimizations for mission-critical push-to-talk (MCPTT) over LTE, which uses VoIP as a base.
Defining Specifications
| Specification | Title |
|---|---|
| TS 26.935 | 3GPP TS 26.935 |