ETFTN

Extended TFT Support Network

Core Network
Introduced in Rel-11
A network capability that allows a User Equipment (UE) to establish multiple Packet Data Protocol (PDP) contexts or Packet Data Unit (PDU) Sessions associated with a single IP address, each with its own Traffic Flow Template (TFT). It enables sophisticated QoS management and service differentiation for applications running on the same device, such as simultaneous voice, video, and data streams.

Description

Extended TFT Support Network (ETFTN) is a GPRS/EPC core network feature defined in 3GPP TS 23.060. It extends the traditional Traffic Flow Template (TFT) mechanism. A TFT is a set of packet filters (based on IP addresses, port numbers, protocol types) used to classify downlink IP traffic and map it to a specific bearer with a defined Quality of Service (QoS). Normally, for a given IP address assigned to a UE, all traffic matching the TFT of a primary PDP context flows on that context. ETFTN allows the network to support 'secondary PDP contexts' (in 3G/GPRS) or dedicated bearers (in 4G/5G) that share the same IP address as the primary context but have different TFTs and QoS profiles.

Architecturally, the feature involves coordination between the UE, the SGSN (Serving GPRS Support Node) in 3G, or the MME/SMF (Mobility Management Entity/Session Management Function) in 4G/5G, and the GGSN/PGW/UPF (Gateway GPRS Support Node/PDN Gateway/User Plane Function). When an application on the UE requires a specific QoS (e.g., low latency for VoIP), the UE can request activation of a secondary PDP context or a dedicated bearer. This request includes a new TFT that describes the traffic flow for that application. The network authorizes the request based on subscriber policies and establishes a new logical bearer (with its own QoS Class Identifier - QCI/5QI, Guaranteed Bit Rate, etc.) between the UE and the gateway, all sharing the same IP address as the default bearer.

How it works: In the downlink, the gateway (GGSN/PGW/UPF) inspects incoming packets for the UE's IP address. It then applies the packet filters from all active TFTs associated with that IP address. The packet is directed onto the bearer whose TFT it matches. If multiple TFTs match, rules of precedence and evaluation order are applied. In the uplink, the UE performs the same classification, directing its own generated traffic into the appropriate bearer based on the uplink packet filters in its TFTs. This allows the network to provide granular QoS, charging, and policy control on a per-application-flow basis without requiring multiple IP addresses for the UE, simplifying network management and preserving IPv4 addresses.

Purpose & Motivation

ETFTN was developed to address the limitation of the basic PDP context model, which tied a single IP address to a single QoS profile. As mobile devices began to run multiple applications simultaneously (e.g., web browsing, email, voice call, video streaming), they needed a way to obtain different QoS levels for different types of traffic without the overhead and resource consumption of establishing entirely separate IP connections (each with its own IP address). The problem was enabling true service differentiation for IMS services like VoLTE, where voice packets require a guaranteed bit rate and low latency bearer, while background data uses a best-effort bearer.

The motivation came from the need to efficiently support IMS and other real-time services over packet-switched networks. Prior to ETFTN, achieving different QoS levels might have required multiple PDP contexts with multiple IP addresses, complicating routing, charging, and network address translation. ETFTN solves this by allowing multiple QoS pipelines (bearers) to be multiplexed over a single IP address assignment. This is more efficient for the network (fewer IP addresses needed, simpler routing state) and for the device (shared IP stack configuration).

It provides the fundamental mechanism for network-controlled QoS that is essential for delivering consistent service quality for voice and video over IP. It allows operators to implement policy and charging control (PCC) rules that trigger the establishment of specific dedicated bearers for specific services, ensuring optimal resource utilization and user experience. Its evolution through releases has been tied to the enhancement of the PCC architecture and the introduction of new QoS parameters for advanced services.

Key Features

  • Enables multiple PDP contexts/PDU sessions with distinct QoS profiles to share a single UE IP address
  • Utilizes Traffic Flow Templates (TFTs) containing uplink and downlink packet filters for traffic classification
  • Allows dynamic establishment of secondary/dedicated bearers triggered by application demand or network policy
  • Provides granular QoS management on a per-application-flow basis (e.g., separate bearers for voice, video, data)
  • Essential for enabling IMS services like VoLTE and ViLTE with guaranteed bit rate bearers
  • Integrates with Policy and Charging Control (PCC) architecture for bearer authorization and management

Evolution Across Releases

Rel-11 Initial

Introduced the Extended TFT Support Network feature in TS 23.060. Formalized the procedures allowing a UE to have multiple PDP contexts (primary and secondary) associated with one IPv4 address or IPv6 prefix, each with an independent TFT. Established the network architecture and signaling flows for secondary context activation, modification, and deactivation based on TFTs.

Defining Specifications

SpecificationTitle
TS 23.060 3GPP TS 23.060