Dedicated EPS Bearer

Description

A dedicated bearer is an EPS (Evolved Packet System) bearer that is associated with uplink packet filters in the UE and downlink packet filters in the PDN Gateway (P-GW), where those filters match only certain packets. Unlike the default bearer, which carries all traffic not matched by any dedicated bearer filter, a dedicated bearer is established to provide a specific Quality of Service (QoS) treatment to a particular traffic flow or set of flows, such as VoLTE media, streaming video, or mission-critical push-to-talk. The packet filters are expressed using Traffic Flow Templates (TFTs), which map IP packet headers (source/destination address, port numbers, protocol type, and optionally IPv6 flow labels) to a specific bearer.

Architecturally, the dedicated bearer sits within the EPS bearer model defined in 3GPP TS 23.401. Each EPS bearer is a logical tunnel between the UE and the P-GW, composed of a radio bearer over the LTE-Uu interface, an S1 bearer over S1-U between the eNodeB and the Serving Gateway (S-GW), and an S5/S8 bearer between the S-GW and the P-GW. A dedicated bearer is always established in the context of an existing PDN connection and therefore always coexists with a default bearer for that PDN. The default bearer is typically a non-GBR (Guaranteed Bit Rate) bearer, while a dedicated bearer may be either GBR or non-GBR depending on the QoS Class Identifier (QCI) and Allocation and Retention Priority (ARP) assigned to it.

Dedicated bearers can be initiated either by the network (network-initiated, typically triggered by the PCRF/PCF via the P-GW when an application requires specific QoS) or by the UE (UE-initiated, using the Bearer Resource Allocation/Modification procedures defined in 3GPP TS 24.301). The signaling path for network-initiated dedicated bearer activation runs from the PCRF through the P-GW, S-GW, MME, and eNodeB to the UE, using NAS messages such as Activate Dedicated EPS Bearer Context Request and the corresponding E-RAB Setup procedures over S1. The UE stores the uplink TFT and applies it to outgoing packets, while the P-GW stores the downlink TFT and applies it to incoming packets.

QoS enforcement for a dedicated bearer relies on several parameters: the QCI (which defines packet forwarding treatment such as scheduling weight, admission threshold, and queue management), the ARP (which determines whether a bearer can be established or modified under resource constraints and whether it can preempt other bearers), and for GBR bearers, the GBR and Maximum Bit Rate (MBR) values for both uplink and downlink. The MME and eNodeB use these parameters to decide admission control and scheduling. If the network cannot maintain the GBR, it may modify or release the bearer.

Dedicated bearers are fundamental to differentiated service delivery in LTE and are the mechanism by which operators enforce policy and charging rules from the PCRF. They enable services such as IMS voice (QCI 1), conversational video (QCI 2), real-time gaming (QCI 3), buffered streaming (QCI 6/7), and operator-specific services to receive appropriate treatment even when the default bearer is congested. In 5G, the equivalent concept is the QoS flow within a PDU session, but the dedicated bearer model remains relevant for EPS interworking and for UEs operating in EPC.

Purpose & Motivation

The dedicated bearer concept was introduced in Release 8 as part of the EPS architecture to solve a fundamental limitation of earlier GPRS/UMTS bearer models: the inability to provide fine-grained, flow-specific QoS treatment within a single PDN connection. In GPRS and early UMTS, QoS was largely tied to the PDP context, and establishing multiple PDP contexts for different services was cumbersome and inefficient. The dedicated bearer model allows a single PDN connection to support multiple bearers, each with its own QoS profile and packet filters, so that different applications can receive differentiated treatment without requiring separate IP addresses or PDP contexts.

This design addresses the need for operators to enforce policy and charging control (PCC) at the flow level. The PCRF can dynamically trigger the establishment of a dedicated bearer when an application requiring specific QoS is detected, ensuring that services like VoLTE receive guaranteed bit rates and low latency while best-effort traffic continues on the default bearer. It also enables the UE to request additional resources for specific flows, supporting applications that need to signal their QoS requirements.

By separating the always-on default bearer from service-specific dedicated bearers, the architecture provides both ubiquitous connectivity and service differentiation. This was a key enabler for the widespread adoption of VoLTE and other real-time services over LTE, and it laid the conceptual groundwork for the QoS flow model in 5G.