Negotiated Quality of Service

Description

Negotiated QoS is a core concept in 3GPP bearer and QoS management, defined in TS 21.905 and TS 23.057. When an application or user requests a certain QoS profile, the network does not simply accept or reject the request; instead, it evaluates the requested QoS attributes against currently available network resources and negotiates each attribute down to a level that can be supported. This negotiation ensures that the network can commit to providing the agreed QoS without over-committing resources. The negotiated QoS profile becomes the binding contract between the user equipment (UE) and the network for the lifetime of the bearer or QoS flow.

Architecturally, negotiated QoS is handled by the session management function (SMF) and policy control function (PCF) in 5G, or by the GPRS support nodes (SGSN/GGSN) in earlier generations. The UE requests a bearer or QoS flow with a desired QoS profile, including parameters such as guaranteed bit rate (GBR), maximum bit rate (MBR), QoS class identifier (QCI) or 5QI, allocation and retention priority (ARP), and others. The network then checks its resource availability, policy constraints, and subscription data. If resources are insufficient, the network may propose a lower QoS. The UE can accept or reject the negotiated profile. Once accepted, the bearer network is obligated to attempt to provide adequate resources to support all negotiated QoS profiles.

The negotiation process is dynamic and can be triggered by changes in network conditions, such as congestion or handover. For example, during a handover, the target cell may not have the same resources as the source cell, leading to a re-negotiation of QoS. Similarly, if the network experiences congestion, it may downgrade the QoS of existing bearers to maintain service for all users. This dynamic adjustment is essential for maintaining stability and fairness in multi-service networks.

Negotiated QoS is closely tied to the concept of bearer-level QoS and flow-based QoS. In 4G LTE, QoS is enforced at the bearer level (EPS bearer), while in 5G, it is enforced at the QoS flow level within a PDU session. Despite these architectural differences, the principle of negotiation remains: the network must ensure that the QoS it grants is actually deliverable. This prevents the network from promising more than it can deliver, which would lead to service degradation and user dissatisfaction.

The term has been present since Rel-4 and continues to be relevant through Rel-19, reflecting its enduring importance. The definition has remained essentially unchanged, indicating a stable and well-understood concept. However, the implementation details have evolved with new radio and core network technologies, such as the introduction of network slicing, where QoS negotiation may also consider slice-specific resource availability.

Purpose & Motivation

Negotiated QoS exists to solve the fundamental problem of resource scarcity in telecommunications networks. Networks have finite capacity, and not all QoS requests can be granted at the desired level. Without negotiation, a network would either have to reject requests outright or accept them and risk failing to deliver, leading to poor user experience. Negotiation allows the network to gracefully degrade QoS to a level it can support, while still providing service. This is particularly important in mobile networks where resources fluctuate due to mobility, interference, and varying traffic loads.

Historically, early packet-switched networks like GPRS introduced the concept of QoS profiles but lacked sophisticated negotiation mechanisms. The need for negotiation became apparent as networks began supporting multiple services with different QoS requirements, such as voice over IP, video streaming, and web browsing. The 3GPP defined negotiated QoS in Rel-4 to provide a standardized way for networks to manage QoS requests in a fair and efficient manner. It addresses the limitations of static QoS provisioning, where QoS parameters are fixed and cannot adapt to changing conditions.

Negotiated QoS also supports policy control and charging. By negotiating QoS, the network can enforce policies based on subscription, service type, and network conditions. It enables operators to offer differentiated services and to manage network resources effectively. The concept has been extended in later releases to support new services like mission-critical communications and massive IoT, where QoS requirements can be diverse and stringent.