Description
N9 is the 3GPP reference point between two User Plane Functions (UPFs) in the 5G Core. It is a user-plane interface used to forward packet data between UPFs when a PDU Session involves more than one UPF. In the Rel-15 5G architecture, N9 is primarily associated with GTP-U tunnels over IP transport, allowing one UPF to deliver user traffic to another UPF after applying packet detection, forwarding, and QoS enforcement rules. N9 is not a radio interface and not an interface to the Data Network; instead, it connects internal 5GC user-plane functions and enables flexible placement of UPFs across the core, aggregation sites, edge sites, or different network slices.
The main reason N9 exists is to support distributed user-plane architectures. A UE's PDU Session may be anchored at a PSA-UPF close to the Data Network, while traffic may first pass through an Intermediate UPF (I-UPF) near the radio access network. The I-UPF receives user data from the gNB over N3 and forwards it to the PSA-UPF over N9. This allows the network to apply local functions such as traffic steering, uplink classification, branching, or edge routing without moving the session anchor. When the UE moves, the N3 endpoint can change to a new gNB or new I-UPF, while the N9 connection to the anchor UPF can remain stable, preserving session continuity.
N9 is also central to features such as Uplink Classifier (UL CL) and Branching Point. A UL CL UPF can inspect uplink traffic and decide whether to forward it to a central PSA-UPF via N9 or to a local edge UPF via another path. In some deployments, the edge UPF connects to a local Data Network through N6, while the central UPF connects to the public Internet. A Branching Point can duplicate or distribute traffic to multiple UPFs over N9, enabling multicast-like delivery or simultaneous access to different service instances. These capabilities are difficult to achieve with a single monolithic gateway and are a key motivation for the inter-UPF reference point.
The SMF controls N9 indirectly through N4. It selects the set of UPFs involved in a session and installs rules in each UPF that define packet detection, forwarding actions, QoS enforcement, and tunnel endpoints. The UPFs then use N9 to carry user-plane packets with appropriate GTP-U tunnel information, QFI markings, and session context. The transport network between UPFs must provide connectivity and may be shared with other traffic, but N9 itself is a logical reference point defined by 3GPP. This means operators can implement N9 over IP/MPLS, Ethernet, or other transport technologies as long as the UPFs can exchange user-plane packets reliably.
N9 plays an important role in network slicing and edge computing because it allows different UPFs to be placed according to slice requirements. A low-latency slice may use a short N9 path between an edge I-UPF and an edge PSA-UPF, while a broadband slice may use a longer path through central UPFs. N9 can also support multi-homing and traffic steering policies, where different flows in the same PDU Session are forwarded to different UPFs or different DNs. Overall, N9 is the internal user-plane glue that makes 5G's distributed UPF architecture possible, enabling local breakout, session continuity, traffic classification, and scalable deployment of user-plane functions.
Purpose & Motivation
N9 was introduced to solve the limitations of a single-anchor user-plane model. In earlier packet core architectures, user traffic typically passed through a fixed gateway function, and inter-node user-plane interfaces were limited to a small set of functions such as SGW-to-PGW. 5G requires more flexible placement of user-plane functions because of edge computing, network slicing, ultra-low-latency services, and distributed deployments. N9 provides a standardized reference point between multiple UPFs so that traffic can be forwarded through a chain or tree of user-plane functions without collapsing all functions into one node.
The motivation for N9 is closely tied to UL CL, Branching Point, and local breakout. Operators need to steer some traffic to a central network and other traffic to an edge network, sometimes within the same PDU Session. A single UPF cannot easily provide both centralized anchoring and distributed edge processing while preserving session continuity. By allowing one UPF to forward traffic to another UPF over N9, the architecture can separate the access-side UPF from the anchor-side UPF and place each function where it is most useful. This reduces backhaul traffic, lowers latency for edge applications, and supports service isolation.
N9 also supports scalability. As traffic grows, operators can deploy additional UPFs near the radio edge or in regional aggregation sites. N9 allows those UPFs to work together while the SMF centrally controls their forwarding rules. This is important for cloud-native 5G Core deployments, where UPFs can be instantiated dynamically. Without N9, the 5G user plane would be less flexible, and many advanced traffic steering and edge computing features would require non-standard or proprietary inter-UPF mechanisms.
What changed in its specifications
The latest release-to-release comparison of specifications that define N9. Each summary describes the whole specification, not only this term.
Rel-19 added MPQUIC steering to multi-access PDU sessions, XR and media handling, and an Energy Information Function. It also added Local Offloading Management for edge computing, Vertical Federated Learning for AI/ML, and Ambient IoT text.
Full comparison in TS 23.501 →Technology family
Core network interfaces and reference points · Core Network
- N9 counterpart in another generation: S5
“The UPF shown in Figure 4.2.10-1 can be connected via an N9 reference point, instead of the N3 reference point.” TS 23.501
Members and relations were proposed independently by several models and kept only where at least two agree; quotes are verified verbatim against the specifications. Release = first approved version of the defining specification.
Detected Changes Across Releases
verified against 3GPP sourcesDocumented changes in 1 releases, each verified against the text of 3GPP specifications or approved change requests and shown with the exact wording. The release timeline above shows the same releases. A release without an entry has no documented change for this term.
The N9 reference point between two UPFs is defined for the first time in the 5G system architecture specification.
- “N9: Reference point between two UPFs.” TS 23.501 v15.13.0
Defining Specifications
3GPP specifications that define or reference N9, with the latest known release. Sourced from the 3GPP document catalog — see methodology.
| Specification | Title | Release |
|---|---|---|
| TS 23.501 vk30 | 5G System Architecture | Rel-20 read |