Multiprotocol Label Switching Architecture

Description

Multiprotocol Label Switching (MPLS) is a data-carrying technique that operates between the traditional OSI Layer 2 (Data Link) and Layer 3 (Network), often called Layer 2.5. Within the 3GPP architecture, it is not a 3GPP-invented protocol but is referenced and utilized as a preferred transport technology for the IP-based core network, known as the Evolved Packet Core (EPC) in LTE and the 5G Core (5GC). MPLS works by assigning short, fixed-length labels to packets as they enter the network. These labels are used to make forwarding decisions, replacing complex IP longest-prefix-match lookups with simple label-based switching. An MPLS network consists of Label Edge Routers (LERs) at the boundary and Label Switch Routers (LSRs) in the core. LERs classify incoming IP packets, assign an appropriate MPLS label based on Forwarding Equivalence Class (FEC), and push the label onto the packet to form an MPLS header. Core LSRs then swap the incoming label with an outgoing label based on their Label Forwarding Information Base (LFIB) and forward the packet along a Label Switched Path (LSP).

The key components of MPLS in a 3GPP context include the LSPs themselves, which are unidirectional paths through the network. These paths can be established via signaling protocols like Label Distribution Protocol (LDP) or Resource Reservation Protocol with Traffic Engineering extensions (RSVP-TE). Traffic Engineering (MPLS-TE) is a critical aspect, allowing network operators to explicitly route traffic along specific paths to avoid congestion, meet QoS requirements, and provide resilience through Fast Reroute (FRR) mechanisms. In the 3GPP core, MPLS tunnels are used to carry GTP-U tunnels (user plane traffic) and GTP-C signaling between network functions over the backbone. For example, the S5/S8 interface between Serving Gateway (SGW) and Packet Data Network Gateway (PGW) or the N9 interface between UPFs in 5G can be transported over MPLS LSPs.

Its role is to provide a scalable, reliable, and efficient transport layer for the mobile core's IP traffic. By using labels, MPLS decouples the forwarding decision from the underlying data link technology (Ethernet, PPP, etc.) and the final IP destination, enabling a unified transport plane. This allows for the creation of virtual private networks (Layer 3 VPNs or L3VPNs) to securely separate traffic from different operators or services. For 3GPP, this means the same physical backbone can carry traffic for multiple mobile operators or network slices with strict isolation. MPLS's traffic engineering capabilities are vital for ensuring that latency-sensitive voice or real-time gaming traffic receives priority over best-effort data, directly supporting the differentiated QoS requirements defined by 3GPP.

Purpose & Motivation

MPLS was adopted into the 3GPP architectural considerations to address the limitations of traditional IP routing for building carrier-grade mobile backhaul and core networks. As mobile networks evolved from circuit-switched to all-IP architectures with 3GPP Release 5 and beyond, the need for a robust, scalable, and controllable transport technology became paramount. Traditional IP routing, while flexible, suffers from unpredictable paths, limited traffic engineering, and slower convergence times, which are unacceptable for real-time voice and mission-critical mobile services.

The historical motivation was to leverage a proven, high-performance transport technology already widely deployed in fixed-line ISP backbones. MPLS solves several key problems: First, it provides faster packet forwarding through simple label swapping, improving network performance. Second, and most importantly, it introduces sophisticated Traffic Engineering (TE), allowing operators to control the exact path of traffic flows to optimize resource utilization and guarantee bandwidth and latency for specific services—a direct enabler for 3GPP's QoS classes. Third, it enhances network resilience with mechanisms like MPLS Fast Reroute, which can restore connectivity in sub-50ms upon a link failure, crucial for maintaining call continuity. Finally, MPLS VPNs provide a natural way to implement network slicing concepts even in the transport layer, offering secure multi-tenancy. By referencing MPLS in specifications like 23.207 (QoS architecture) and 23.802 (architecture enhancements for end-to-end QoS), 3GPP provided a blueprint for integrating this powerful transport technology to meet the stringent reliability, quality, and scalability demands of modern mobile networks.

Key Features

• Label-based forwarding for high-speed packet switching independent of underlying link layer
• Explicit Traffic Engineering (MPLS-TE) for optimal path selection and resource reservation
• Fast Reroute (FRR) for sub-50ms recovery from link or node failures
• Support for Layer 3 VPNs (L3VPN) to create isolated virtual networks over a shared infrastructure
• Separation of control plane (IP routing protocols, LDP, RSVP-TE) and data plane (label switching)
• Ability to transport multiple protocols (IPv4, IPv6, Ethernet) over a unified MPLS backbone

Evolution Across Releases

Defining Specifications