What is MPLS?

Multiprotocol Label Switching (MPLS) is a networking technology that directs data from one node to the next using labels rather than a network address. These labels identify paths between the desired endpoints, where the use of network addresses identify the endpoints themselves. Packet-forwarding decisions are made solely on the contents of the labels, so there is no need to examine the packet itself.

MPLS does not directly align with the OSI seven-layer hierarchy and is often referred to as a layer 2.5 protocol. It was designed to provide a unified data-carrying service for circuit-based clients and packet-switching clients that provide a datagram service model. 

Network operators have adopted MPLS because it can encapsulate and forward a range of network protocols. MPLS is ideal for network and service convergence in the service provider and enterprise markets. It is often used with IP but can also carry time-division multiplexing (TDM) traffic.

With MPLS, the first time a packet enters the network, it is assigned a specific forwarding class of service (CoS), also known as the forwarding equivalence class (FEC). This involves appending a label to the incoming packet, often to represent the type of traffic it carries. For example:

• Voice and video services may be labeled to indicate their real-time requirements.
• Traffic for operational technology (OT) and other essential functions could be labeled to indicate its mission-critical nature.
• Internet or email traffic could be given a best-effort label. 

These labels ensure each packet receives the appropriate service quality across the network. MPLS routers establish a label-switched path (LSP)—a predetermined path for forwarding the traffic across the MPLS network—based on the criteria in the FEC. LSPs are unidirectional, so return traffic is sent over a different LSP.

IP/MPLS can guarantee the LSP symmetry required for legacy applications such as line differential protection for power grids. In these cases, traffic engineering policy will enforce the requirement and ensure that symmetry is present in the primary and redundant paths. 

There are two kinds of routers in an MPLS network: edge routers and transit routers. For IP traffic, each packet passes through an MPLS edge router first (and last, as it exits the network). The edge router analyzes the information in the packet header to determine the CoS required for the application data labels the packet accordingly. 

Most innovation occurs at edge routers because vendors develop specialized network processors to perform complex calculations fast enough to match line speed so that the edge router is not adding latency and delay as it ensures QoS.

The label switch router (LSR) or transit router uses the CoS and QoS information in the MPLS label to route each packet through the network. It replaces the incoming label with a new label for the next hop in the LSP.

Today, enterprises and service providers are replacing many networking protocols with IP/MPLS. This is because IP/MPLS is a multiservice network technology that can meet the performance needs of all applications, from the simplest to the most mission-critical. This growing adoption is also driven by its proven ability to support power utility requirements and smart grid applications.

Two other big factors are accelerating IP/MPLS adoption. The first is that application developers are writing IP-based applications. Any vertical business that wants to update its software to take advantage of the latest and greatest innovations must eventually make the move to IP. 

The second factor is the economies of scale created by the pervasive deployment of IP networking systems. Several decades ago, choosing a specialized network technology for a specific application was likely only marginally more expensive. There were many competing technologies and no technology had market dominance. Today, the world has settled on IP and IP/MPLS as network technologies of choice. Since the advent of 4G/LTE, even specialized mobile networks have moved to using IP and IP/MPLS.

Historically, IP/MPLS networks largely replaced circuit-based networks that businesses used to connect branch offices to headquarters. These were originally run on T1 or E1 TDM dedicated layer 2 links, so business-critical QoS could be assured. For similar reasons, TDM was used for mission-critical applications in a variety of verticals, including power utility and air traffic control networks, where delays were simply not acceptable.

The strength of TDM was also its weakness. With lines dedicated to a branch or power utility control network, network capacity often went unused. The great advantage of an IP packet-based network such as the internet is that all users share the basic managed connectivity. This means the service can scale, taking more bandwidth when needed and less when not. 

The QoS and reliability capabilities of IP/MPLS networks ushered in the age of virtual private networks (VPNs) for mission-critical enterprise services and service provider networks. As part of a digital transformation strategy, an enterprise can use IP/MPLS for branch offices and specialized vertical applications and realize the cost advantages of a shared physical infrastructure. 

By using labels, the MPLS network can logically or virtually separate the VPN data from other data traversing the network. Adding line-rate encryption abilities to the edge router ensures that packets are completely secure as they pass across the network. 

MPLS will play a key role as data networking evolves. This ultra-reliable technology is well suited for the delivery of mission-critical services and real-time applications, two areas of increasing importance within enterprise digital transformation strategies.

Segment routing is emerging as a pivotal advancement that builds on the foundation provided by IP/MPLS. It uses the established IP/MPLS protocol stack to deliver scalable and resilient data networking solutions.