jump to navigation

What the Heck is a MPLS NGN? April 14, 2009

Posted by HubTechInsider in Fiber Optics, Telecommunications.
Tags: , , , ,

Ever since I heard of MPLS NGN, I have been excited about the potential for the latest backbone networking technology and wanted to find out more about it. After reading through several books on MPLS NGNs, their architecture, the advantages, and what their potential for ILEC provisoners as well as CLEC access providers truly is, I think I am ready to outline the definition of a MPLS NGN, describe in an extremely non-technical way how they work, what they do, and what kinds of services they will enable in the future. I also try and expand just a bit on why I think they are so important, and what kinds of traditional weakness and deficiencies is the networks that have gone before they are able to address. And addressing on the fly is really at the heart of what a MPLS NGN does so well: 

General Architecture of a Multiprotocol Label Switching, Next Generation Network

MPLS is an acronym for Multiprotocol Label Switching. A NGN is a Next Generation Network. 

MPLS was created to address the weaknesses in traditional IP networks. Please recall that IP was designed to support “best effort” services. In other words, routers contain no inherent perception of the existence of or proper functioning of connections or rings; they see the ports and addresses that are available to their discovery via priority cues and routing tables. Simply put, IP routing lacks intelligence. So-called “Least cost” routing was designed to conduct traffic along the network using the shortest possible number of hops, which means traffic on the network could potentially take shorter, congested paths rather than the potentially more efficient longer, uncongested paths, leading to network “hotspots” and degrading network performance.

The MPLS environment, which has been gaining increased attention, was born out of Cisco’s tag switching. MPLS was originally proposed by the IETF (Internet Engineering Task Force) in 1997, with the core specifications being finalized in 2000. MPLS’s ability to plot static paths through an IP network gives service providers the traffic-engineering ability they crave, and the capability for provisioning (in the telecom sense of that word) VPNs is greatly strengthened. In fact, MPLS provides a very solid base for VPNs – and with increased capability for traffic engineering, service providers are able to tightly control and maintain QoS as well as optimize network utilization.

Although technically not an IP network, despite the fact that it can run in routers and uses IP routing protocols like OSPF and IS-IS, MPLS is one of the most significant developments in IP. To truly understand why this is, you also need to know that although it can also use repurposed ATM switch hardware, MPLS is, again technically, not an ATM network. 

MPLS is another type of network entirely: MPLS is a service-enabling technology. Think of MPLS like a general purpose, tunneling technology. As such, it is capable of carrying both IP and non-IP payloads. It uses what is called “label switching” to transport cells or packets over any data link layer throughout the network.

Much like the inband and out-of-band signaling on the PSTN, MPLS separates the forwarding, or transport, plane from the control plane. By so doing, it enables the capability to run the control plane on devices which cannot actually understand IP or recognize the boundaries of incoming packets. MPLS itself is an encapsulating protocol that has the ability to transport a number of other protocols. These protocols are encapsulated with a label that at each hop is swapped. The label is a number, or UID (Unique Identifier) that identifies a set of data flows along a particular logical link. They are only of local significance and they must change as a packets follow along a predetermined path – they literally switch.

MPLS’s potential to untie IP and optical switching under one route-provisioning umbrella is of great benefit, but it was designed to address two problems inherent in IP networks: IP sends all traffic over the same route between two points, and it cannot absolutely guarantee network resources, because as you will recall, IP is a connectionless protocol. These two shortcomings, in times of heavy network traffic, lead to some routes becoming underutilized while others become congested. Lacking control over the routing assignments, the provider cannot steer traffic from congested to less busy routes. So one key differentiator between IP and MPLS is the simple fact the MPLS networks can steer packets between two points along different paths depending upon their switching MPLS labels.


No comments yet — be the first.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: