Multi-Protocol Label Switching (MPLS) has the potential to overcome many of the scalability, performance, management and integration challenges confronting carriers as they develop IP networks and services. Yet vendors have failed to capture the confidence of the incumbents by not addressing their needs. Historically, vendors have concentrated on new entrants and promoting niche developments that play to their own strengths.
"The industry needs to focus on those parts of MPLS that have real customers, who will be around long enough to sustain the next round of technology adaptation," says Stevenson.
MPLS provides the predictability of routing performance needed to support differentiated IP services. It addresses fundamental routing scalability problems, and helps carriers manage network capacity.
"It is becoming capable of integrating many services like IP VPN and transparent LAN services over a common switching platform, thereby reducing the cost of offering a full service portfolio," says Stevenson. "It will help reduce the duplication of switching platforms for data services, and also help in the evolution to a next generation voice network."
But while new entrants see MPLS as a key to competition, incumbents have reservations about a less than mature technology. When the industry was still awash with venture capital, vendors tended to concentrate on extensions to MPLS that helped new entrants, particularly in the metro Ethernet market, rather than addressing incumbents' concerns about routing performance and reliability.
"The vendors have been caught between a fading market for radical innovation, and a more stable market that they have neglected," says Stevenson.
MPLS technology is far from perfect and development should be focusing on rugged solutions that meet the needs of incumbents. Yet, development is moving well ahead of real market needs, with vendors interpreting the technology in different ways to suit their strengths and ambitions, and producing disjointed solutions that have proved less capable, flexible and reliable than promised.
"There's a danger that dilution of effort away from the original goal of improving performance, scalability, reliability and resilience of IP networks will delay migration towards next-generation systems, and hold back the introduction of new services and applications," says Stevenson.
While new entrant carriers may be happy to try leading-edge solutions, incumbents are wary of changes that might adversely impact service levels for voice and other legacy services. "Carriers will not migrate major portions of their legacy network onto an IP backbone until they have confidence in MPLS," he says.
MPLS is essentially a long-term technology, concerned with developments to the core of the network that are needed to underpin evolution of all carrier services onto a switched IP core. "This is a very complex issue and the technology must work." The traditional IP approach of making a solution that almost works and then applying sticking plasters won't wash with the incumbents. They need a stepwise approach to network evolution that leverages legacy assets and minimises upgrade costs for recently purchased routing and switching equipment.
The industry dialogue about MPLS must not degenerate into yet another round of inflated claims for IP technology. "A period of stability is needed where the focus is on MPLS solutions more robust so that they can be the basis for service differentiation and sustainable growth," Stevenson concludes.
Related Ovum research: MPLS: Short-term Fix or Long-term Solution? Access@Ovum and IP-Services@Ovum Advisory Services. For more information e-mail: firstname.lastname@example.org or visit www.ovum.com