Survey indicates longevity for frame, ATM

In a recent survey conducted by the MPLS and Frame Relay Alliance, which we discussed last time, there was some clear consensus as to how the alliance should organize its work items.

The top ranking went to ATM-frame relay-Ethernet service interworking, described as providing any-to-any connectivity when different access protocols are used across an MPLS-based network backbone. Frame relay-MPLS interworking took the second spot. ATM-MPLS interworking received the third position overall and was ranked third by non-service providers, but it only ranked fifth (out of six choices) by the service providers.

These results seem to signal longevity for frame relay and ATM with the addition of Ethernet transport services into the mix.

If there was a surprise, it was the relatively low priority for an MPLS user-to-network interface (UNI), which defines CPE-based MPLS connections. Overall, the ranking was fourth out of six items, although the service providers ranked this third, ahead of interworking with ATM.

This probably has nothing to do with the merit of the work. Rather, an MPLS UNI has faced some really tough competition. In addition, enterprises weren’t part of the survey respondent mix – and an MPLS UNI would directly affect their capabilities and service choices.

Unsurprisingly, voice over MPLS (VoMPLS) took fifth place out the six items. With the local voice infrastructure already seemingly dedicated to transition to voice over IP, providing quality-of-service mechanisms for appropriately transporting IP traffic without translating the formatting to ATM Adaptation Layer-2 is both sufficient and sane. Also, there already exists a VoMPLS implementation agreement.

It’s not good news for the ATM community, though, that the service providers actually ranked this capability above service interworking with ATM.