Summary

MPLS QoS is a fundamental component of an MPLS offering. You do not need to have MPLS to provide a QoS in the IP network because of the options through MPLS, such as DiffServ-aware TE and MPLS QoS. However, it is often argued that an MPLS network provides much better QoS than an IP network.

QoS is typically associated with SLAs and performance guarantees. There are two types of guarantees, and in fact the the word guarantee as used in an SLA can have two different meanings: an absolute guarantee or a statistical guarantee. In practice, providers of IP services use the term statistical guarantee to describe a specified quality level of a certain percent of the time at a given level of demand. This is in fact reasonable because the traffic demand is statistical and in effect unbounded.

An IP/MPLS NGN can provide an SLA that is an almost absolute guarantee for data rates, delay, and packet loss by using a combination of admission control and MPLS DiffServ. The admission control function enables the rejection of calls when the network cannot guarantee QoS. By separating the admission control function from data plane queuing, a compromise is struck between absolute QoS and scalability.

This chapter has covered the basic building blocks of IP QoS and discussed how these building blocks can be used in an MPLS network to deliver QoS to end customers. MPLS QoS provides a means to deliver a stringent QoS model on a packet network. Combined with MPLS TE, MPLS QoS can provide a powerful tool for delivering tight boundary SLAs. For finer control of network QoS, DiffServ-aware TE is a useful tool.