文摘
New kinds of services and data flows, such as cloud computing, server virtualization, or quality of service (QoS)-aware Internet services, have led to higher demands on network resources. Thus, they have also led to increased bandwidth consumption and unexpected congestion, especially in backbones. In this article a novel proposal is studied, with the aim of improving the performance of delay-sensitive flows. The congestion impact on those QoS-aware flows that require a high level of reliability and delay is analyzed. This management proposal has been implemented on multiprotocol label switching transport profile (MPLS-TP), which is a packet transport networks technology that provides QoS by means of flow differentiation in Internet backbones. In order to improve performance in congested networks, a guarantee of service is proposed. It allows local recovery of lost traffic of MPLS-TP privileged flows, and in order to fulfill the requirements for implementation on MPLS-TP, a minimum set of extensions to RSVP-TE has also been proposed to provide self-management capable routes. Finally, in this paper we have carried out a performance improvement measurement by means of an analytical model and simulations. Keywords MPLS-TP Congestion RSVP-TE Guarantee of service Local retransmissions