On Improved Modeling of Survivability-Supported Traffic Engineering in Multi-Service Networks

摘要

In this plenary lecture, a number of optimisation models and iterate heuristic algorithms that address the survivability-supported Traffic Engineering (TE) problem in multi-service networks for multimedia are presented. In such networks traffic demands with different Quality of Service (QoS) and survivability requirements (e.g. existence of a node disjoint backup path for each primary path) inhere. The optimisation models for engineering the QoS traffic with different survivability prerequisites and the Best-Effort (BE) traffic are based on special admission control/routing Mixed Integer Programming (MIP) and Linear Programming (LP) optimisation sub-problems, which are solved sequentially. LP relaxations of the MIP sub-problems are also provided. The iterative heuristic TE algorithms are based on a modified version of the Dijkstra's algorithm. The above methods are used for the solution of the TE problem in two different networks and their performance is compared. An integrated approach based on the aforementioned work, for prioritised 1:1/1+1 protection and restoration-supported TE in multi-service networks, in the case of single or multiple node/link failure(s), is also presented and tested.