HYBRID BACKUP MODEL WITH GENETIC BRANCH POINT OPTIMIZATION FOR SINGLE ELEMENT FAILURE IN MPLS NETWORKS

Abstract

In this paper, we present an approach for enhancing current MPLS resilience, called Hybrid Backup model (HB), in order to restore a QoS protection path from both node or link failures. At present, there are three main MPLS fault management methods, i.e., global, reverse and local backup models. Firstly, the global backup model is the cheapest model but it suffers from high packet losses and long restoration time. Secondly, the reverse backup model improves the packet loss problem; however, the delay problem still remains. Thirdly, the local backup method is seemed to be the best choice in case of the minimum restoration time and packet losses, nevertheless, it has high cost in term of number of path switch label switching router (PSL), path merge label switching router (PML), label usage, and bandwidth reserved. Therefore, we proposed a new approach based on hybrid of four switching types which are global, global reverse, local and local reverse switching types. To evaluate number and location of branch point, branch point optimization based on Genetic Algorithm is then proposed. The proposed model can be reduce the cost of the local backup model, while it still maintains fast restoration and low packet losses. Furthermore, it can improve some significant network performances such as bandwidth reserved, rejection probability, and total throughput. According to performance comparison between all backup models, numerical and simulation results are presented to support the proposed backup model.