Design and Performance Evaluation of Intelligent LSPs in Multiprotocol Label Switching Networks

Abstract

Multi Protocol Label Switching (MPLS) is defined by the Internet Engineering Task Force (IETF). It is a network expertise that broadcasts traffic efficiently and supports Quality of Service (QoS) on the Internet. In the present research work, fuzzy based control methodology has been adopted that deals with the uncertainties and high variability appearing in the network. The essential design of traffic regulation is successful liberation of packets by implementing fuzzy based congestion control methodology. Based on the systematic literature survey and studying associated matter related to the problem area, a fuzzy based solution is developed that does not change the existing conventional algorithms but uses operations of them in order to improve the network performance. The selection of the fuzzy logic method is due to the simplicity and expert defined rules governing the routing system. It can be modified and tweaked easily to improve the performance. The work presented is an effort of implementing a fuzzy based decision making component for high volume traffic MPLS networks by implementing Traffic Engineering, Quality of Service and Multipath routing. The approach explicitly proves to be successful in solving the issues and challenges pertaining to stability, scalability in high volume and dynamic traffic. Furthermore, it handles congestion by higher link utilization and provides efficient rerouting of traffic along with fault tolerance in the network. Fuzzy Controller consists of two sub fuzzy systems- Label Switched Path setup System (LsS) and Traffic Splitting System (TSS). The computation of dynamic status of Load and Delay is utilized by LsS to arrange the paths in order of preference. The attained Link Capacity and Utilization Rate are employed by TSS for maintaining congestion free path. The system is to facilitate better decision making for splitting the traffic in different promising paths. This is apparent from the series of simulations carried out on different traffic scenarios.