PERFORMANCE ANALYSIS OF ROUTING WITH GENETIC ALGORITHM METHOD IN OBS NETWORK

• Main Author: SIREGAR (NIM 23206335), SIMON
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/13116
• Institution: Institut Teknologi Bandung

Recent studies show that OBS architecture involves a routing and wavelength assignment (RWA) problem which is NP-complete. A common solution of routing problems in OBS networks by using Open Shortest Path First (OSPF) protocol. OSPF is defacto routing protocol in current interdomain IP networks. Because of the RWA problems in OBS network, to find exact solution a set of link-weights that optimizes network performance by using OSPF will be to complex too be solved ini reasonable computation time.<p>In this thesis, we proposed a routing algorithm using genetic algorithm (GA) to solve the RWA problem in OBS networks, which is developed in NS-2. This routing will be used in network simulation and will be compared with the existing routing in OBS networks. The perfomance metrics used in this thesis are end-toend delay, probability of blocking dan throughput. Simulation results show that the proposed routing scheme can reduce probability of blocking about 30 %, increase througput about twice of the existing one and still have an optimum end-to-end delay even with higher value.