Research On Resource Allocation Algorithm Of Ant Colony Optimization For Scheduled Traffic In EONs

With the rapid development of Internet and the diversity of communication traffic, the network bandwidth demand is increasing explosively. The contradiction between the fixed-size wavelength bandwidth allocation in conventional Wavelength Division Multiplexing(WDM) optical networks and the variable bandwidth requirements of user traffic is presenting sharply. By introducing the bandwidth-variable transmission and switching technologies, Elastic Optical Networks(EONs) which can achieve flexible and effective allocation of spectrum bandwidth resources has been proposed in recent years. Routing and Spectrum Assignment(RSA) problem is one of the key problems in EONs. Traditional RSA research focuses on the static traffic and the dynamic traffic, but with the virtualization of network resources and the change of the way on user bandwidth requirements, the resources allocation of the scheduled traffic is becoming a problem to be solved in future application.Sponsored by the project from the National Science Foundation of China(NSFC), the novel intelligent algorithm and formalization of resource optimization in hybrid EONs, this thesis gives a brief review of the development status of optical networks technologies, and summarizes the research status of EONs. It presents the EONs architecture,the key components and the main technologies of EONs. A brief introduction to the Orthogonal Frequency Division Multiplexing(OFDM) modulation technology is given. This thesis makes the analysis on the RSA problem in EONs and discusses the RSA methods and research status. It also addresses the key theory basis of Ant Colony Optimization(ACO) algorithm and traffic models of EONs.This thesis studies on the RSA problem for scheduled traffic in EONs, and tries to optimize two main goals including blocking probability and spectrum efficiency. To solve the RSA problem for scheduled traffic in EONs, we built an Integer Linear Programming(ILP) model which includes two objective functions and some constrained conditions, converting RSA problem into a combinatorial optimization problem. We proposed a RSA algorithm named Ant Colony Optimization for Scheduled Traffic(ACOST) to solve this problem. Taking advantage of a modified ACO algorithm and introducing residual frequency slots ratio to determine the selection probability of next nodes, the path is selected by integrately considering both the time coincidence ratio of the traffic and the weight of such a light path.The performance evaluation software of the proposed algorithm is implemented via MATLAB and simulated in NSF network topology. We compare the simulation results with those from other exiting algorithms to verify the feasibility and effectiveness of ACOST algorithm, and make the analysis on the effect of some key parameters in the proposed algorithm on the simulation results. The simulation results show that the proposed algorithm in this thesis achieves lower blocking probability and improves spectrum utilization, and can be used to solve the RSA problem for scheduled traffic in EONs.