Routing frequency slot allocation method based on evolutionary multiple objectives in elastic optical network

Abstract

The invention discloses a routing frequency slot allocation method based on evolutionary multiple objectives in an elastic optical network, and mainly aims to solve the optimization problem of routing and spectrum allocation in the elastic optical network. The method comprises the following specific steps: inputting network topology information and initial resource configuration information; selecting initial resources for static services to configure K candidate paths; performing routing and spectrum allocation chromosome coding to obtain a parent population P<t>; performing crossover and mutation operations to generate a new population Q<t>; combining the initial parent population P<t> with the population Q<t> obtained by optimization to generate a new population R<t>= P<t> is a union of Q<t>; forming a next-generation population P<t+1> by an elitist strategy; and performing iteration to obtain an optimal routing and spectrum resource allocation scheme in the elastic optical network. Through adoption of the method, the frequency slot number and the blocking rate are minimized; link congestion is reduced; spectrum resources are allocated equally; service congestion is reduced; the network resource utilization ratio is increased; a plurality of resource allocation schemes with large bandwidth change ranges or high spectrum utilization ratios are provided for operators or specific to the demands of data center interconnection and the like; and the bandwidth demands of different applications are met.

Description

technical field [0001] The invention belongs to the field of network communication, and mainly relates to simultaneously optimizing resource consumption and blocking rate in an elastic optical network, in particular to a routing frequency slot allocation method based on evolutionary multi-objective in an elastic optical network. It is suitable for solving RSA problems of route selection and resource allocation in static services. technical background [0002] With the explosive growth of Internet services, how to improve spectrum utilization and meet the needs of different users has become an engineering problem to be solved urgently. The traditional wavelength division multiplexing WDM network has the advantages of high transmission rate, high cost performance, and wide application. It uses a fixed-size wavelength as the minimum granularity, and a guard bandwidth needs to be reserved between adjacent wavelengths, so it is difficult to flexibly adapt to different granularity...

AI Technical Summary

Problems solved by technology

[0007] There are many studies on multi-objective routing and wavelength allocation (RWA) algorithms in WDM networks. The RWA problem in WDM networks is to allocate wavelengths for user service requests on existing optical paths or new optical paths. RSA in elastic optical networks The problem is to allocate spectrum for user service requests in the same environment, but the solution to the RSA problem is more complicated, because when solving this problem, the change range of the protection bandwidth and the bandwidth required by the service must be considered. At present, the literature on multi-objective research in elastic optical networks is less Less, more normalization methods are used, the implementation is more complicated, and the existing research cannot well apply multi-objective optimization to solve the actual problem of elastic optical network RSA, and cannot well realize the relationship between resource usage and blocking rate. compromise

Images