Matching Factor Based Spectrum Allocation And Defragmentation For Time-varying Over-wavelenth Services

Recently, with the development of big data, cloud computing and data center services, time varying feature of over-wavelength services brings heavy pressure to the spectrum efficiency of optical networks.Meanwhile, exponential growth transmission traffic put forward the requirement of evolution to SDM (Space Division Multiplexing) to the elastic optical networks.With the support of National 863 project, this paper aims at the problem of resource management and defragmentation in elastic networks.The strategy of spectrum assignment and defragmentation based on matching factor is proposed. Our algorithm utilizes the time varying feature of over-wavelength super-channel in elastic optical networks to realize the spectrum sharing between super-channels, and solves the problem of resource efficiency in elastic optical networks and its evolution of SDM networks. The main contribution and innovations of this paper are summarized as follows:Firstly, from the perspective of time varying feature of over-wavelength services, we propose the concept of matching factor and corresponding spectrum allocation for super channels. Matching factor describes the complementary degree between rate function of super-channels in time domain, thus it can realize the optimal spectrum sharing. The simulation results show that proposed algorithm achieves significant decrease of the network blocking probability.Secondly, to address the fragmentation issues in elastic optical networks, this paper starts with the super-channels of time varying over-wavelength service, studies the entire network oriented flexible grid defragmentation strategy. With the support of matching factor, our strategy measures matching degree between all established time varying super-channels to allocate them in an optimal order in elastic optical networks. Simulation results show that proposed strategy significantly optimizes the network performance.Thirdly, to solve the problem of spectrum assignment with complex constraint condition in the evolution to SDM networks, we present the concept of "Contaminated Area" based on inter-core crosstalk in the context of multi-core-fiber (MCF) . With the combination of matching factor and contaminated area, a resource allocation algorithm for time varying over-wavelength services to reduce potential crosstalk is presented. At last, the simulation results show that this algorithm can reduce over 50% network blocking probability.