Research On Routing And Spectrum Assignment And Optimization In Co-existing Fixed And Flexible Grid Optical Networks

With the further revolution of information technology, computer science technologies have largely processed and Internet applications have widely spread. A new kind of service with huge capacity grows with each passing day,such as high definition video service, traffic request in data center, distributed services and etc. The trend of explosive traffic growth forces researchers and operators to continually explore advanced optical transmission and networking solutions to achieve the purpose of improving network capacity with efficiently networking. In this context, the flexible grid optical network with large capacity,dynamic flexibility and high spectral efficiency, which is based on flexible grid technology, has become the industry consensus as an inevitable choice for the next generation. Different from the traditional fixed grid technology, the flexible grid technology has broken the constraints of the channel center frequency at fixed intervals, more fully adapts to changes in a variety of granularities transmission service request, can provide efficient spectral resources on-demand assignment and establishs flexible optical superchannels.These features will greatly improve the resource utilization to increase the network capacity. The thesis, based on flexible grid technonlogy, focuses on the migration from fixed grid to flexible grid optical network, solves the problem of service provisioning in coexisting fixed grid and flexible grid optical network, considering the capacity limitation of standard single mode fiber, furtherly investigates the flexible optical network based on multi-core fiber, and explores the future optical network routing spectrum control mechanism and optimization technology. The contributions are summarized s as follows:Firstly, aiming at the scenario of migration toward flexible grid optical network, around inter-operate requirements between fixed grid equipments and flexible grid equipmennts, this thesis studies the spectrum resources assignment in coexisting fixed grid and flexbile grid oprical nework, proposes three spectrum assignment constraints in this scenatio: spectrum congruity constraint,spectrum continuity constraint and spectral compatibility constraint, and designs a new kind of migration strategy : measurable migration strategy based on upgrade probability. The proposed strategy is based on the migration from fixed grid to flexible grid, optimizes High Degree First (HDF) strategy and High Traffic First (HTF) strategy by modelizing the fixed grid optical network to realize controllable network migration. By investigating the migration with different parameters, it can be found that the proposed measurable migration strategies have various performances due to face the various scenarios or upgrade demands. But, all of measurable migration strategies point out that the network capacity can be enlarged and bandwidth blocking ratio is also decreased sharply with only a few fixed grid nodes being upgraded to flexible grid nodes. In addition, with processing migration, the results conclude that the benefits from migration is gradually reducing.Secondly, this thesis investigates the routing and spectrum contral mechanism with the network measurable migration in co-exsisting fixed and flexible grid optical network, introduces the concept of Potential Upgrade Nodes Group (PUNG), proposes three different PUNG construction schemes,which are Maximum Value Scheme (MVS), Connected Nodes Scheme (CNS)and Largest Weighted Link Scheme (LWLS), furtherly optimizes the proposed measurable migration to flexible grid optical network, and verifies the influence of different PUNG construction schemes on network migration by simulation. Except the management of network migration, based on the PUNG,it can provision migration-aware services, and proposes Migration-aware Service Provisioning Strategy (MSP). Through simulating and analyzing the proposed strategy, compared to the traditional Non-migration-aware Service Provisioning Strategy (NSP), MSP can reduce connection interruption ratio effectively, and provide stable connections for service requests with the ganrantee of bandwidth blocking ration.Lastly, the gallop of Internet drives network capacity further increase, with the limitation of Shannon Nonlinear Theory, and Standard Single-Mode Fiber(SSMF) in the foreseeable future as the transmission medium has been unable to provide sufficient available capacity for users' requests. Flexible optical network with multi-core fibers based on Space Division Multiplexing (SDM )becomes the trend of next generation optical network. The multi-dimensional resources model not only improves the network capacity but also brings more complex problems about routing and resource assignment. Hence, based on the physical constrains of multi-core fibers, this thesis studies Rouing, Core and Spectrum Assignment (RCSA) algorithm with inter-core crosstalk constraints,and proposes a novel Crosstalk-Aware Provisioning strategy with Dedicated Path Protection (CAP-DPP) and a novel Distance Adaptive Routing, Core and Spectrum Assignment algorithm (DA-RCSA). By simulating the two proposed algorithms on the platform, it can be found that the proposed algorithms can reduce the blocking probability effectively, due to the mitigation of crosstalk,in multi-core fibers elastic optical network (MCF-EON) considering inter-core crosstalk.