Resource Allocation Strategy For Optimizing Spectrum Utilization And Energy Consumption In Elastic Optical Networks

With the popularity of some applications such as three-dimensional videos on demand,e-learning and cloud computing,the demand for network bandwidth is growing exponentially,which brings a great challenge to communication networks.To meet the needs of the future Internet,high-capacity,high-speed and flexible elastic optical networks are considered as a better solution for next generation optical transport network.In elastic optical networks,the spectrum is divided into narrow slot and can choose a different modulation level according to the transmission distance.As the result,the spectrum utilization is greatly improved.However,spectrum contiguity and continuity constraints must be guaranteed by traffic transmission in elastic optical networks,which limits the further improvement of spectrum utilization.With the rapid growth of data services,the transmission capacity of backbone network has faced a great challenge.As the explosive growth of network scale and the great advances of the Internet technology,the power consumption of networks will keep increasing at a very fast speed.What's more,the Information and Communication Technology(ICT)industry has become the world's fifth major energy-consuming industries.Thus,energy consumption has restricted the development of the ICT industry.How to make the network more efficient has become an urgent problem to solve.In this thesis,an efficient spectrum allocation strategy is studied in elastic optical networks.In the premise of improving the spectrum utilization,reducing the energy consumption in elastic optical networks.For the multicast spectrum allocation problem in the static scenario,an energy-efficient multicast grooming strategy based on light-tree split for elastic optical networks is proposed in chapter 3.The strategy firstly finds the the maximum grooming light-tree that can carry the current request from the routed light trees,and grooms the request based on light-tree split to improve the light-tree sharing ratios and reduce the consumption of the transponders.If the maximum grooming light-tree can not carry the current request or does not exist,and the request is about to be blocked because of scanty spectrum resource,the algorithm tries to dynamically divide the destinations into different sub-trees according to the modulation level,which can reduce the traffic blocking rate and avoid the additional spending because of light-tree spitting.In the process of spectrum allocation,the grooming success rate can be improved bypreferentially allocating spectrum for the requests participating in light-tree grooming.Simulation results show that the proposed strategy can significantly reduce the light tree transmission energy consumption and improve the spectrum utilization.Generally,the dynamic set up and tear down of links will fragment the available spectrum resources and induce the spectrum fragmentation problem.For the unicast request that has a known holding time under the dynamic scenario,a holding-time-aware energy-efficient routing algorithm for elastic optical networks is proposed in chapter 4.The algorithm encourages choosing the path with the least additional energy consumption during the holding time of request,and selects the path with the minimal energy cost to improve the energy efficiency.In the process of spectrum allocation,according to the length of business duration,choosing a different spectrum allocation to ensure the continuity of available spectrum.According to the departure time of the request,we sort the spectrum block according to ascending order in the time difference with the average departure time of its adjacent requests.By computing the spectrum coherence degree,the spectral block that has the least impact on the spectrum continuity is determined.Simulation results show that the proposed algorithm can effectively reduce the blocking probability and improve the energy efficiency.