Research On Resource Allocation Algorithm Of Airborne Sensor Optical Interconnection Network

With the rapid increase in the degree of informationization of airborne systems,the performance of airborne sensor networks is also higher.Wavelength Division Multiplex(WDM)optical networks have gradually become the research direction of nextgeneration airborne sensor networks due to the advantages of high bandwidth,low latency,transparent transmission protocol,flexible and scalable optical networks.Based on this,this thesis studies the related technologies of airborne sensor network resource allocation for WDM optical interconnection network.In order to meet the multicast requirements in the airborne sensor network,the multicast routing and wavelength allocation and multicast traffic grooming algorithms are mainly studied.In the WDM network of airborne sensors,in order to meet the specific service requirements of the airborne network and the quality of service(QoS)constraints in the service,while minimizing the consumption of multiple network resources,this thesis studies multicast routing and wavelength assignment algorithms with the provision of QoS.Based on the systematic description and analysis of the problem,the mathematical model of integer linear programming is proposed,and a multi-objective evolutionary algorithm is proposed which combines non-dominant sorting genetic algorithm(NSGA2)and tabu search algorithm(TS).And it is applied to solve the problem of multicast routing and wavelength assignment that provides QoS guarantee.In the application of the algorithm,this thesis designs a specific individual coding scheme for the specific problems studied in this thesis,and redefines the inter-individual dominance relationship with constraints for specific problems.At the same time,for the genetic operations used in evolutionary algorithms,this thesis designs specific initialization strategies,selection strategies,and crossover and mutation strategies.For the designed algorithm,the distribution characteristics of the solution in the target space and the convergence parameters and diversity parameters defined by the algorithm are compared by simulation.The results show that the average convergence of the algorithm is 40% and 78% more than the comparison algorithm NSGA2_MRWA and SPEA_MRWA,the diversity parameter decreased by an average of 3.84% and 5.11% in the case of a larger request.For the data transmission of airborne sensor optical networks with multiple transmission rates and the mismatch between low-speed data and wavelength bandwidth,this thesis studies the traffic grooming problem.This thesis first analyzes the flow characteristics of airborne sensors,and describes the traffic grooming problem of the lowspeed data transmission.On this basis,a mathematical model of traffic grooming based on optical tree is established.An optical tree shared traffic grooming algorithm based on the bandwidth of progressively decreasing bandwidth is proposed,and the genetic algorithm is used to route and wavelength.This thesis simulates and analyzes different topologies,different scales of traffics,and different types of traffics,and compares them with other algorithms.Firstly,the other algorithms are compared with the consumption of optical resources such as wavelength resources and transmitter resources and splitter resources.Finally,the above factors are combined to compare the normalized network cost.The simulation results show that the traffic grooming algorithm proposed in this thesis uses less network resources than the other two algorithms.On average,the traffic grooming algorithm in this thesis is 58% lower than the network without using the algorithm,and 17% and 29% lower than the comparative algorithms SLTSG and LTSG,respectively.