Research On Scheduling Method Of Multi-level Cryptographic Service In Cloud Environment

Cloud computing has become the latest development direction of distributed computing.As a new computing mode,cloud computing has the characteristics of convenience,high reliability and on-demand service.However,the emergence of various types of job requests and large amounts of high concurrency problems in cloud environment brings great challenges to cryptographic services.Facing the complex cloud computing environment,the shortcomings of traditional scheduling methods and existing cryptographic service scheduling methods in solving the above-mentioned problems are increasingly highlighted,and the attributes such as user types and processing capabilities of computing nodes are different in the cloud environment.Reasonable and effective scheduling methods are the key to effectively handle job requests and provide high-quality services in the cloud environment.To enable the system to provide high-quality cryptographic services to the outside world,this thesis constructs a multi-level scheduling model from the perspective of users and the system,and proposes a scheduling method that integrates user attributes,task attributes and node attributes.The main work and innovations of this thesis are as follows:(1)Analyze the demand of users and system scheduling in cloud environment,and build a multi-level scheduling model.Considering the security and accuracy of scheduling and the overall performance of the system,a multi-level scheduling model is built based on the actual scenarios of heterogeneous node performance in cloud environment.The purpose of parallel processing of different job requests is achieved by integrating cryptographic computation units of various algorithms such as encryption,decryption and Hash.Through multi-level scheduling processing,mapping with computing nodes is completed according to different job requests with different user attributes,thus completing processing of job requests.(2)Based on the constructed model,a scheduling method integrating user attributes,task attributes and node attributes is proposed.Considering the security requirements of users for job request processing,a scheduling method based on user attributes is proposed by analyzing user attributes to establish a mapping with corresponding rank queues.By analyzing the task attributes in the job request and classifying according to different task attributes,the mapping with the specific algorithm buffer is established to simplify the complexity of subsequent scheduling.Considering the various types of computing node resources,heterogeneous performance and other node attributes under the cloud environment,from the perspective of system completion time balance,the method of selecting job requests for computing nodes is adopted,and the appropriate job requests are selected by weighing the difference of completion time of each computing node,so that the matching between job requests and specific computing nodes is realized,and the reliability processing of job requests is further realized.Finally,based on the model proposed in this thesis,it is verified by C/C++ simulation under Windows.The load balance of the scheduling method proposed in this thesis is analyzed from different angles,and other algorithms are selected for comparative verification in the maximum completion time.The experimental results show that the scheduling strategy and method proposed in this thesis can more flexibly process job requests and node matching,and is more suitable for cloud computing environments with diverse requirements.