Software Defined Fog Computing Structure Model And Application Research In Network Defense

With the development of intelligent mobile terminals and the popularity of Internet of Things(Io T)technology,traditional cloud computing architectures have been unable to withstand the rapidly increasing data size and data processing pressure.Fog calculation has been considered as one of the "last mile of artificial intelligence" and one of the important applications of 5G communication technology because of its advantages in terms of computing power marginalization,low latency,and reduced cloud computing burden.However,in the application and deployment scenarios of fog calculation,there are still three problems: 1.There are differences and dynamics in the processing requirements of user data in fog calculation.On the one hand,different fog computing users have different requirements for the data preprocessing strategy in the fog node,and each user's needs exhibit characteristics that change with time;on the other hand,different content types(such as video,audio,etc.)data The pre-processing requirements are also quite different;2.The relative solidification and static characteristics of the existing fog computing node architecture make it impossible to achieve efficient and flexible utilization of the fog computing node resources;3.The current fog computing node architecture is not well Attacks against fog computing systems may result in insufficient security protection for edge networks.The software-defined networking technology that has emerged and flourished in recent years has pointed out a new direction for solving the above problems in fog computing.SDN technology decouples the control plane of the system from the data plane(Data Plane),and greatly enhances the scalability of the fog computing system through the logic centralized controller and the open API(Application Programming Interface,API).Flexible refactoring.However,there is no related research on applying software definition technology to fog nodes and network dynamic defense based on fog calculation.Based on the above-mentioned problems and background,this paper proposes a software-defined fog computing structure model,and discusses the application of the architecture in network boundary dynamic defense.The purpose is to realize the reconfigurable of the fog node architecture and service orchestration through the software-defined method,and realize the efficient utilization of resources in a single fog node while satisfying the user's differentiated requirements.The main work is as follows: 1)The softwaredefined hierarchical design of the existing fog node architecture provides the basis for the flexible,efficient and reusable service combination and arrangement in the fog node.2)The basic operations involved in the data processing in the fog node are abstracted and encapsulated into different service modules,and the resource pool is uniformly added,changed,deleted,etc.to meet the diversity requirements.3)Perform graph-based modeling on the service orchestration process to achieve efficient service reconfiguration and optimization.4)Applying the proposed model to the network security defense,and adding and combining the security function modules to implement a more complex hybrid linkage defense strategy.5)Carrying out the simulation platform construction and testing,analyzing and evaluating the delay cost and benefit of the architecture,which proves that the softwaredefined fog computing architecture proposed in this paper has low delay,loss and good scalability.