Research On Recommendation And Selection Of Terminal Security Policy Under Edge Computing

With the advancement of the construction of the power Internet of Things,the traditional cloud computing-based service model has problems such as insufficient security protection,network delays or interruptions,and data processing bottlenecks.Therefore,the smart grid has introduced edge computing to overcome the shortcomings of the current power system.The development of this new power grid architecture also brings new security protection requirements.On the one hand,the edge network environment is complex and unstable,which increases the possibility of edge devices being attacked;on the other hand,traditional security protection systems are difficult to adapt to edge terminals with limited resources.Therefore,researching the security of power grid terminals under edge computing,selecting and recommending appropriate security strategies for security protection,is of great significance and necessity for the security applications of edge computing.Based on the existing edge computing security architecture,this thesis studies the edge side terminal security protection strategy,breaks through the previous passive protection system,and selects the "appropriate" security protection strategy by evaluating the terminal risk before the occurrence of security threats.The main research work of this thesis is as follows:(1)This thesis designed an edge computing-side terminal access security protection strategy selection plan,analyzed the security threats that power grid terminals may face,proposed a security risk calculation method,and established a mathematical model for security strategy selection.Besides,a WOA-based terminal security strategy selection algorithm and an improved IWOA algorithm are designed.The simulation results showed that the proposed solution can quickly select a suitable security strategy for edge-side access terminals and recommend it to security managers,reducing the security of the edge computing system.Risk and defense costs.(2)This thesis introduced a real-time recommendation method for terminal security policies based on machine learning.A trained security policy selection model was deployed on the edge to select and recommend terminal security policies in real time.In order to adapt to the dynamic changes of terminal risks and security requirements,methods of offline training and online judgment at the edge were proposed,and a security policy selection model was designed based on the LSTM network.The simulation results showed that the proposed method can continuously update and optimize the security policy selection model,improved the performance of the model,and recommend terminal security policies in real time at the edge.(3)This thesis proposed a security reinforcement method for edge computing terminals based on graphs and potential functions,a risk map of the terminal security protection system is constructed,and importance is evaluated according to the topological potential of the graph nodes,so as to determine the weak points of the terminal security protection system that need to be reinforced.The simulation experiment results show that the proposed method can upgrade and strengthen the key security mechanism of the terminal,improve the performance of the terminal security protection system,and is beneficial to the security management of the edge computing system.(4)The terminal security strategy recommendation was implemented on the power edge computing system platform built,and the terminal hierarchical security strategy recommendation plan was designed using the analytic hierarchy process,and the implementation process was explained in detail with the smart metering system as the application scenario.The solution in this article can be applied to the actual edge computing system platform.It has practical application value.