Research On Optimal Allocation Of Edge Computing Resources Based On Comprehensive Trust

With the continuous progress of Internet of Things technology,the mainstream computing mode is gradually moving from traditional distributed computing to the era of big data and edge computing.The deep research of edge computing,the security of edge computing resource allocation environment and the optimization of edge computing resource have become the most important issues worth considering at present.This dissertation mainly explores two aspects and focuses on the optimal allocation of edge computing resources with the shortest time and the lowest cost under the condition of keeping the configuration environment safe and reliable.Under the three-layer structure of cloud-edge cloud-terminal,this dissertation takes the edge cloud as the research center,vertically extends to the marginalization of resources,and marginalizes the original centralized computing,storage and caching functions,so as to provide high-quality services close to users.Horizontal extension is the global nature of resources.The edge cloud is regarded as a resource pool,and the coordination and unity of resources are realized through the coordination between the edge cloud and the edge cloud.Firstly,trust model is added to edge calculation.The direct trust is calculated by using the past interaction history and time decay factor.The recommendation trust is determined by combining the value,difference and reliability of the recommendation nodes.On the basis of direct and recommended trust,information entropy is added,respective weights are corrected and summed to obtain the comprehensive trust.The improved trust model provides a secure configuration environment for resource interaction between edge clouds.Secondly,resource allocation of edge cloud is analyzed,and an edge computing resource allocation model based on comprehensive trust is established,with running time,computing cost and trust as optimization objectives.Aiming at the deficiency of Non-dominated Sorting Genetic Algorithm ?(NSGA-?),a Entropy-Non-dominated Sorting Genetic Algorithm ?(E-NSGA-?)based on information entropy was proposed.The comparison of E-NSGA-? algorithm with NSGA-? and Strength Pareto evolutionary algorithm(SPEA2)verifies the effectiveness of the improved E-NSGA-? algorithm.Finally,the model is solved by ENSGA-? algorithm.Through simulation,it can be concluded that under the condition of satisfying the constraint,on the basis of high efficiency and low cost,this model also considers the reliability of resource allocation of edge cloud.E-NSGA-? improves the convergence of resources,makes the distribution of resources in the edge layer more uniform,and finally realizes the optimal allocation of resources.