Research On Trusted Cooperative Service And Computation Offloading Strategy In Edge Computing

With the development and popularization of the Internet of Things,big data and 5G technology,cloud computing can no longer meet the massive data and computing tasks generated by network edge devices.Edge computing arises at the historic moment.By offloading computing tasks to computable devices at the edge of the network,edge computing relieves the computing load of the cloud center while expanding the computing resources of network edge nodes,providing a new solution for realizing " local and nearly " services.The selection of offloading nodes,comprehensive performance evaluation and cooperative service strategy directly affect the quality of service of edge computing tasks offloading,which is the key research issue of trusted cooperative service in edge computing.The main research work of this paper is as follows.The scheme of cooperative service in edge computing based on the comprehensive performance evaluation(CSEC-CPE)is proposed by constructing task-driven cooperative virtual service pools based on node comprehensive performance evaluation.The probabilistic prediction model of node connection is constructed to suppress the Matthew effect.In addition,Direct trust and recommended trust were integrated to build a calculation model of comprehensive trust of nodes.Combined with the factors affecting the selection of members such as comprehensive trust,robustness,and resource sharing,a comprehensive performance evaluation model was built to select members.Then,based on task driven,the leader node quickly selects credible,robust,and excellent performance candidate member nodes according to the comprehensive performance evaluation of the candidate nodes.The member nodes complete the autonomous integration of edge computing resources and build virtual service pools.Then CSEC-CPE develops several factors including the load balancing ability,package loss rate,latency and so on.The game model of cooperative service utility is established to prove that it has Nash equilibrium stability.Based on the open data set of Route Views,a marine edge computing simulation system is designed and implemented to simulate cooperative services,computation offloading and interactive behavior in edge computing.Experimental results demonstrate that,compared with KNN and K-means clustering algorithm,the proposed scheme in-creases the aggregation degree of members by 64.4% and 51.56%,the average aggregation ability of the system by 34.5% and 15.67%,and the comprehensive trust degree by12.66% and 28.99%,respectively.Besides,compared with the AODV and SR routing algorithms,the task reception rate is improved by 64.41% and 51.65%,and the collaborative service success rate is improved by 52.54% and 40.63%,respectively.In the constructed trusted cooperative service pool,the rational use of the energy of the members is conducive to the benign offloading of computing.And when the energy of the node is exhausted,the disconnection will easily cause the network to jitter,and the task cannot be completed.Therefore,the energy consumption of edge nodes is used as the evaluation factor of task offloading in the trusted cooperative service pool,and one of the trusted clusters is used as an application scenario,and a migration strategy model with energy consumption constraints is established to improve nodes caused by energy exhaustion.In this way,the instability of the edge network caused by the energy exhaustion of the node and the service quality of the edge computing is improved.According to the computing power,energy and other demand characteristics of the source node task and the remaining energy and computing capacity of the edge node,the task is offloaded to a suitable node for execution.A hybrid offloading strategy based on node energy in edge computing is proposed.This strategy mainly adopts a hybrid offloading mode of uplink and parallel by analyzing the requirements of tasks.When the task calculation amount is huge,the uplink offloading with energy consumption as the constraint is adopted,and the parallel offloading—intra-cluster execution,with the delay as the constraint is adopted for the delay-sensitive tasks.Taking into account the heterogeneity of edge nodes and the power difference,this paper uses the remaining power supply time of the edge node to indicate the state of the node,and comprehensively considers the load and energy utilization rate of the edge offloading node,and transforms the offloading energy minimization problem into the residual energy maximization problem.Additionally,constrained optimization problems are directly solved with high algorithm complexity.In this paper,particle swarm optimization(PSO)and simulated annealing(SA)algorithms are used to optimize HOS-NE,improve the stability and throughput of the network,and achieve load balancing.Experimental results reveal that compared with random execution algorithm and greedy algorithm,HOS-NE based on intelligent algorithm optimization increases the network throughput and remaining energy by 16.7% and 28.6%,respectively,and reduces the packet loss rate by 5%?10%.