Research On Task Offloading Decision Method For Edge Computing In Industrial Internet

As modern industry continues to develop in the direction of intelligence,the need for computing resources for industrial equipment has become more urgent.Traditional cloud computing cannot meet the demand for real-time computing resources and data security protection in industrial scenarios due to its architecture.As a new computing paradigm after cloud computing,edge computing places servers on the edge side of the network close to end devices and can provide low-latency and secure computing services to industrial devices through computational task offloading.Computational task offloading policies can have a huge impact on the execution effect of industrial computing tasks,and unreasonable task offloading cannot improve the performance of industrial Internet edge computing.At the same time,the explosive growth of offload traffic puts tremendous pressure on the transport network,and the uneven distribution of multiple edge servers due to the amount of offload tasks also poses a load balancing problem.To address the above problems,this paper introduces data compression technology and load balancing technology into edge computing task offloading to carry out research,and proposes a joint data compression and security protection edge computing task offloading method and a joint load balancing and fuzzy security protection multi-edge server task offloading method,the main work is as follows.(1)An optimization model with joint data compression and security protection for computational offloading in a single-edge server-multi-terminal scenario is established.In this model,in order to address the load of oversized tasks on the link bandwidth in the Industrial Internet,a data compression model is established to determine the optimal compression ratio by modelling the compression ratio as a non-linear function of the computational load of compression and decompression,which in turn reduces the transmission delay of the tasks.In addition,a security protection model has been established to set different security levels for each task in order to improve the security of the data and to meet the computing requirements of different tasks.In order to solve the task offloading strategy,a task offloading algorithm based on an improved simulated annealing-particle swarm algorithm is designed.Simulation results show that the offloading model has significant effects in terms of data security protection,latency and cost reduction,with a 17.41% reduction in the target value after adding the compression model.Compared to existing offloading algorithms for edge computing,the proposed algorithm has a better level of convergence and offloading effect,and can reduce the weighted cost by 27%.(2)An optimization model with joint load balancing and fuzzy security protection for computational offloading in a multi-edge server-multi-terminal scenario is established.In this model,a load balancing model is established in order to offload tasks to appropriate edge servers,taking into account the CPU,memory and bandwidth of the edge server and using the idea of variance to calculate the load balancing variance.The idea of triangular fuzzy number is also introduced into the security protection model to evaluate its security level.A two-layer task offloading algorithm based on adaptive genetic and particle swarm algorithms is designed to solve this task offloading strategy under the constraints of task offloading decision,load balancing and data security.Simulation experiments show that the proposed model is more effective in reducing the weighted target value,and the target value is reduced by 14.5% after adding the load balancing model,and also the proposed algorithm has better performance compared with other two-layer optimization algorithms.