Research On Distributed Edge Computing Resource Allocation Optimization Method Based On Intelligent Prediction

Most of the existing resource allocation models and algorithms for edge computing and fog computing have the following two assumptions by default:one is that the edge resources are limited,and the other one is that the demand of user equipment for resources is determined(or can be expressed by a constant).However,in the existing research,these two assumptions do not necessarily meet the needs of the real market.In most cases,the demand of user equipment for resources is not determined,but changes dynamically.Therefore,the uncertainty of resource demand needs to be considered in the optimal allocation of resources.In this paper,we believe that the allocation of edge computing resources is carried out in an uncertain,dynamic and predictable environment.Therefore,this paper proposes a distributed edge computing resource allocation optimization method based on intelligent prediction.The main contributions of this paper are as follows:(1)The real data sets are collected,cleaned,analyzed and sorted by MATLAB,and the time series of Internet data are established.The SARIMA model combined with grid search in time series prediction is used to predict the extracted Internet data,model the regional data corresponding to the specific ID of a specific city,and then use three performance standards:MSE(mean square error),RMSE(root mean square error)and MAE(mean absolute value error)to evaluate the prediction effect of the model.In addition to achieve the effect of experimental prediction,random data samples for optimization are randomly generated.Because the data is collected from different cities and different regions.Therefore,after establishing the prediction model,we also consider whether the different models established in different cities or regions have applicability.(2)Based on the realization of data dynamic prediction,we propose the system model in the research of distributed edge computing resource allocation optimization method based on Intelligent Prediction and the edge computing resource allocation problem based on prediction.The computing resources in resource allocation are defined in the computing model,and the communication model is used to define the communication resources in the resource allocation cycle.A benefit function is used to obtain the cost of providing edge computing services.The solution algorithm of the problem is given.Finally,test samples are generated and numerical simulation is carried out.Numerical simulation shows that the change of benefit function value is basically consistent with the change of sample data volume.The model in this paper can effectively solve the optimized resource allocation scheme according to the change of predicted user requested data volume.