Research On Static And Dynamic Data Placement Strategy In Fog Computing

With the gradual maturity of Internet of things technology,the access to a large number of devices has led to explosive growth of data.Traditional cloud computing services are more and more difficult to meet the needs of users for low delay.Fog computing has come into being.It extends cloud computing to the edge of the network,reducing the pressure on cloud computing centers and shortening the time for user requests to be responded to.Due to the large number of fog nodes,wide geographical distribution,heterogeneous performance and dynamic characteristics,it is a huge challenge to place data on the fog computing platform to meet users’ real-time access and data security requirements.Existing studies have shown that the data placement problem on fog computing platforms is an NP-hard problem.The existing data placement strategies either only consider the time performance of the task,that is,the fastest response time,and take the security performance as the constraint of the problem;or only consider the security of the data,and sacrifice the time performance.But it is a vital goal for users both to ensure data security and to meet low latency needs.In view of this,this paper established a static data placement model and a dynamic data placement model,respectively,by considering the time performance and data security in the fog computing platform simultaneously as optimization objectives.And a high-performance algorithm for solving the model is presented.The main research results are as follows:1.The static data placement model and its algorithm in fog calculation are studied.First,metrics for temporal performance and security performance of data placement problems in fog computing platforms are proposed.Then,the static data placement model of the fog computing platform is established based on this,and an interactive multi-objective evolutionary algorithm is designed to solve the model.The innovativeness of the proposed algorithm includes three aspects:(1)reducing the search space of the problem and reducing the temporal complexity of the algorithm by incorporating the user’s preference information in the algorithm;(2)The temporal performance of the data placement policy(the average response time for a user to read and write a request)is improved by placing the data on the fog node with high centrality;(3)The security performance of the data placement policy is improved in two ways: first,to avoid placing the data on fog nodes which are vulnerable to network attack;The second is to increase the geodesic distance of the fog node storing adjacent blocks of data.Finally,the static data placement algorithm proposed in this paper is compared with the existing latest and commonly used data placement algorithms through simulation experiments.The experimental results show that the static data placement algorithm proposed in this paper can obtain a data placement strategy with better time performance and security performance.2.The dynamic data placement model and its algorithm in fog computing are studied.First,the reasons for the dynamic characteristics of the fog computing platform are clarified,and the impact of the dynamic characteristics on the data placement strategy is analyzed.Then,metrics for temporal performance and security performance considering the dynamic characteristics of fog computing platforms are proposed.Based on this,the dynamic data placement model of the fog computing platform is established,and a dynamic multiobjective evolutionary algorithm is designed to solve the model.In the proposed algorithm,the infeasible data placement strategy can be quickly repaired by designing a dynamic adjustment operator,and the unusable fog nodes in the placement strategy can be replaced based on the user’s preference information.Finally,through simulation experiments,the performance of the dynamic data placement algorithm proposed in this paper and the existing latest and commonly used data placement algorithms are compared when the fog computing platform is attacked by the network.The experimental results not only show that the algorithm proposed in this paper can quickly respond to the dynamic changes of the fog computing platform,but also show that the data placement strategy generated by the proposed algorithm can effectively reduce the possibility of data backup being attacked by network.