Research On Optimization Of Data Edge Storage In Fog Computing

Various virtualization resources store in the cloud through central ized cloud computing controlling.Cloud computing provide various s ervices for Internet users to meet their needs through the network.In rec ent years,smart phone,wearable devices and in-vehicle terminals are ex periencing explosive growth,the wireless network sensing and transmiss ion technologies are improving,and the area of wireless coverage networ k is further expanded.Now cloud computing be able to meet the Inter net users'(PC users)services,but unable to meet all new intelligent mo bile terminal diversified services.To adapt the new situation,Cisco has pr oposed the conception of fog computing.According to the mobile users behavior characteristics and consumption demand of intelligent termina 1,some data will be cached in fog computing node near the user by forec asting,based on the idea of providing localized service.Fog computing will process of the end-user's request in real-time,and effectively overc ome shortcoming of the cloud computing.However,the fog computing n ode cached space compared to cloud computing is very limited,in spite of the fog has computing power and storage capacity.Therefore it is n ecessary to selectively store some data from the cloud to meet deman d of mobile terminal users.Specially,some data in fog computing nod e can be replaced when the buffer space is insufficient to make the fog co mputing node keep best storage state.In this paper,we study the placement of cached data in fog computing under the condition that the storage space of fog computing is limite d.We focus on the problem of data replacement when the storage space o f edge storage nodes is insufficient.First of all,analyzing the characterist ics of data storage in fog computing,explaining the related concepts an d the utility maximization model of fog computing edge storage data obje cts is established,called Max-Utility model.After introducing the basic c haracteristics and advantages of ant colony algorithm,the traditional ant c olony algorithm is used to solve the Max-Utility model.Finally,throug h MATLAB simulation,we can test the feasibility of the proposed algorit hm from several indexes such as cache hit rate and response delay.