Research On Duplication Location Strategies In The Mobile Computing System

There are many categories of computer network，such as wireless networks, cablenetworks, hybrid networks. With the number of mobile host gradually increased, the demandfor mobile access is growing, ubiquitous mobile computing era will shift from personalcomputing era. Terminal equipment of a computer or other information in a wirelessenvironment use mobile computing to achieve data transmission as well as the sharing ofresources, and use it facilitately to transfer the data to any of the needs of the client, so as toachieve accurately, timely, reliable data services.However, in the mobile computing system, data information services are not stability,when the data are erroneous，a node or link is failure，data is not available，in addition，withthe accumulation of time，data is increased，the nodes that storing data catach a accident witha big probability, this will cause some data not to be available. Therefore duplicating data intime and storing data on multiple nodes is important, we transfer resource to nodes thatresponsing requirement of mobile computing timely, and this will avoid affecting the datafault-tolerant performance becourse of the failure of network and computers and can fullyimprove the data access service quality.The paper puts forward duplication location strategies in the different mobile computingsystem. According to the specific characteristics of different mobile computing system, wedesigned the replica placement scheme which is adapting to specific environment. In the fixedbase station, this paper puts forward AFRR strategy according to access frequency, inselecting the duplication of storage node, with the consideration of node storage space andaccess patterns, this paper puts forward AMTT strategy, to reduce transmission time ofaccessing data. At the same time using parallel replication to replace duplication of a singledata, will reduce the time consuming of data replication. Without base stations, the ALmethod is improved, nodes maintain access record, and the grouping of all the mobile nodesis managed, data needs is to be shared between multiple nodes, and this can reduce theaverage path length, also improve the query succeed rate. In the environment of multiple basestations, because the nodes are moving from this base station to that base station. Weconsidered mobile nodes and access characteristics. According to response timecommunication cost and RNC node access frequency, we distribute the duplication in different way, and optimize the replica placement scheme.Finally, through the experiment, we evaluate the performance of data replica placementstrategy, and the performance of the algorithm focuses on comparing different constraints, weverify the placement strategy of this paper to improve the quality of data access service.