The Cost Model And Its Optimization Based On Distributed System In Moving Object Connection Operation

The cost model can provide a reference model to solve the problem in mining the valuable information of the hidden problems,save the time and reduce the cost.Through the cost model,we can estimate the cost function based on past and present information to help people make appropriate decisions.Through the cost model,we can reveal the cost source of each step in the specific query operation,abstract the problem and even reveal the essence of the problem.Through establishing the cost model,the parameters of the influence become obvious,which is beneficial to the computer processing,and the monitoring of the known parameters can also be directly monitored and calculated by computer.The maximum effect of the cost model is the prediction at the unknown cost and the evaluation when the cost is known.Transplanting the cost model to distributed system can predict the cost of the query based on distributed system before running,in order to achieve the purpose of optimization.It also can find the way which can be optimized through the estimated cost of actual consumption cost and model comparison after the query.By establishing the cost model of moving objects based on distributed system,the evaluation of new index can calculate the cost in advance,to determine the feasibility of the index,avoiding wasting a lot of time in nothing.Meanwhile,we can predict the cost in advance before the behavior is done.To be compared with other indexes,we could choose a more suitable method to improve query efficiency.Porting the cost model to a moving object is suitable in the theme of the development of the world.With the rapid development of intelligent devices,it is necessary to optimize the query of moving objects.And the evaluation and prediction the optimal cost before optimized will make the query optimization of the moving object be more reasonable,more based and more convenient.The main research work and contributions are summarized as follows:(1)An optimization algorithm of the selection plan of semi-join query based on distributed database.It proposes a query plan selection algorithm based on semi-join query on distributed database.According to the different locations in querying,various query schemes based on semi join cost differently,so that we can choose the most suitable query plan to improve query efficiency.In order to improve the query efficiency of distributed database system,a new cost model is used to evaluate and transmit the cost of execution and optimization before executing the semi-join scheme.Because of eliminating the irrelevant data to the connection,the useless data in the connection operation relationship is effectively reduced,and the execution method with less execution cost can be selected.Firstly the query execution cost model for distributed database is analyzed.Secondly,the join operator mode,the transmission method of join relation,meanwhile execution sites in semi-join are researched.The execution costs of evaluation methods are calculated,and a feasible query plan selection algorithm is presented.Finally,the execution site,connection method and transmission method are determined.(2)Distributed index and cost calculation of moving objects in traffic data based on star topology.The traffic data is collected according to the sensors distributed in each position,so that the data is gather by place.And according to the index design of star network topology,a distributed index based on star topology is proposed.This implements a variety of mobile object query algorithm,and gives the index query processing cost calculation by the index.