Research On Incremental Group Nearest Neighbor Query Method In Mobile Enviroment

The improvement and integration of computing techonology and wireless communication dose benefit to the improvement of mobile computing techonology.During the mobile environment,user can move from one palce to another without time and space constrait.when user moves from one place to another,the asoiciated imformation should write back to the database.At that time,request result associated with user position will also change and differ as time goes.so,giving some effective query method to cope with user request becomes a highlight just now and also a challenging research task.Given two groups of objects, group nearest neighbor(GNN) query retrieves objects from one group with the minimum sum of distances to all objects in the other group. Some research work has done in this aspect for static objects,but we can't find effective method for moving objects in dynamically changing environment.Re-executing a query after each update is a solution to this problem(REEGNN). However, the query process doesn't make use of the result of previous queries in this method, which needs high cost.We propose an incremental group nearest neighbor query method (INCGNN) .At the beginning of the query,we give the k+1_th group nearest neighbor according to a split list .As the next update time comes,keep the result updated,then we can get the k group nearest neighbor easily in the new update period.As the group nearest neighbor query has its own feature, we also give three query optimize methods, which include the improved k+1 group nearest neighbor query algorithm, the Ellipse-based pruning method and the delayed update method.Experiments result shows that INCGNN outperforms REEGNN significantly and the above optimize methods could enhance the query performance efficiently.At last,we analyse the condition in which our method perform well and the superior aspects.