一种基于P2P的移动对象查询框架

近年来,时空数据查询方法的研究成为人们普遍关注的研究热点.但大部分研究主要集中在集中式环境,在分布式环境下对海量时空数据进行高效的轨迹查询和窗口查询是一件十分有意义且具有挑战性的工作.设计了一种基于P2P的解决方案,提出了对移动对象运动空间进行双层划分的方法来同时支持两种查询.应用网格过滤技术有效地解决了数据频繁更新的问题.对运动空间进行高效的划分,具有比空间填充曲线方法更好的负载平衡性,同时设计了高效的Overlay--SmartChord来支持窗口查询.实验结果表明,和现有方案相比所提方案可以有效减少更新通信量,负载平衡性和路由效率有显著提高.

A P2P-Based Framework for Moving Objects Query Processing

Recently,more and more attention has been drawn to research of temporal-spatio queries.But most of the research mainly focuses on the centralized environment.It is very meaningful and challenging to execute trajectory queries and window queries over massive distributed temporal-spatio data.In this paper,a new partitioning method is proposed,in which two tier grids are used to handle moving object data in different granularity.The top layer distributes data in the server to the P2P network,in order to balance the load among the nodes.The base layer filters the in-cell movement of the objects,which avoids intensive extra updates.A novel overlay,SmartChord,has been designed to support window queries,which leads to a higher routing efficiency.The experimental results show that,compared with existing techniques,the scheme in this article can effectively reduce the update traffic,and that both load balancing and routing efficiency have been improved significantly.