一种分布式环境中的近邻检测方法

 为了提高分布式环境中近邻检测的效率,本文提出了一种基于近邻框的检测方法.近邻就是在用户周围一定物理距离范围内其所关心的朋友.在大规模近邻检测中,一般的检测方法研究都关注减少系统内用户客户端和服务器之间位置更新消息的数目,以降低服务器的负担.本文使用近邻框概念,通过移动用户之间位置关系的简单判断来取代欧氏距离或者最短距离的计算,来提高系统的处理效率.同时在用户客户端结合地图信息对自身的移动区域进行自适应地预测,减少客户端和服务器之间的消息交互.论文讨论了近邻框检测的一般步骤,检测过程中疑似近邻用户的处理,并对系统性能展开了分析,通过实验验证了近邻框检测的可行性.实验结果表明本文的方法在不同环境下都能较好地完成近邻查询,方法中使用的优化技术可以显著提高系统的整体效率.

A Proximity Detection Algorithm in Distributed Environment

A proximity box based detection algorithm is presented to improve the efficiency of proximity detecting in distributed environment.Proximity detection is to find each pair of friends such that the distance between them is within a given threshold.Servers in Location-Based Services are likely to be the bottlenecks in large scale proximity detecting,therefore,the state-art-of proximity detection methods are designed with the goal to reduce the server load.In this paper,we propose a detecting method based on proximity box instead of the Euclidean distance or the shorted distance,thereby improving the processing efficiency of the system.At user client,an adaptive method with the road map information is introduced to predicting the user’s security moving zone for reducing the message exchange between the client and server.The detailed steps of the proximity detection algorithm are given and approaches to deal with suspected proximity user are discussed.An indexing algorithm for moving users and proximity relations between friends is also discussed.The experimental results indicate the algorithm is effective in a real city map.The results also prove that proposed index algorithm can significantly improve the overall efficiency of the system.