SGER: Distributed Spatial Partitioning Algorithms for Scalable Processing of Mobile Location Queries

IIS-0640219 Ling Liu lingliu@cc.gatech.eduGeorgia Institute of InstituteSGER：用于移动的位置查询的可扩展处理的分布式空间划分算法对移动的对象的位置查询的监视和评估已经成为许多基于位置的应用的必要条件，例如车队管理、货物跟踪和位置感知广告。 两个主要的成本决定了系统的性能和可扩展性：移动的客户端的位置更新的无线通信成本和在数据库服务器的位置查询评估成本。本研究计画设计并发展一个分散式空间位置感知架构与一套空间分割演算法，以最小化位置更新与位置查询评估的成本，同时提供有效且可扩充的位置服务。该项目首先利用对象安全区域和查询监控区域来利用空间局部性，然后开发了将查询监控区域与对象安全区域仔细结合的技术以及有效划分服务器负载并降低能量和带宽消耗的算法。一个概念验证原型的分布式空间划分方法的优势，传统的以服务器为中心的方法或纯对等的方法在系统的效率和可扩展性。由此产生的产品预计将有一个显着的实际影响广泛的目前和未来的移动的基于位置的应用。原型实施和实验研究还将用作研究生和高年级本科生课程的教育工具，涵盖移动的计算和先进的互联网应用开发，并将通过以下网站传播：http://disl.cc.gatech.edu/MobiEyes/web。

IIS-0640219Ling Liu lingliu@cc.gatech.eduGeorgia Institute of InstituteSGER: Distributed Spatial Partitioning Algorithms for Scalable Processing of Mobile Location Queries Monitoring and evaluation of location queries over mobile objects has become the necessity of many location-based applications, such as fleet management, cargo tracking, and location aware advertisement. Two predominant costs determine the system performance and scalability: the wireless communication cost for location updates of mobile clients and the location query evaluation cost at the database server. This research project designs and develops a distributed spatial locality aware architecture and a suite of spatial partitioning algorithms to minimize the cost of both location updates and location query evaluations, while providing efficient and scalable location-based services. The project first exploits the spatial locality by capitalizing on object safe region and query monitoring region, and then develops the techniques for carefully combining the query monitoring region with the object safe region and the algorithms for effectively partitioning the server loads and reducing the energy and bandwidth consumption. A proof-of-concept prototype is produced to demonstrate the advantages of the distributed spatial partitioning approach over the conventional server-centric approach or the pure peer to peer approach in terms of system efficiency and scalability. The resulting products are expected to have a significant practical impact on a broad range of present and future mobile location-based applications. The prototype implementation and experimental research will also be used as an educational tool for graduate and senior undergraduate courses covering mobile computing and advanced Internet application development and will be disseminated via the following Web site http://disl.cc.gatech.edu/MobiEyes/web.