CISE Research Resources: Resources for Research in Scalable Parallel Computing and Networking Simulation

EIA-0130847 Gopal Gupta德克萨斯大学达拉斯CIE研究资源：可扩展并行计算和网络模拟研究资源PI寻求获得一个由六个4-CPU多处理器组成的网络和一个与100Mbit/S以太网连接的数据存储系统。该设备将用于构建可扩展的并行和分布式系统，用于表逻辑编程(LP)，用于计算几何项目，以及用于定量研究移动自组织网络(MANET)。逻辑编程项目寻求将备选方案的动态重新排序与堆栈拆分相结合，以在SMP的分布式网络上实现或并行表LP系统。计算几何项目涉及加权区域最优轨迹问题的可扩展并行和分布式算法的设计和实现的研究，在手术规划、地理信息系统和放射治疗中的应用。并行实施将在开关电源网络上进行。MANET项目将对MANET中的单向无线链路进行定量评估：干扰和电池寿命对单向链路发生的影响，以及此类链路的MAC子层问题。我们将通过模拟来评估同时使用双向链路和单向链路的路由协议。多处理器网络和数据存储系统将分别用于进行大型模拟和存储大型模拟数据。

EIA-0130847 Gopal GuptaUniversity of Texas DallasCISE Research Resources: Resources for Research in Scalable Parallel Computing and Networking SimulationThe PIs seek to acquire a network of six 4-cpu multiprocessors and a data storage system connected with 100 Mbit/s Ethernet. This equipment will be used for building scalable parallel and distributed systems for tabled logic programming (LP), for computational geometry projects, as well as for quantitatively studying mobile ad-hoc networks (MANETs). The logic programming project seeks to combine dynamic reordering of alternatives with stack-splitting for realizing or-parallel tabled LP systems on distributed networks of SMPs. The computational geometry projects are related to research on design and implementation of scalable parallel and distributed algorithms for the weighted regions optimal trajectory problem, with applications in surgery planning, geographic information systems, and radiation therapy. The parallel implementations will be carried out on the network of SMPs. The MANET project will conduct a quantitative evaluation of unidirectional wireless links in MANETs: the impact of interference and battery life on the occurrence of unidirectional links, and MAC sub-layer issues for such links. Routing protocols designed to use both bidirectional and unidirectional links will be evaluated via simulations. The network of multiprocessors and the data storage system will be used, respectively, for conducting large simulations and for storing large simulation data.