CISE Research Instrumentation: Parallel Computing For Wireless Networking Research

9729863 Goodman，大卫J. WINLAB（无线信息网络实验室），罗格斯大学 CISE研究仪器：无线网络研究的并行计算 这项研究仪器补助金使以下研究项目： - 电信系统的可扩展模拟。U-NII波段无线系统的干扰和共存。多跳分组无线电网络中的路由选择和功率控制。干扰消除。美国国家科学基金会/工业合作研究中心WINLAB将购买一台高性能14处理器SMP计算机（SUN UltraSparc HPC 4000），以支持无线数据网络中的计算密集型研究项目，这些项目具有相当大的未开发并行性。新的和正在进行的项目，严重依赖于这样的计算平台包括：-虚拟无线测试床的无线电传播，移动性，网络流量和协议的并行模拟的发展。Testbed逐渐发展成为标准的研究和教育工具。在最近分配给无执照国家信息基础设施（U-NII）的5 GHz频段内，对无线系统的干扰和鲁棒协议进行了调查。多跳无线电网络中的功率控制和可行路由方案的研究。新CDMA系统实现复杂多用户检测器的试验台设计。 购买的多处理器SUN UltraSparc将大大加快在这些领域的研究，以及在其他项目，涉及DIMACS（NSF S T中心）和ECE和CS部门在罗格斯大学，目前没有一个类似的高性能多处理器与并行线程支持。

9729863 Goodman, David J. WINLAB (Wireless Information Network Laboratory),Rutgers University CISE Research Instrumentation: Parallel Computing For Wireless Networking Research This research instrumentation grant enables the following research projects: - Scalable Simulations of Telecommunication Systems- Interference and Coexistence of U-NII Band Wireless Systems - Routing and Power Control in Multihop Packet Radio Networks - Interference Cancellation. WINLAB, an NSF/Industry Cooperative Research Center, will purchase a high performance 14-processor SMP computer (SUN UltraSparc HPC 4000) to support computation-intensive research projects in wireless data networking which exhibit considerable degree of untapped parallelism. The new and ongoing projects that critically depend on such a computing platform include:- Development of a Virtual Wireless Testbed for parallel simulations of radio propagation, mobility, network traffic and protocols. The Testbed evolves to become a standard research and education tool.- Investigations of interference and robust protocols for the wireless systems in the recently allocated 5 GHz bands for the Unlicensed National Information Infrastructure (U-NII).- Investigations of power control and feasible routing schemes in multihop radio networks.- Design of a testbed implementing sophisticated multiuser detectors for new CDMA systems. The purchase of the multiprocessor SUN UltraSparc will greatly accelerate research in these areas, as well as in other projects involving DIMACS (an NSF S&T Center) and ECE and CS Departments at Rutgers, which currently do not have a comparable high performance multiprocessor with parallel thread support.