9617355 Phatak, Dhananjay S. Gupta, Vipul State University of New York, Binghamton CISE Research Instrumentation: Computation Engines and Mobile Computing Testbed This research instrumentation grant enables research in the following projects: -Fault Tolerance of Feedforward Artificial Neural Networks (ANNs) and Synthesis of Robust Nets-Constructive Algorithms to Incrementally Build Feedforward ANNs that Lead to Efficient Hardware Implementations-Novel Number Representations for High Speed and Low Power Arithmetic Circuits- Mobile and High Performance Networking-Network Reconfiguration and Processor Scheduling in Parallel Systems The project supports collaborative research in multiple areas of Electrical Engineering and Computer Science. The equipment consists of computation engines supporting a small wireless Local Area Network and peripherals. It is envisaged to serve a dual purpose: - providing the computational power and graphical environment required by CPU/memory intensive projects in Neural Networks, Signal Processing, and VLSI CAD, and -providing a test-bed for mobile computing research. Neural net research includes synthesis of fault tolerant networks, and constructive algorithms for efficient hardware implementations. VLSI CAD focuses on low power designs crucial for today's laptops and portables. Ultrafast algorithms and VLSI architectures are also being developed for real-time arithmetic processing. As the computing paradigm continues to shift from institutional to personal to mobile/ubiquitous, supporting nomadic users becomes an interesting research challenge. The proposed instrumentation supports exploration of mobility supporting protocols, security and privacy for mobile users, specialized protocols for wireless media and adaptive applications. In addition to these state-of-the-art research projects, the proposed setup also enables novel experiments in integrating conventional server functions (CPU/memory/data-storage/retrieval) with mobility support services.
