Abdelsalam Heddaya - US grants

Project BURDS introduces event based distributed data replication as a practical method for including redundacy in distributed systems. Data redundacy helps improve the availability of data under failures, and may also enhance the response time of data access by substituting local accesses for remote ones. Traditionally, data is replicated by copying a data object's state, or value. BURDS explores use of partially replicating state transition (or event) histories instead, in an attempt to show that they are more suitable for representing abstract data types in distributed computing environments. In addition BURDS, research intends to develop an experimental technique that ensures that results of performance comparisons between different replication methods are repeatable and analyzable. The complexity, the deep layering, and the non determinism of the environment of a distributed system render these goals particularly hard to achieve. However, recently available tools acquired for BURDS have explicit support for these goals. Finally, the implementation activity of BURDS should contribute to theoretical investigations into problems that are illuminated by BURDS. Two examples of these are: (1) modeling the potential information flow patterns in networks that may fail by partitioning, and (2) transforming an event history into a value.

A workstation cluster will be provided for researchers at Boston University for research in the Department of Computer Science. This equipment is provided under the Instrumentation Grants for Research in Computer and Information Science and Engineering program. The research for which the equipment is to be used will be principally in the area of reconfigurable reliable data storage, but will also be used for database management, performance analysis, knowledge acquisition, and database query optimization.

This award is for the development of an interdisciplinary curriculum for undergraduate students in computer science, the natural sciences and engineering. The curriculum will focus on massively parallel computing and will use Boston University's 64-node CM-5. The courses will be project- oriented emphasizing direct experience with computational problem-solving in the sciences and with programming in a massively parallel environement. This award is for the development of an interdisciplinary curriculum for undergraduate students in computer science, the natural sciences and engineering. The six project oriented courses that are proposed: one in fundamental methods and five in advanced computational topics, will emphasize direct experience with computational problem- solving in the sciences and with programming in a massively parallel environment. The results of the curriculum and materials development will be disseminated through summer workshops, publications, and presentations at professional meetings.

CDA-9529403 Chen, Marina Bestrides, Azer Heddaya, Abdelsalam Boston University Real-Time, Multimedia and High Performance Computing in Distributed Systems The requested instrumentation is to enhance and support a rapidly growing research group in areas of high performance computing and communication, digital multimedia, and real-time systems. The instrumentation consists of a 12 x 12 ATM switch connecting nodes that serve as compute engines, files servers, and video clients which provides high bandwidth access to the university's supercomputing facility. Five research projects will directly benefit from the requested instrumentation: 1. Parallel compilers, 2. Performance prediction of high-bandwidth networking environments connecting a heterogeneous set of computers and supercomputers, 3. Efficient multi-threaded shared memory parallel computing on distributed systems, 4. Information dispersal and retrieval protocols to provide for real-time, fault-tolerant comunication, and 5. 3D extension on deformable shape modeling for content-based retrieval for video and image databases. The advancement of these projects depends on emerging technology of high-bandwidth networking environments connecting a heterogeneous set of computers and supercomputers, and the availability of a network that delivers a guaranteed bandwidth and real-time graphics and video I/O capabilities.