Over the past twenty years, I have focused my
research on parallel and distributed systems, including architectures,
algorithms and applications.
During 1990’s I focused mainly on
parallel and distributed simulation (PADS). PADS deals
with the execution of a large simulation program on parallel/distributed
computers. The main motivation behind PADS is that many practical simulations
are computation bounded and require large amount of computer memory. PADS
research is aimed at (i) reducing the execution time
of the computation bounded simulation programs (in some cases from several days
to a couple of minutes) by means of parallel processing techniques, and (ii)
partitioning and running large simulation models (those that cannot be run on a
single computer due to the memory limitation) on a number of processing
elements (e.g. a shared/distributed memory multiprocessor or a cluster of work
stations). From an academic point of view PADS is very challenging, since it
addresses several of the most important issues of parallel/distributed
processing, e.g., partitioning, load balancing, synchronization, efficient
memory management, efficient IPC, programming and debugging of
parallel/distributed systems, and hardware supports for parallel programming.
In the
recent years, my research has been on interoperability and composability
of simulation models, symbiotic simulations, semantic Web, service composition and context aware service
discovery, optimization of business process models, and application of IT in
health care.
=============