Topic of Research:
Algorithm Design for Parallel/Distributed Processing Models

Student Researchers:
Jake Krohn and Dan O'Brien

Faculty Advisor:
Dr. Dian Lopez

Aims of Research:
As the physical limits of computational speed are reached and the presence of inexpensive processors in the marketplace increases, it becomes imperative to search for alternative methods of computation that offer increased speed over existing models while still maintaining the affordability and flexibility of current systems.

Our research focuses on the design and testing of benchmarks for a parallel/distributed processor allocation model, taking into account practical limitations such as network latency and communication time between processors. Phase I of our research focused on the design and implementation of a test bed against which the results of parallel/distributed allocation algorithms can be compared. This test bed finds optimal scheduling solutions for NP-hard problems (given small precedence task graphs) through highly recursive brute force computation. These results provide benchmarks in a form that is versatile enough to be utilized by other researchers in the field. Having successfully completed and tested this phase of the research, our current efforts are directed towards the design and implementation of approximation task allocation algorithms that will be tested against our benchmarks.