Application of Grid computing for designing a class of optimal periodic nonuniform sampling sequences

Designing periodic nonuniform sampling sequences for digital alias free signal processing is a computationally extensive problem where sequential single computer based solutions could easily run for days or even weeks. In order to reduce computation time, the sequential algorithm needed to be parallelized making it possible to execute parts of the calculations on different nodes of a computational Grid at the same time, reducing the overall runtime of the application. This paper presents and compares two different Grid based implementations representing the two main Grid approaches at the moment. The first solution utilizes a production Grid environment based on GEMLCA and the P-Grade Grid portal, and the second represents a BOINC desktop Grid-based solution.