Network transportation and optimization for grid-enabled visualization techniques

Grid Computing denotes an approach to utilize distributed resources which are not subject to centralized control. This approach fulfills computing requirements arising within the context of todays high-performance computing applications, especially in the field of computational science and engineering.The Grid Visualization Kernel GVK aims at exploiting the power of the grid to provide visualization services to the scientific user. The possibly large amounts of data involved in such a visualization process require optimized network transmission in order to achieve the required performance goals and minimize latency for interactive applications. Possible optimization techniques, such as data compression the use of multiple connections, are studied in this paper.