High Performance System Modeling and Performance Evaluation for Grid Computing

초록

This paper introduces a Quantization-based System Modeling (QSM) which performs reduction of data traffic requirements in grid computing. The QSM is based on a quantization of system state and provides a high performance grid computing. For realization of QSM in grid computing, this paper introduces how to perform data transmission reduction using quantization approach and suggests the more efficient concept and realization, e.g. quantization prediction. For a realization of the approach, we develop a state update transmission reduction component, e.g. quantization prediction component, which provides behavior and characteristic of a quantization prediction concept. A kinetics of spaceship is taken for a case study to evaluate performance. Empirical results, performing data traffic reduction and system execution time saving, apparently show system performance improvement through the quantizationbased approach in grid computing.