Research On Grid-based Remotely Sensed Data Processing Node

The spatial information can be collected, shared, organized, and processed with SIG (Spatial Information Grid). SIG is the infrastructure that manages and processes spatial data according to users' demand. Remotely sensed data is one of the most important spatial information sources, so the research on architectures and technical supports of Grid-based remotely sensed data processing node is the significant part of the research on SIGBased on the current research progress of Grid technology and SIG, a detailed research on the Grid-based remotely sensed data processing node has been conduced in this thesis. Firstly, the architectures of Grid-based remotely sensed data processing node is presented and the functions involved in the architectures have been expatiated on. Node needs Grid-based security and internal managing mechanisms. In this thesis a tasks scheduling algorithm is also presented based on priorities and multiply queue. With the Grid middleware, the method of internal tasks monitoring and resource monitoring are given in the thesis. The remote sensing models and algorithms have to be registered and updated in SIG The conception of remote sensing meta-model is proposed, which is the mode of service registering, updating and the implementation specification. Generally, the remotely sensed data processing models are complicated and the amount of the data is large. A research on the mode of remote sensing algorithms based on Grid computing has been conducted. The conditions of usage and analysis of executing efficiency in each mode are presented with examples. At the end of the thesis, the key technology of each mode structure and some of the middleware examples are also presented.