Scheduling Geo-processing Workflow Applications with QoS

With the development of OGC Web service program (OWS), Web GIS has progressed towards a service-oriented paradigm that enables a new way to distributed geospatial computing. Many applications require geo-processing workflow (GPW) to take advantage of spatial databases or computing resources distributed over Web based their control or task. This paper presents a dual-layer framework to implement geo-processing workflow, in which an abstract workflow is defined in abstract layer to represent a user-specific task flowchart and a concrete workflow is orchestrated in concrete layer to bind geospatial Web services distributed world wide. Within the framework, we present a tree-coded genetic algorithm (GA) approach to address optimization problems in the process of instantiating an abstract workflow as a concrete workflow. The algorithm performance is analyzed through an example and well verified by the result from integer linear programming (ILP).