On-the-fly situation composition within smart spaces

Advances in pervasive computing systems have made smart computing spaces a reality. These smart spaces are sources of large amount of data required for context aware pervasive applications to function autonomously. In this paper we present a situation aware reasoning system that composes situations at runtime based on available information from the smart spaces. Our proposed system R-CS uses situation composition on-the-fly to compute temporal situations that best represent the real world situation (contextual information). Our proposed situation composition algorithm is dependent on underlying sensor data (hardware and software). These sensory data are prone to errors like inaccuracy, old data, data ambiguity etc. R-CS proposes algorithms that incorporate sensor data errors estimation techniques into our proposed dynamic situation composition based reasoning system. R-CS is built as an extension to Context Spaces, a fixed situation set based reasoning system. We implement R-CS dynamic situation composition algorithms over context spaces and validate our proposed R-CS model against context spaces' fixed situation reasoning model.Validerad; 2009; 20090705 (jerker