Large-scale context management

Summary form only given. Most pervasive computing systems are context-aware and thus are able to dynamically adapt their behavior to context changes. For many applications the relevant context is limited in both size and scope. However there is an emerging class of applications whose context may include a huge amount of entities possibly dispersed over the entire globe. Those applications, including logistics, production systems, traffic control or energy management, rely on highly dynamic context information that is captured by a huge number of networked sensors and potentially shared by a wide spectrum of applications. Due to the distributed nature as well as the varying dynamics and quality of context information, scalable context management becomes a challenging task. In this talk we give an overview on context management and present some approaches to increase scalability. Finally, we will discuss future research directions in that field.