Context map for navigating the physical world

Pervasive computing environments are composed of numerous smart entities (objects and human alike) which are interconnected through contextual links in order to create a Web of physical objects. The contextual links can be based on matching context attribute-values (e.g., co-location) or social connections. We call such a Web of smart physical objects as context map. Context maps can be used for context-aware search and browse of the physical world. However, changes of dynamic context values over time may render a context map inconsistent. So, it is important to update contextual links with changes in specific context values. Given the asynchronous nature of pervasive environments, it is non-trivial to detect events generated by contextual changes in real time. We propose two algorithms for instantaneous and periodic detection of events with concurrent timing relations. Our algorithms have low time complexity and they can address the needs of different types of pervasive computing applications. We have evaluated our proposed algorithms through simulations as well as testbed experiments.