Model Checking for Sentient Computing: An Axiomatic Approach

Eleftheria Katsiri, Alan Mycroft
2005 International Conference on Mobile Data Management  
Sentient Computing allows applications to best interact with their physical environment, by becoming aware of their surroundings. Awareness is achieved by means of a sensor infrastructure that helps maintain a model that represents the current state of the dynamically changing world. This model can be seen as a concrete interpretation of the physical environment and conceptually stands between the physical world and the abstract view of applications. A number of factors such as the
more » ... ty of physical space and the precision of the sensor technology may introduce errors and inconsistencies between the physical world and the model. On the other hand, the abstract view of the application domain needs to be correct and compatible with the concrete model, especially in the case of distributed, heterogeneous environments where applications need to interact seamlessly with several different concrete models. The contribution of the work described in this paper is that it looks at the above problem as a constraint-satisfaction problem to which it applies classical, logical satisfiability, in order to produce a model-based solution. Our system is similar to a classical model-checker ; it checks the satisfiability of the application requirements against the world model, as well as the consistency of the world model with the properties of the actual physical environment. Because it is model-based, it is appropriate for distributed, heterogenous environments. Our system forms part of SCAFOS, a generic, distributed middleware framework for contextawareness, implemented in the first author's PhD dissertation. An implementation that uses the theorem prover system SPASS, is also discussed in this paper.
dblp:conf/mdm/KatsiriM05 fatcat:zmojiy6wefh33ewuotf7uukvci