Sensor systems/networks are becoming increasingly useful platforms to provide solutions for a wide variety of real world applications. Their diffusion, together with their applications have led to a tremendous heterogeneity in the logic for interfacing them as well as collecting data from these systems.
Nowadays the interaction with multiple sensor systems is required by lots of applications, like those typical in the emerging pervasive computed paradigm. Moreover monitoring applications of wide geographical areas have highlighted the research problem of the integrated management and correlation of data coming from various networks that cooperate for a common objective. However in order to incorporate sensing infrastructures into computing paradigms based on the application development for multiple sensor systems, specific integration frameworks for accessing different data sources are needed.
The SeNsIM (Sensor Networks Integration and Management) framework is a more general integration platforms for heterogeneous sensor systems. SeNsIM (i) makes possible the deployment of applications based on multiple sensor systems and (ii) allows a generic user or an application to easily access different data sources. The architectural model for the integration has been realized by exploiting the wrapper-mediator paradigm, in which a mediator module accesses the sensor data by means of ad hoc connectors (wrappers), one for each networks. The mediator is responsible to format and forward user requests to the different networks, while the wrappers are responsible to translate the incoming queries and forward them to the underlying sensors. XML has been used as a modeling language to describe sensor systems in order to facilitate sharing of structured data across them.
Current SeNsIM implementation allows the integration of TinyDB based sensor networks, but we are going to extend wrapper modules to the most common WSN middlewares and to other sensor systems as well.