Middleware Solutions for IoT-based Application in Disrupted Network Environments

Thanks to the Internet of Things (IoT), sensors, actuators, computational and storage resources, are getting more and more embedded in everyday life, radically transforming human interaction and perception of their surroundings. More specifically, IoT allows Information Technology (IT) services to exploit a capillary network of heterogeneous devices to acquire real time information and to autonomously interact with the environment. However, due to their distributed architecture, IoT-based services are highly dependent on a reliable network infrastructure that can effectively connect edge devices to IT services and users' devices. Therefore, disrupted network environments prevent the effective use of IoT. Post-disaster urban environments and tactical environments are characterized by an unreliable network infrastructure that negatively affects the capability of nodes to share information. However, middleware can provide an effective tool to enable IoT-based applications in these challenging environments. In fact, by means of a middleware based approach it is possible to define a ubiquitous IoT management layer that can rapidly adapt to the network characteristics and re-enable access to IoT devices. More specifically, middleware can form a federated solution capable of autonomously locating IoT devices in highly fragmented network environments, thus allowing applications to discover and directly connect to such resources. Moreover, middleware can also support the communication between applications and IoT devices by implementing information-centring communication paradigms that better exploit the scarce network resources. In particular, communication paradigms such as publish-subscribe and Information Centric Networking (ICN) decouple the communications' endpoints and enable more resilient information sharing in highly disrupted environments. This thesis investigates the effectiveness of middleware for proactive discovery and the information-centring communication paradigm to enable the use of IoT-based applications in disrupted network environments. To this end, this thesis also presents a distributed discovery middleware for disaster recovery environments and experimentally evaluates the capabilities of multiple information centric communication protocols within tactical networks. The research presented in this thesis is the result of the collaboration with international institutes and a research period at the Florida Institute for Human and Machine Cognition (IHMC), FL, USA.