Current trends towards the Future Internet are envisaging the conception of novel services endowed with context-aware and autonomic capabilities to improve end users' quality of life. The Internet of Things paradigm is expected to contribute towards this ambitious vision by proposing models and mechanisms enabling the creation of networks of "smart things" on a large scale. It is widely recognized that efficient mechanisms for discovering available resources and capabilities are required to realize such vision. The contribution of this work consists in a novel discovery service for the Internet of Things. The proposed solution adopts a peer-to-peer approach for guaranteeing scalability, robustness, and easy maintenance of the overall system. While most existing peer-to-peer discovery services proposed for the IoT support solely exact match queries on a single attribute (i.e., the object identifier), our solution can handle multiattribute and range queries. We defined a layered approach by distinguishing three main aspects: multiattribute indexing, range query support, peer-to-peer routing. We chose to adopt an over-DHT indexing scheme to guarantee ease of design and implementation principles. We report on the implementation of a Proof of Concept in a dangerous goods monitoring scenario, and, finally, we discuss test results for structural properties and query performance evaluation.

No abstract

Maritime-surveillance operators still demand for an integrated maritime picture better supporting international coordination for their operations, as looked for in the European area. In this area, many data-integration efforts have been interpreted in the past as the problem of designing, building and maintaining huge centralized repositories. Current research activities are instead leveraging service-oriented principles to achieve more flexible and network-centric solutions to systems and data integration. In this direction, this article reports on the design of a SOA platform, the "Service and Application Integration" (SAI) system, targeting novel approaches for legacy data and systems integration in the maritime surveillance domain. We have developed a proof-of-concept of the main system capabilities to assess feasibility of our approach and to evaluate how the SAI middleware architecture can fit application requirements for dynamic data search, aggregation and delivery in the distributed maritime domain.

No abstract

Service-Oriented Computing (SOC) is a wide and complex research area. Despite the huge effort in both industrial and academics initiatives, several challenges need to be addressed in order to effectively realize the SOC vision. One of the most relevant issues is the need of effective, flexible, reliable, low cost solutions for dynamic service brokering and composition. This paper presents results of an ongoing work on the design and development of a service-and messageoriented middleware for atomic and composite service brokering, named SAI middleware. The SAI middleware offers a set of features for service brokering and dynamic composition, while also guaranteeing loose coupling between service providers and consumers and relaxing the prerequisites for service providers to publish their capabilities in an interoperability domain. SAI dynamic composition is based on an Artificial Intelligence planning approach and on the adoption of an ontology-based functional profile encoding information for enabling automatic information extraction and combination in the service composition chain. Our main contribution consists in addressing these issues in a holistic way, as required to effectively support the SOA vision in real application scenarios, while not optimizing single aspects yet.