Taming the IoT data deluge: An innovative information-centric service model for fog computing applications.

Fog Computing is a new computation paradigm, recently emerged from the convergence of IoT, WSN, mobile computing, edge computing, and Cloud Computing, which is particularly well suited for Smart City environments. Fog Computing aims at supporting the development of time-sensitive, location-, social-, and context-aware applications by using computational resources in close proximity of information producers and consumers, such as increasingly common cheap and powerful modern hardware platforms. However, realizing Fog Computing solutions for Smart Cities represents a very challenging task, because of the massive amount of data to process, the strict resource and time constraints, and the significant dynamicity and heterogeneity of computation and network resources. These formidable challenges suggest taking into consideration new information and service model solutions that explore several trade-offs between processing speed and accuracy. Along these guidelines, we designed the SPF Fog-as-a-Service platform, which proposes a new information-centric and utility-based service model and allows the definition of self-adaptive and composition-friendly services, which can execute either on edge devices or in the Cloud. In numerous evaluations, SPF proved to be a very effective platform for running Fog services on heterogeneous devices with significantly different computational capabilities while also demonstrating remarkable ease of development and management characteristics.