Dynamic configuration for sensing as a service model in the internet of things paradigm

Abstract

The Internet of Things (IoT) is a dynamic global information network consisting of Internet Connected Objects (ICOs), such as RFIDs, sensors, actuators, as well as other instruments and smart appliances that are becoming an integral component of the future Internet. In parallel, the sensing as a service model has emerged as a business solution that enables new business opportunities and adds additional value to IoT technologies. Currently, ICOs outnumber both people and computers connected to the Internet, and their population is expected to grow to 50 billion in the next 5 to 10 years. To be able to develop IoT applications, such ICOs must become dynamically integrated into emerging information networks supported by architecturally scalable and economically feasible Internet service delivery models, such as cloud computing. In this thesis, we proposed a number of novel approaches that enable non-technical users to dynamically configure sensing as a service platforms without knowing technical details. Our overall objective is to simplify the tasks that need to be done in the sensing as a service domain. Therefore, anyone with no technical background can successfully use these systems to achieve their own goals. Further, our solutions also accommodate technical experts to perform their tasks much faster with less effort. Firstly, we propose an approach called Context-aware Dynamic Discovery of Things (CADDOT) which facilitates efficient and effective sensor discovery and configuration. Then, we propose Mobile Sensor Data Processing Engine (MOSDEN), a plug-in based IoT middleware for mobile devices, that allows collecting, processing, fusing, and filtering sensor data without programming efforts. MOSDEN establishes the link between sensors and cloud IoT platforms. Together, CADDOT and MOSDEN allow non-technical users to connect sensors to cloud services easily and efficiently without requiring much technical expertise. Secondly, we propose the Context-aware Sensor Configuration Model (CASCOM) to addresses the challenge of automated context-aware configuration of filtering, fusion, and reasoning mechanisms in IoT middleware. In this cloud solution, non-technical sensor data consumers are allowed to define their requirements in an abstract way. CASCOM is capable of identifying the sensors and data processing components that need to be composed together in order to produce the desired outcome. As outcome, CASCOM produces data streams that can be fed into other applications/services easily where further analysis or visualization may occur. To support CASCOM, we propose the Context-aware Sensor Search, Selection and Ranking Model (CASSARAM) to address the challenge of efficiently selecting a subset of relevant sensors out of a large set of sensors with similar functionality and capabilities. It is evident that sensing as a service platforms require significant improvements in order to bring them to a level that non-technical personnel would be able to use them. This thesis demonstrates some of the solutions and future research directions. Show more