Technologies for Greener Internet of Things Systems

The Internet of Things processing paradigm arises from the ever increasing tendency for decentralization of the hardware used for processing. Innovative services combine the use of cloud and mobile computational resources in order to provide intuitive and agreeable user experiences. Furthermore, the internet of things paradigm extends to include sensor networks and distributed data collection infrastructures. The essential implication is the necessity to handle big volumes of possibly streaming data. For the case of streaming data, storage in a data warehouse is in general not feasible; a real-time initial processing phase is necessary that will sample, select, organize or summarize the available information. Such systems are categorized in the Big Data Processing paradigm. The two paradigms are correlated, since Internet of Things applications are beneficial when there are “a lot of Things” and hence the amount of data classifies as “Big Data”. The internet of things information processing paradigm is one of the rare occasions where the demand for Green Computing systems does not compete with the need for performance. Indeed, the volumes of data that need to be processed are overwhelming to such an extent that approaches which use unlimited amounts of power, for processing, storage and the associated hardware cooling are simply not feasible. Additionally, remotely operating components that form the distributed processing network, such as smart mobile devices or remote interconnected sensors, need to be Green if they are to be viable. This research focuses on algorithmic developments that make the real-time collection, summarization, analysis and decision making based on streaming data greener.