A Graph-Based Cloud Architecture for Big Stream Real-Time Applications in the Internet of Things

The Internet of Things (IoT) will consist of billions of interconnected heterogeneous devices denoted as "smart objects." Smart objects are generally sensor/actuator-equipped and have constrained resources in terms of: (i) processing capabilities; (ii) available ROM/RAM; and (iii) communication reliability. To meet low-latency requirements, real-time IoT applications must rely on specific architectures designed in order to handle and process gigantic (in terms of number of sources of information and rate of received data) streams of data coming from smart objects. We refer to this smart object-generated data stream as "Big Stream," in contrast to traditional "Big Data" scenarios, where real-time constraints are not considered. In this paper, we propose a novel Cloud architecture for Big Stream applications that can efficiently handle data coming from deployed smart objects through a graph-based processing platform and deliver processed data to consumer applications with lowest latency.