A Software Toolkit for Complex Sensor Systems in Fog Environments

Smart things (such as sensors, and embedded devices) are going to become an essential source of data within the Internet of Things (IoT)—equipped with sensors and actuators; those devices can gather different information not only from their internal states but also about the environment and entities they interact with. A huge variety and amount of data are generated from various devices and sensors, which need to be processed and responded to in near real-time where the cloud is becoming a dispensable part of that process. This data is used as a basis for various kinds of decision algorithms, machine learning, and artificial intelligence in general. Hence, the amount of data that is being created and sent over the network is vastly increasing. More and more traffic in the network becomes a burden for low-bandwidth and high-latency networks. Purely cloud-based solutions are not able to overcome these issues as the physical distance between the user, edge devices, and the cloud services is an essential factor for transmission latency and response times. The next logical step is pushing cloud services to the edge of the network—to the devices collecting the actual data and at the same time processing data as well. Acquiring data and performing decisions locally on those devices instead of a physically distant cloud server will significantly reduce the amount of data that is sent through the network, reduce the required bandwidth and increase data security. These are the main principles of fog computing, which we will use as the fundamental paradigm developing a software toolkit considering the locality and connectivity of the devices in the underlying infrastructure to facilitate the integration of smart things within the network edge.