A Framework for Delivering IoT Services with Virtual Sensors
Nandini Mukherjee, Sunanda Bose, Himadri Sekhar Ray · 2020
Sensing and cloud technologies are two primary ingredients for developing IoT services. However, there are several challenges which need to be handled for delivering uninterrupted sensing services over the internet. These challenges include: handling heterogeneous sensing devices, non-interoperability issues due to different standards and protocols, acquiring sensing devices from external network by IoT applications, reducing data transmission from the sensing devices to cloud, and handling huge amounts of unstructured and semi-structured data from the cloud. Over the last decade, many IoT platforms have been proposed by the researchers, and solutions have been developed based on these platforms. However, the platforms are non-interoperable, and solutions are developed based on their own IoT infrastructure, devices, APIs, and data formats. In this research work, a framework is proposed based on sensor virtualization at the IaaS layer of the cloud. Sensors are treated as the members of a shared resource pool, just like computational resources in a cloud platform. These sensors can be acquired on a demand basis, shared by different applications, and released when no longer required by the applications. An architecture is developed for plugging the physical sensors with the cloud platform and virtualizing them to handle heterogeneity and facilitate shared access. The proposed solution has been implemented in remote healthcare delivery applications. In this chapter, a kiosk-based application is discussed, and its implementation and its impact on the society are presented as a case study.