Energy-efficient Data Replication in Virtualized Sensor Networks with Decoupled Service Model

Traditional single-application Wireless Sensor Networks (WSN) can be resource-inefficient when multiple such WSNs are deployed in the same geographical area for similar applications. This can be addressed by deploying shared sensor/IoT networks which allow a physical network infrastructure to be virtualized and shared among multiple applications belonging to different clients. The aim of this paper is to focus on reducing the overall energy expenditure of such a system in which the network infrastructure and the services offered to end-clients are decoupled via virtualization. To achieve this, we formulate the concept of data replication for serving multiple data service providers and develop an algorithmic framework for deciding and placing suitable data replication locations. Extensive simulation is performed to demonstrate that the algorithm provides better network-wide energy consumption when compared to existing algorithms on non-virtualized single-application networks.