Information Centric Sensor Network Management Via Community Structure

Self-organizing dense sensor networks are expected to resolve the challenges of spectral efficiency, energy efficiency, and device management. This paper explores the use of information dependence among sensors to form a community structure of sensor data, and uses this structure to develop an information- centric processing methodology to achieve self-organizing dense sensor networks. Moreover, a data aggregator adopts ℓ₁ regularization to easily enhance energy efficiency with uncertain sensor availability. Combining this community structure and these data recovery algorithms, a novel self-organizing device management scheme is proposed to mitigate the sensor maintenance costs. This approach is justified through simulations and analysis, and is envisioned to have future application in the Internet of Things.