Architectural Optimization of Hierarchically Organized Wireless Sensor Networks for Energy, Cost, and Security

摘要

Sensor networks are being increasingly used to monitor public and private infrastructure as well as industrial processes. Thousands of sensors distributed over a geographical area periodically take measurements and pass them to a central base-station which is equipped with data analyticscapability or the ability to connect to a cloud service for data storage and/or data analytics. The nodes of the sensor network are organized hierarchically. Security of the transmitted information is addressed through encryption. Elliptic Curve Cryptography is an attractive proposition forsensor networks since it offers a high degree of security while requiring smaller key-lengths and fewer data transmissions. In this paper, we consider a stand-alone hierarchically organized sensor network and a number of architectural and system-level techniques that can help bring down theoverall energy consumption and setup cost. Since sensor nodes are powered by batteries or through harvested energy, these techniques are important to prevent unintended interruptions in the operation of sensor nodes between two maintenance cycles. We describe a simulator that can estimatethe energy requirement of the sensor network. Using this simulator, we implemented an optimization framework based on Simulated Annealing to assign the best processor widths and ECC curves to each level of the network.