Optimizing the Localization of a Wireless Sensor Network in Real Time Based on a Low-Cost Microcontroller

In this paper, a low-cost microcontroller-based system that uses the pedometer measurement and communication ranging between neighboring nodes of a wireless sensor network for localization is presented. Unlike most of the existing methods that require good network connectivity, the proposed system works well in a sparse network. As the localization requires solving of nonlinear equations in real time, two optimization approaches, namely, the Gauss-Newton algorithm and the particle swarm optimization have been studied. The localization and optimization algorithms have been implemented with a microcontroller. The performance has been evaluated with experimental results.