A Wireless Sensor Networks' Analytics System for Predicting Performance in On-Demand Deployments

The need for advanced tools that provide efficient design of on-demand deployment of wireless sensor networks (WSN) is critical for meeting our nation's demand for increased intelligence, reconnaissance, and surveillance. For practical applications, WSN deployments can be time consuming and error prone since they have the utmost challenge of guaranteeing connectivity and proper area coverage upon deployment. This creates an unmet demand for decision-support systems that help manage this complex process. This paper presents research to develop a system for predicting optimal deployments of WSN. Specifically, it presents results of image processing algorithms for terrain classification, results of modeling WSN signal propagation under different terrain conditions, results of optimization and visualization techniques for high-dimensional deployments, and system architecture for efficient integration and future deployment. Results show a feasible approach that can be used to automatically determine areas of high signal obstruction-which is essential to estimate obstruction parameters in simulations-and mapping of accurate WSN path-loss models to enhance the overall decision-making process during predeployment of large-scale WSN.