A model-based approach to the design and implementation of Computational Sensor Networks (CSNs) is proposed. This high-level paradigm for the development and application of sensor device networks provides a strong scientific computing foundation, as well as the basis for robust software engineering practice. The three major components of this approach include (1) models of phenomena to be monitored, (2) models of sensors and actuators, and (3) models of the sensor network computation. We propose guiding principles to identify the state or structure of the phenomenon being sensed, or of the sensor network itself. This is called computational modeling. These methods are then incorporated into the operational system of the sensor network and adapted to system performance requirements to produce a mapping of the computation onto the system architecture. This is called real-time computational mapping and allows modification of system parameters according to real-time performance measures. This book deals with the development of a mathematical and modular software development framework to achieve computational sensor networks.
This text proposes a model-based approach to the design and implementation of Computational Sensor Networks (CSNs). This high-level paradigm for the development and application of sensor device networks provides a strong scientific computing foundation, as well as the basis for robust software engineering practices. Building upon a model-based approach the text discusses computational modeling of sensor networks and covers real-time computational mapping that allows for modification of system parameters according to real-time performance measures.
Drawing upon years of theoretical development and practical experience, and using numerous examples and illustrative applications, Thomas Henderson covers the sensor network as a computational science tool and describes:
A computational framework for sensor networks;
Emergent algorithms for cluster formation, coordinate frames, gradients, reaction-diffusion pattern formation, and level set computation in a sensor network;
How CSNs can be used to develop models to probe the structure of the sensor system itself by solving inverse problems based on strong models (e.g., PDEs) of the physical phenomena;
How CSNs can be used to model sensor network systems, including surveillance, environment modeling, structure monitoring, and biomedical monitoring;
Computational Sensor Networks is a must have book for sensor network application engineers, computer engineers, computer scientists and those involved in the development, design and building of sensor networks in an industrial, research and an academic environment.
