Multicriteria large scale heterogeneous sensor network deployment framework

Optimizing the deployment of sensors is critical to the success of sensor networks. The development of a framework to manage this multicriteria problem is required for any deployment larger than a few sensors. In this dissertation, a frame-work is developed to advance the current state of art by integrating Native American mapping and genetic algorithm heuristics to find sufficient solutions to large scale heterogeneous sensor deployment problems.;The framework provides a layered architecture. The bottom layer consists of the Native American hunting heuristic, which is the mapping, target points, and deployment points. The top layer comprises an abstract representation of the problem to be solved and the necessary constraints. The interconnected patterns are defined in layered dimensions. We explore integrating a genetic algorithm for the multicriteria optimization problem solving aspect. The genetic algorithm uses the map to optimize the population of candidate solutions. Genetic algorithm rules are defined, to guide the designer in applying the evolutionary algorithm. Similar to how a hunting leader may have optimize the positioning of braves in the forest, the genetic algorithm approach pursues sufficient solutions for the deployment of sensors.;This evolved multicriteria framework is a decision support system for studying complex deployment problems. We developed the sensor network investigation toolkit to provide the designer with a tool to manage and tune the system. Several visualization tools are discussed allowing for insight into the characteristics of the experiments. We conclude with a summary and a review of future research items.;Since the initial documentation and discovery, the Native American hunting heuristics captures the complexity of many mapping concepts in a manner that is both simple and efficient. By following simple rules that require only minimal memory and complexity, large groups of hunters can coordinate in a robust manner that results in the whole tribe of hunters scouting and monitoring in a coherent, synchronized way that is capable of reacting to and covering an environment. Taking inspiration from the "rules of thumb" that evolved naturally from Native American societies, our sensor deployment problem framework is modeled based on the formulation of these rules.