Extração de dados em redes de nós sensores sem fio utilizando robôs móveis

Abstract

The goal of this work is to propose methods for data extraction in wireless sensor networks using mobile robots. This approach represents an alternative to the routing algorithms used in sensor networks that try to minimize energy consumption associated with data transmission. The methods developed here are motivated by applications such as environment monitoring and inhospitable environment exploration. In addition to the data extraction task, the robots can be used for node deployment and re-calibration, data fusion and coverage and connectivity repair. Aspects related to robot navigation and environment exploration are addressed as well. Two methods for data extraction were implemented: i) Reactive Method in which the data collected in the sensor node is modelled as a potential function whose gradient attracts the robot and ii) Hybrid Method, that includes heuristics for the resolution of Vehicle Routing Problem (VRP), where the objective is to distribute a set of clients among a set of vehicles and create optimized routes in every group. These methods try to minimize variables as path trajectory, total task time, data loss, robots energy, etc. Both were evaluated in a simulated environment with 60 sensor nodes and up to 10 robots. Finally, numerical results related to path trajectory, efficiency, data loss and data extracted by the robot are compared and discussed.